Yamaha NS1000M - Tweaks [Archive] - The Art of Sound Forum

Qwin

29-01-2015, 18:47

http://www.jkwynn.co.uk/Project_Images/Yam/yam_01.jpg

These are the new arrivals - the legendary Yamaha NS-1000M (Monitors).

Set them up in my not ideal locations, but there's not much I can do about that, short of moving house.

They seem to sound as folks have described for examples of this age, with slightly forward mid and treble units, which are ruthlessly honest in their rendition and with a soft'ish bass. They are thirty odd years old and I suspect there are some tired parts in there somewhere. I aim to get them sounding as good as new and deserving of their reputation, but I must finish the Celestion 66’s first. :rolleyes:

http://www.jkwynn.co.uk/Project_Images/Yam/yam_02.jpg
A quick and simple first task is to change the spring clips that are used to connect the speaker wire, for some decent binding posts.

http://www.jkwynn.co.uk/Project_Images/Yam/yam_03.jpg
I came across a Japanese seller on Fleebay that was selling these ready to fit sets of binding posts, that are mounted on a Yamaha compatible fixing plate, seemed ideal. On arrival I was disappointed to find they were made of white metal. I had on hand a very similar looking free standing set from another supplier that have gold plated solid brass contacts, they were so similar that it was easy to swap them over on the mounting plate.

http://www.jkwynn.co.uk/Project_Images/Yam/yam_04.jpg
That's the binding posts fitted, I don't believe in using expensive connectors but I do expect them to be brass under the inevitable Gold flash plating. These will make a much better contact than the original spring terminals.
Something concerned me when I was fitting these, the cables from the connectors to the x-over board are single strand bell wire, I am not kidding! I will get round to sorting that along the way.

http://www.jkwynn.co.uk/Project_Images/Yam/yam_05.jpg
Here is a closer look, for those that are not familiar with this speaker, it is a three way sporting a 12" paper Woofer in a sealed box of solid construction. The Midrange and Tweeter are Beryllium Copper domes and were state of the art at the time of manufacture. The large White Yamaha Logo is missing from the space to the right of the tweeter (both speakers), there is no sign of it having ever being applied. I think it was screen printed directly on to the cabinet as it doesn't look like a sticker or transfer on other examples I have looked at. No sign of it and the black wood grain finish looks perfect with no sign of over painting etc, it's a mystery. I think the logo was a bit large and did nothing for the looks so I am not that bothered at the moment. I will just enjoy them for now while I finish fettling the 66's.

walpurgis

29-01-2015, 18:59

Soft sounding bass is not something I've associated with the NS-1000. They should sound very tight and punchy, but they do require an amp that drives them firmly. I'd go so far as to say that some of the tightest bass I've heard has been from these.

They look like nice examples.

Qwin

29-01-2015, 19:45

They are not in bad nick, a few small marks thats all.

The 2 parallel caps on the bass low pass are 94uf (2x47uf) and are the only electrolytics on the crossover and I think renewing these very old parts would put some kick back into the bass, which is otherwise quite deep and well controlled. I renewed the same components (by location) on my Celestion 66's and it tranformed the bass on those to another level. The rest of the caps used are metalised paper in oil and have a very long life, though they are not the best sounding devices by todays standards. I will get round to removing the small rear panel on which the x-over is mounted for an inspection, I think you can check the x-over in this way without having to removing the bass drivers.

I have heard they are very picky about amps and my Sugden A21a has probably not got the current reserves to power the bass, which I understand has a fair impedance rise as frequency falls.

Luxman 70's SS and Yamaha B2 etc also valves are supposed to work well.

AlfaGTV

29-01-2015, 21:07

Congratulations on a pair of excellent speakers!
My pair won't be going anywhere soon! ;)
But i dont agree with you on the "soft bass" comment as these are tightest punchiest speakers ever played in my listeing room! (And there have been some, Audio Physic Tempo, Monitor Audio Platinum PL-300, Dali 40SE, Avalon Idea/NP2.0, Von Scheikert VR2 etc)
They do benefit from an amp with a high damping factor though, at the moment mine are powered by a Karan KA S270. Not the flimsiest amp i assure you! :)
Will be following this thread with great interest!

Qwin

30-01-2015, 08:42

Mike - By reputation the bass should be stonking, and I have read much about careful amp matching, so I am taking that into consideration.

I still think the 30 year old electrolytics in the bass filter will be holding things back, they are cheap to replace and keeping to electrolytics will not change the voicing, just put it back to where it was from new. As I mentioned this had a dramatic effect on my Celestion 66's and several people commented on how good the bass was on these, including Marco who had his own unmodified set as a reference.

I will get round to changing these caps in the near future and report back on how they effect things, electrolytics of this age really do need replacing, if only to safeguard the drivers.

YNWaN

30-01-2015, 08:47

The badges were self adhesive printed metalized plastic film and pretty easy to peel off (I still have the ones I removed from mine).

Reffc

30-01-2015, 10:21

Nice find Ken and cracking speakers.

As you suspect, that softish bass may be partly due to aged 'lytics as the raised ESR has a similar effect to reducing bass sensitivity and reducing efficiency...considerably so. Replace those with like for like new ones and the bass should be transformed.

I'd be using minimum 1mm cross section, preferably 1,5mm for the bass connection wires (15 to 18 AWG). 18 to 20 AWG is more than adequate for the mid/Hf.

I know that Mark is to modest to mention it, but I have not seen a better or more professional restoration of Yamaha NS1000's than the ones Mark has done, so I'm sure that Mark may have some top tips for you. His attention to detail with things like the internal woofer brace (a la Roksan Darius....a beam and locking bolt/spike arrangement) is a good example of vibration control/transfer.

YNWaN

30-01-2015, 11:34

Thanks Paul, very kind of you to say so :). Yes, the bass could well be the two electrolytic caps aging - easy enough to change and to measure. I can't remember exactly now but I think I used 1.5mm cable for the mid and tweeter and doubled it up for the bass (I used a pro grade OFC copper - nothing exotic).

I was pleased with the brace I fitted to the cab to fit against the dimple that is in the rear of the bass driver. The idea was to mechanically ground and brace it to the rear corners of the cabinet (rather than the centre of the rear panel where there will be the greatest flex). Unfortunately it's not possible to compare the effects of such a change as it isn't removable and took a while to make and fit. It really does brace against the back of the driver though.

Having seen Ken's work on his Celestions I'm sure he will do an equally super job of his new Yammies :).

AlfaGTV

30-01-2015, 11:37

Good thinking Qwin, keep us posted on the progress, will you?
There are also some talk about the modifications to the x-over by the famed builder of speakers Troels Gravesen;
http://www.troelsgravesen.dk/Yamaha-NS1000.htm

Personally, i am less bothered with electrolytics that are coupled in parallell with the drivers, than the occasional one in series. Your trials may well prove me wrong though! :)

YNWaN

30-01-2015, 12:17

Troels' article on the Yammies is interesting but ultimately a bit of a tease as, although he seems to give a lot of detail, he doesn't actually give enough to emulate what he has done (it is a commercial venture after all so not surprising). The biggest change he makes is to alter the bass alignment from a 2nd order slope to 3rd order (as he describes it); however, it isn't really third order in the true sense - more a version of. The bass on mine is actively driven by an amp with a digital front end and it is easy to alter the crossover. I experimented quite a bit with different alignments and did settle on 3rd order.

Qwin

30-01-2015, 13:08

Paul - yes I have seen Marks build thread on another Forum, like you say fantastic build quality and attention to detail.

Mark - Thanks for the heads up on the stickers.

Mike - I find Troels x-over builds a bit of a gamble. He seems to build whatever the program suggests and does not experiment much beyond this. Programs only get you in the right ball park and a lot of tweaking by ear and measuring is usually needed after that. Also he didn't even realise that the stock caps (7x 3.5uf) in the mid range and the single 2.7uf in the Tweeter filter are not electrolytics, which doesn't inspire confidance. There is the point allready made that he is in the business of encouraging the purchase of Jantzen parts, for whom he is a major share holder if not out right owner.

I agree the wire used internally could do with upgrading, especially from x-over to bass driver and the wire from binding posts to x-over is a joke, one strand of copper as thin as that used on telephone or CAT5 twisted pairs, what were they thinking?

YNWaN

30-01-2015, 15:30

Oddly, someone on another forum contacted me only a couple of days ago about my opinion on some replacement copies of NS-1000M badges that were for sale on eBay (not by me) - they were £30 a pair and looked to be extremely close to the originals. They were laser etched which is how I did my own.

Qwin

30-01-2015, 16:17

These may be the badges which I also spotted:

http://www.ebay.co.uk/itm/271745481141?_trksid=p2060778.m1438.l2649&ssPageName=STRK%3AMEBIDX%3AIT

I might just stick a round tuning fork logo on there, looks more refined and there are many of them around in the motor bike section like this:

http://www.ebay.co.uk/itm/291258532452?_trksid=p2060778.m1438.l2649&ssPageName=STRK%3AMEBIDX%3AIT

YNWaN

30-01-2015, 17:02

Yes, that's them - as you say, plenty of the three tuning fork logo badges in the motorcycle section.

Qwin

20-03-2015, 18:34

http://www.jkwynn.co.uk/Project_Images/Yam/yam_06.jpg

I took the small rear panel off the speakers to reveal the crossovers. The two large black caps (47uf) are aluminium electrolytics, the rest are film caps, metallised paper. All the silver caps are the same value (3.5uf), there is a single value used parallel in the midrange bandpass filter and six coupled together along the front for the series cap total of 21uf. But where are the coils you might ask?

http://www.jkwynn.co.uk/Project_Images/Yam/yam_07.jpg

Four coils are mounted below the PCB with hot melt adhesive and are connected to solder tags along three of its edges. These connections have all to be de-soldered before the board can be removed to allow access to the rear face for removal of the capacitors.

http://www.jkwynn.co.uk/Project_Images/Yam/yam_08.jpg

The connections for the coils were all unmade and the board removed. All the 3.5uf caps have been removed, the ones that were mounted vertically in the moulded plastic pockets were a pain to take out, as the hot melt glue used to fix them did not want to move. It’s now ready for me to start fitting the new caps. :)

Qwin

20-03-2015, 21:56

http://www.jkwynn.co.uk/Project_Images/Yam/yam_09.jpg

The caps have been fitted, Ansar Supersound (PP) for the midrange bandpass, a Mundorf Supreme (PP) for the series tweeter and Mundorf Ecap (Lytics) for the parallel pair in the low frequency filter. The Supremes are a bit pricey, but I only needed a 2.7uf so not too painful, its the big one, with the two Mundorf Lytics behind it. The three small Ansars on the right are 1.5 + 1 + 1 = 3.5 (3.48 actual). The Ansars on the lower left should make 21uf and are 8 + 6.8 + 6.8 = 21.6 (21.29 actual). The Lytics 47 + 47 = 94 measured 95.4uf. The 2.7uf Supreme measured 2.71, so all values are very close to ideal.

I replaced the bell wire between the binding posts and the x-over with van damme 2.5mm square OFC, I soldered it to the back of the posts.

I left the internal wire as standard, as best as I could measure, the bass driver is wired with 15 AWG which should be up to the job.

AlfaGTV

21-03-2015, 09:37

Great job! But what you waiting for dude? How do they sound???? :)
Regards Mike

Qwin

21-03-2015, 10:45

Doing the other crossover now, just took a coffee break.

It should all be finished by the end of the day and I will give it a few days before I comment to allow things to settle.

The Mundorf Supremes and Mundorf Ecaps were pre burned for a week, but the Ansars were fitted more or less when they arrived, I had to order a couple of extra 6.8's as one was at 4% tolerance, and the other three were better than 1% and almost exactly the same as each other (within 0.02uf). One of the two additional caps also fitted into this tight match so all was good in the end.

Well I better get back to it.:)

Qwin

21-03-2015, 16:12

Well thats the other one back together.

Got them playing to get some hours on them, all the solder joints where disturbed on the boards, these only take a couple of hours to bed in. The Ansar Supersound will take the longest as the other caps, having been pre burned, will settle in a couple of hours. I have found Ansars burn in pretty quickly, 30/40 hrs usually does it, the worst I came accross was Sonicap Gen1 which didn't reach their best till 150 hrs, some recon it takes 180 hrs, but I couldn't find any improvement after 150. The sound I can hear playing in the back ground is only slightly different to what I was hearing before the re cap, but in a good way. :rolleyes:
I will give a full report in a few days.

AlfaGTV

21-03-2015, 19:35

Excellent, looking forward to it! :popcorn:

Qwin

25-03-2015, 18:44

Listening results for the Re Capped speakers.

I listened to many tracks on a Sugden CD21 and on my Turntable, via a Sugden A21a Integrated Amp and also through a Croft Micro 25 Pre/Phono and Proton AA-1150 Power Amp.

As a starting point let me try and describe how the NS1000M sounded before any changes:
I found there was a tendency to favour the upper frequencies, with the midrange and tweeter sounding forward and slightly edgy with the lower frequencies sounding slightly detached and lacking a bit of energy. I used the level controls to compensate and for both upper drivers adjusted to ten to the hour; 12 o’clock being flat. This is what a lot of people end up doing and probably fuels the comments that careful amp matching is important with these speakers to avoid brightness. I would rate my kit as fairly neutral in presentation, so will not comment further on that, only to say that my gear/speakers seemed to fit this trend.

After the re cap compared to the original sound
The results were not as dramatic as when I re capped my Celestion 66’s, I think this is because the 66’s use quite a few electrolytics which would have deteriated to quite an extent, hence the big improvement. The NS1000M use film caps (metalized paper) apart from the large value Lytics used in parallel in the bass filter. These film caps have a long life so will not have deteriated to any great degree, but are not quite as good as modern Polyprops. I think this is what accounts for the smaller impact the changes made.

After 7 hours playing
The sound is generally slightly warmer now but the mids are edgy and timing is off which it never was previously. Listening to one of my killer tracks for timing, Gershwins Rhapsody in Blue by Pittsburgh Symphony/Previn on Philips, the fast percussive piano became a jumble of sound and the tune was almost lost. After 40 hours, maybe less the sound had smoothed out and timing had improved a lot. I didn’t listen in between so couldn’t say precisely at what point things improved or if it was gradual.

After 60 hours compared to the original sound
I now have both the level controls set to flat.
There is no longer any harshness in the treble or prevalence towards the upper frequencies.
The mid range is warmer/richer sounding than before.
Overall, a more balanced sound, with no area standing out or dominating.
Trumpet and Sax are notably sweeter with no sign of harshness.
Piano notes have better timber and seem more realistic.
Bass has a bit more of a punch and better grain/texture.

On the bass, I had read where one owner had taken all three sheets of wadding out from behind the bass driver and he thought it improved the sound. I took only one sheet out and found the upper bass level was slightly more elevated but it was a bit thin sounding to me. I put the sheet back in and felt it sounded richer/fuller so left it at that.

The improvements were not night and day but overall the presentation is slightly more even handed and I would say more musical, from what is basically a studio monitor, the later being a trait which monitors are not renowned for. I would speculate that they are also more amp friendly now, though they have lost none of that crisp top end, which the Beryllium drivers are capable of delivering. The Mundorf Supremes are doing a good job there and over all I would say the level of detail has improved slightly, but like I mentioned, nothing dramatic.

So a slight improvement in most areas with nothing negative to report.

I am very happy with the way this went and will not be tampering any further.

The following link details the parts I used in the re cap and also the measurements I took from the caps taken out, plus measurements from the stock inductors. These may be of use to anyone looking to attempt any x-over work:

http://www.jkwynn.co.uk/Project_Images/Yam/NS1000M_Recap.pdf

Also this link for the Owners Manual may be of interest:

http://www.jkwynn.co.uk/Project_Images/Yam/NS1000M Owners.pdf

For the time being, that is job done, I will be doing some minor cosmetic touching up when time permits and that’s about it.

Great sounding speaker and the sound I have been searching for. :D

AlfaGTV

25-03-2015, 21:14

Thanks very much Qwin! Excellent summary!

snorbans

27-03-2015, 08:13

Qwin

27-03-2015, 14:27

Hi Ian
Big Tannoys 12/15" are not laid back and can bop with the best of them, the 15" can be a physical experiance.

My comments were comparing my old original caps with the replacements and while the mid range is warmer and richer than it was I would not say it is as mellow as big Tannoy's.

The recapped NS-1000M are fast, tight and punchy with an absolute clarity and lack of distortion from the upper mids upward that I have not had before and only experianced at hifi shows from speakers costing many thousands of pounds. Being an infinate baffle design they work well in small rooms, with a lack of boom and room effects till the volume gets silly. Although an old 70's design they sound like modern speakers, they are very revealing and will highlight any weaknesses elsewhere in your system. I am going to have to upgrade my speaker leads for a start, I am using Van Damme Blue, which is a good entry level cable, but I borowed some Atlas that seemed to expand the frequency range at both ends so I will be looking to make a change soon. :)

walpurgis

27-03-2015, 15:30

I'm selling all my speakers to get a pair of ns1000 I am sold on the description of a detailed mid range and crisp HF. If that's the case, They would represent a departure and contrast with my Brit speakers. Never heard them but the other obvious step on is big Tannoys but again, from reading only they may be a bit refined and laid back. Would this be a reasonable summary of the speakers respective talents?

Laid back? Not on your Nellie! The proper big Tannoys can be devastatingly dynamic and certainly have as "detailed mid range and crisp HF" as the Yamaha NS-1000, albeit with a different tonal signature. The NS-1000s are a very fine speaker, but I find the Tannoy sound a bit more coherent and integrated. Nothing images like Tannoys.

Wakefield Turntables

27-03-2015, 15:41

My 15"MG sound very good listening to SOft Machine at the moment.:eek:

snorbans

27-03-2015, 17:02

Im in trouble with the Tannoy lovers aren't i. More listening required but for now this thread has sold me on the NS1000 - thanks!

Ali Tait

27-03-2015, 17:29

Laid back? Not on your Nellie! The proper big Tannoys can be devastatingly dynamic and certainly have as "detailed mid range and crisp HF" as the Yamaha NS-1000, albeit with a different tonal signature. The NS-1000s are a very fine speaker, but I find the Tannoy sound a bit more coherent and integrated. Nothing images like Tannoys.

Statics image better. :D

walpurgis

27-03-2015, 17:32

Statics image better. :D

Heard plenty, owned some. Not sure I'd agree Ali. Very different sound and presentation anyway (as you well know ;)).

(how's the new job by the way?)

Ali Tait

27-03-2015, 17:45

LOL, yes quite right, just pulling yer leg. :-)

Job is good thanks, if a bit on the manic side, the usual loads to do and very little time to do it..

Qwin

12-04-2015, 15:27

Hi Mark

To be honest, they are sounding pretty good as they are, the new caps nocked off any rough edges in the sound and I am enjoying what they do.

So I want to keep these pretty much as standard, the only plans I have, is to do a little cosmetic work on some paint scrapes and build some stands.

Would you be interested in selling your original Yamaha logo stickers?

Qwin

12-04-2015, 22:04

If you find them drop me a PM as I would be intrested in buying them, the replicas I have seen are all a lot thicker.

There is a guy selling originals in Australia as part of a set including the driver securing screws, but not come accross anything more local.

Qwin

09-05-2015, 21:40

http://www.jkwynn.co.uk/Project_Images/Yam/Belden_9497.jpg

http://www.jkwynn.co.uk/Project_Images/Yam/DIY_9497.JPG

I decided to replace the wiring inside the cabinets. I found many great reviews for the Belden 9497 (top picture) and it was used by Shindo Labs. Trouble is it is thin and only good for low powered valve amps into efficient speakers, it just wont handle larger current. I decided to make my own beefier version of this cable and purchased some 2.5mm square, marine, tinned, thin wall cable. I twisted this together using a cordless drill and as you can see from the two specs it is very similar, using similar diameter strands just more of them. I tried this out as a speaker cable running from the amp and it is very natural sounding with phenomenal bass, so I replaced the stock internal cables with this and while I was at it, bypassed the level controls and ran the cables along neater paths and not bunched together which could result in crosstalk.

walpurgis

09-05-2015, 22:35

Hmm. I've got a couple of pairs of speakers I'll be rewiring soon. I was going to try Nordost Wyrewizard Spellbinder. I know some people don't like their cables, but I've used Nordost in the past and had decent results.

Qwin

10-05-2015, 07:50

The beauty about this stuff is the wire is cheap at less than £1/m you need two pieces per run so just under £2/m, cheap enough to suck it and see. I purchased it off the Bay: http://www.ebay.co.uk/itm/251872885379?_trksid=p2057872.m2749.l2649&var=550764015035&ssPageName=STRK%3AMEBIDX%3AIT

The Beldon 9497 wire has been used on some pretty expensive gear, (give it a google), Ongaku owners etc and Mr Kondo is rumered to have used it on his own system, the Japanese love it. Do what I did and try this DIY version between Amp and Speakers first, to get a taste of what it does. It looks puney even in my 2.5mm version but produces richer and deeper bass than all my 6mm cables, go figure? The Belden spec sheet is available on line and there is nothing remarkable about its construction, it isn't even OFC and just PVC insulation, so my clone should be pretty close in performance.

I might try making some 4mm square as my main speaker cable.

Firebottle

10-05-2015, 09:31

... even in my 2.5mm version but produces richer and deeper bass than all my 6mm cables, go figure?

Go figure indeed.
Are the 6mm cables you compared to moulded together or enclosed, as opposed to separate runs of the marine grade?

:) Alan

Jimbo

10-05-2015, 10:13

Ali Tait

10-05-2015, 10:59

Yes, tried some thick cable on the OB's, ruined the bass, turned all loose and wooley.

Ali Tait

10-05-2015, 11:03

Trying some of this at the moment, sounds good-

http://www.ebay.co.uk/itm/291058740745?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

Qwin

10-05-2015, 14:02

Firebottle - The 6mm was Van Damme Blue which is in a jacket with only a slight twist and also the Van Damme Hifi which is shot gun style but with a 10mm spacing. The 2.5mm I made up as a twisted pair.

Ali - I had my eye on that add (12 AWG = 3.3mm[ ]) but he only has one legnth left and I needed two pieces - he bought it as a job lot, ex military.

Ali Tait

10-05-2015, 15:06

Aye, it sounds pretty good, shame he's only got 5m left, I could have rewired the OB's internally.

AlfaGTV

10-05-2015, 20:18

Give single strand installation cable a try! Preferably with a similar twist to ones you show pictured. That, will give you an eye opener ;)

AlfaGTV

10-05-2015, 20:40

And as an extra tweak, try one twisted pair per conductor. (In total two twisted pairs per speaker connection)

337alant

13-05-2015, 21:19

Plenty of decent wire here

http://stores.ebay.co.uk/Hollow-State-Electronics/cable-wire-/_i.html?_fsub=1754266013&_sid=22307073&_trksid=p4634.c0.m322

Alan

Qwin

28-05-2015, 17:57

Qwin

28-05-2015, 18:10

I've never been completely happy with the passive cross over, either before or after the re-cap. I hear distortion around the mid to tweeter x-over point. I've heard it on other examples of the NS-1000M, I had it on my Celestion 66's and hear it to some degree on almost every passive three way I have come across. It's mostly noticeable on voices, especially Female. I decided to try a simple analogue active filter circuit in an attempt to get rid of this problem, I'm using the ready made modules by KMTech. The passive is a simple circuit without any notch filters or zobel networks so a simple analogue filter using Op-Amps might work without the need for any frequency equalisation etc. The PDF file linked here: http://www.jkwynn.co.uk/Project_Images/Yam/Active_X-over_Layout.pdf
shows the layout I drew up, it has a power supply, also by KMTech, you just need to add a Transformer. The 30VA I used is overkill and one of 15 to 20VA would have done the job if I could have found one. I've tried to keep the AC path well away from everything else and to avoid the DC tracks crossing the signals, which I have kept as short as is practical. I am building high quality volume pots into the outputs to allow channel matching for amplifiers of different power ratings and sensitivity. The standard PCB mounted RCA sockets will be removed from the boards, I just used library pictures of the products for the layout.

http://www.jkwynn.co.uk/Project_Images/Yam/yam_15.jpg

It all fits into a 17" x 1U case which is 250mm deep, this is only marginally bigger than my Croft pre amplifier so it will sit under this on the shelf. I bought a ready made case and started work fitting the power supply.

http://www.jkwynn.co.uk/Project_Images/Yam/yam_16.jpg

This is the back panel and a trial fit of the fused IEC socket, the RCA sockets and the label I made to help identify the inputs/outputs. I went for a traffic light system, Red for Tweeters normaly at the top, Amber for Mids and Green for Bass normaly at the bottom. These all have Left and Right channel variants which are indicated by Red and White RCA sockets/plugs as normal. I will colour code the leads and amp channel inputs, also the speaker outputs to make it all fairly idiot proof. I don't want to accidentally plug a channel with power and frequency tailored for a bass driver into a tweeter, not if I can help it, so the clearer the better.

Qwin

29-05-2015, 17:28

http://www.jkwynn.co.uk/Project_Images/Yam/yam_17.jpg

I fired up the power supply to test the "Power On" LED. I put a standard LED across the whole of the dual voltage output (-9.8/0/+9.8) so about 19.5 volts dc. This will allow it to discharge both pairs of capacitors in the PSU when switching off. This calculates to approx 750R for the anode resistor, but with the 20 mcd LED used I thought this was too bright, and I don't like really bright lights on my gear. As can be seen in the centre of the picture I used a very small pot on the LED and turned it down till it was the brightness I wanted. I then measured the pot resistance (1k2) and soldered a resistor of that value on the long leg.

Qwin

04-06-2015, 14:40

http://www.jkwynn.co.uk/Project_Images/Yam/yam_18.jpg

The crossover modules arrived from KMTech. The Yammies cross at 500 and 6000 Hz, these boards cross at 501 and 6090 Hz which is the closest match using standard resistor/cap values. I'm not using the stock PCB mounted RCA sockets, so I've added extension wires to the boards made from wires I stripped out of some Klotz MC-5000. The input cables are made from the same, but I just stripped off the outer jacket and copper shield on these. As I will probably be using interconnects made from Klotz MC-5000 it made sense to use this on the internal connections.

Qwin

06-06-2015, 14:58

http://www.jkwynn.co.uk/Project_Images/Yam/yam_19.jpg

Got all the wiring done, crossover boards now have power to them and I have wired a couple of cheap 10k dual gang volume pots into the output to control mid and tweeter levels. I didn't see the point in splashing out on expensive pots till I have an idea of how good the crossover is. Must say I am a bit concerned about all those wires and the possibility of crosstalk.

I managed to do a basic check that things are working. I played each pair of outputs in turn through a stereo amp and a pair of passive speakers. This confirmed HF/Mid/LF output from the boards and that the level pots were wired up correctly.

I now have to construct some new rear panels for my Yamaha speakers with three sets of speaker connectors, I want to keep the existing panels with the passive crossovers attached un-altered, in case I ever want to revert to original spec.

Qwin

09-06-2015, 14:06

http://www.jkwynn.co.uk/Project_Images/Yam/yam_20.jpg

Made the rear panels for the speakers, the camera has made them look crooked and a bit scruffy but they are square and straight in the flesh and look much smarter.

I made suitable laminated labels for the panels and mounted the connecting sockets.

I only ever use "Z" type 4mm Banana plugs, so I'm not a fan of bulky multi function binding posts. The panel sockets I used (inset) stand only 5mm off the surface and are high quality items from "Hirschmann" they are solid Brass, Gold plated and rated at 32Amp. RS do them at £2.04 each with free delivery. ;)

sq225917

09-06-2015, 23:31

Looking good. what are you doing with respect to the mixed polarity of the drivers in the original crossovers? How's it affected by the new steeper slopes?

Qwin

10-06-2015, 07:39

sq225917 - The short answer is I don't know.

I'm very much treating this as an experiment with no fixed ideas about the outcome, if phasing is an issue I will try and deal with it at that point.
The effect of the steeper slopes is a suck it and see situation. These simple analogue circuits will not be a good match in a lot of cases, everything crossed that they will work well with the Yamaha's. When I complete the first rough build, over the next few days and able to give a full test, I'll report back on my findings. I'm sure plenty of people will be interested in the results.

sq225917

10-06-2015, 09:12

Count me as interested. The Tweeter in the original crossover is wired in reverse phase to cancel out the effects of the crossover slopes, so you might find with 4th order slopes that this is negated. I'm sure you'll work it through... Looking forward to seeing more.

Qwin

10-06-2015, 11:13

Yes, I'm starting with the wiring as standard, so just the tweeter with reverse phase.

I have the extra complication that my Croft Pre reverses phase but my Nakamichi Power does not (me thinks) so I will have to reverse all the wires going into the speakers to achieve the normal situation. :rolleyes:

Qwin

10-06-2015, 20:42

http://www.jkwynn.co.uk/Project_Images/Yam/yam_21.jpg

Got the new rear panels fitted to the speakers without any issues, everything lined up. I built them blind from some measurements I took the last time I had the passive crossovers out, so a bit of a relief.
I made the decision early on, not to do any cosmetic work on these speakers, till I had them sounding the way I wanted, lesson learnt from my Celestion 66 project.

http://www.jkwynn.co.uk/Project_Images/Yam/yam_22.jpg

Got everything wired up. I had to build a shelf that fits on the power amp so that the crossover can sit on top without interfering with the amp cooling. This location allows short and simple connection runs. I used Van Damme HiFi speaker cable for the Tweeter and Bass runs and VD Blue for the Midrange.

http://www.jkwynn.co.uk/Project_Images/Yam/yam_23.jpg

I said the location of the crossover allowed short and simple connections but this snake pit is still the result. To be fair I've just used whatever leads I had available, this is one of the down sides of going active, lots and lots of cables.

I fired it up, no puff of smoke and had music straight away but it sounded very dull, too much bass, I increased the mid and tweeter levels using the volume pots I had inserted into the crossover output and it was a lot better. From experience of op amp based phono stage builds I know these type of circuits usually take a day or two to settle in so I left it playing all afternoon. When I came back at tea time I was immediately taken by how good it was sounding. I only have some very cheap pots controlling the levels, £3 a piece but even so, this was already sounding better than the passive circuit. The upper mids were cleaner and female vocals were at last sounding distortion free at the crossover point. Bass is strong and deep and timing is very good, the crossover slopes seem to work very well, piano scales sounding seamless. I have the tweeter wired in reverse phase as is normal, I tried changing phase but there was no dramatic difference, the original possibly sounding slightly better, I will come back to this later.

Even after such a short time I can easily see from the positive results that it is worth continuing with this build and investing in some better pots and making some decent short interconnects. :)

AlfaGTV

11-06-2015, 18:26

Very interesting Qwin, thanks for sharing!
Regards Mike

Qwin

22-06-2015, 14:12

http://www.jkwynn.co.uk/Project_Images/Yam/yam_24.jpg

I established the crossover works well and all the parts now have screw fixing holes in the right places, even the new and better quality pots have a location bracket riveted into place. So it was time to completely disassemble things and give the case a coat of paint. I used my favourite single coat finish, which is Black etch primer, good adhesion to tricky surfaces like Aluminium or Zintec, hard wearing, corrosion resistant, good up to 200 deg C and dries to a nice satin finish, what more could you want?

sq225917

22-06-2015, 15:21

Time to get the measurement mic out, that's looking good.

Macca

22-06-2015, 16:05

[

http://www.jkwynn.co.uk/Project_Images/Yam/yam_22.jpg

:)

You seem to have the speakers almost soffit mounted in that photo. Are you thinking of going the whole hog with that?

Reffc

22-06-2015, 17:14

Nice job (as usual) Ken. RE Phase, the ear isn't very good at picking out absolute phase changes, only relative phase, so if the phase differences were significantly out between tweeter and mid, you might notice, except that 4 pole slopes (electric) mean that any overlap is constrained to narrower bands, hence it can be more difficult to spot.

Using a measurement mic (gating the response to remove the room reflections) will show you which way is the better way if your ears don't quite. If you find a dip in response at crossover one way, and a flat crossover response the other, then whatever phase that is will be correct. The crossover points chosen for the original will have taken into account the filter slopes V's driver break up (top end) so whilst you should get a text book 24dB/Octave for the tweeter high pass filter, the same may not be so of the mid, hence if the slopes are a fair way out, the only remedy using the active boards fixed to 4 pole is to lower the crossover frequency accordingly. Same applies at the lower end for the bandpass (high pass part) filter. Only acoustic measurement will confirm although if you can find the driver data sheets, there will be clues there.

Yamaha chose the most sensitive part of human hearing (5KHz) for the upper crossover but that's not unusual as many 3-ways use a XO point between 4 and 5 kHz to limit any waist formation in the polar response. Using steeper slopes allows a safe lowering of that upper X/O point too if needs be. I'd be tempted to get that mic out and see what the raw response is from each driver before setting things in stone and investing any any better parts.

Qwin

22-06-2015, 18:23

Macca - The speakers are surrounded on all sides, which is not ideal but I have no plans to soffit mount.

sq225917/Reffc - Yes a calibrated mic/software would come in handy about now, when I get into the final stages of setting this up it may be necessary.

Note: the NS-1000M cross at 500/6000Hz (501/6090 using these boards).

Qwin

24-06-2015, 17:43

http://www.jkwynn.co.uk/Project_Images/Yam/yam_25.jpg

While I was waiting for parts to arrive I finished the shelf that sits on the Amp.
It now locates on the corners so it can't slip off and is Satin Black to match.

http://www.jkwynn.co.uk/Project_Images/Yam/yam_26.jpg

I reassembled all the parts into the painted case, didn't fit the front panel as I'm still working on it. The areas where the aluminium panels come together were left unpainted for earth continuity. The cheap Pots I used in the trial build have been swapped out for some TKD 2CP-601, 10k, Dual Gang, Log Pots. This pair of Potentiometers, including suitable shaft extension kits, cost more than the crossover circuits, but as they are in the output signal path, I thought it worth the expense. I had considered ladder rail stepped attenuators, or measuring the settings used on the cheap Pots and replacing these with good quality fixed resistors. I figured that the steps in the attenuator would have been too big for accurate driver integration and fixed resistors wouldn’t allow any fine adjustment for things like different room acoustics, a change of interconnects or Amp, so the pots seemed the best compromise.

The new Pots have made an improvement to the sound, if only slightly, they sound more mellow, the originals sound slightly sharp/grainy in comparison. The main difference is the channel matching which is a lot closer on the new ones. I measured to check and they are within a couple of percent compared to eight percent on the originals.

The next thing to try is a bit of op amp rolling. :eyebrows:

Qwin

26-06-2015, 14:49

It’s funny how your standards and expectations increase as you follow the upgrade trail, what was almost acceptable yesterday is not good enough today.
My stock NS-1000M sounded nice, but renewing the caps raised the bar, changing the brand of caps took it a little further and going active cured some of the minor niggles that had bothered me. Putting the active system under scrutiny now has me questioning whether the op amp’s used are the best match for squeezing the last bit out of the Yamaha’s?

Time for a bit of op amp rolling.

The KMTech crossover boards come with Burr Brown OPA2134 fitted as standard. I had been browsing through Rod Elliott’s site (ESP) in particular the Forum section relating to his 09 active crossover. This is also a Linkwitz Riley circuit using op amps, so has a lot in common with the KMTech design. A poster had used OPA2134 in his build and described the sound as……
“somehow the crossover was "muffling". Maybe that’s what "they" mean by the "laid back Burr Brown sound" – muffled”
He dropped in some LME49720 from Texas Instruments and commented that it brought back what he had been missing from his passive circuit……
“there seems to me to be a marked improvement over the OPA2134's in the HP (LP seems unchanged. Everything has smoothly sharpened up and subtle underlying detail and "space" has returned around the vocals.”

Generally the OPA2134 has no major vices but is reported around the forums as having a good warm bass and lower mids, if not the deepest bass available. An ok but not brilliant top end and a so so midrange that is detailed but lacks dynamics and has a poor sound stage and maybe a touch grainy. You have to be careful when reading comments for different applications, many of the views expressed are for use in headphone amps, but I must admit this is describing what I am hearing. The mids do seem to be the weak link and to me sound a little shut in and slightly grainy.

http://www.jkwynn.co.uk/Project_Images/Yam/yam_27.jpg

The comments by the guy on the ESD Forum, regarding the OPA2134 struck a note with me. Therefore his views on the LME49720 in a similar application to mine seemed quite relevant and the spec sheets do show very low figures for distortion and noise level. So I got a set of x10 from RS, they are not expensive, just £21 for the set including next day delivery. There are more exotic parts around but for the time being I will keep it simple and use a like for like dual channel in a dip8 packaged chip. Adaptors for mounting SMD’s and pairs of single channel devices opens up the options (and price point) somewhat and may be looked at along the way, if need be.

I changed all five op amps on each board, as much to make sure they were all burnt in as for any other reason, these devices tend to settle quite quickly, 5/10 hours over a couple of days in my experience. Using a mix of different types may be beneficial and could be tried later.

How do they sound?

I have to agree the mid range is much better, with space around instruments/voices and very transparent. Bass/lower mids is different but I will have to consider this further as to whether it is a better different, same goes for the top end, so more listening required. ;)

Macca

26-06-2015, 14:58

IMV it is worth improving the mid even if it is at the expense of the top and bottom. Obviously no compromise would be best but if you have to compromise the midrange is the worst place to do it.

Qwin

26-06-2015, 15:44

Couldn't agree more, The mids are now sounding very nice so I will work round that.
The bass and treble are just harder to assess as any changes have been more subtle.
It really is starting to sound extremely good overall though.

Qwin

27-06-2015, 13:59

I have had time to try different (read as all) combinations of OPA2134 and LME49720 op amps.

The LME worked best for mids and top end the OPA for bass, the best combination seems to be.......

Input buffer = LME

HP = LME

LP = OPA

BP = (LP) LME
.......(HP) OPA

Best phasing for the tweeter seems to be as standard, which means it is inverted compared to Mid/bass drivers.
I haven't made any measurements but the sound seems pretty good to me with no obvious problems.
It's now fairly well balanced tonaly and probably the best that can be achieved using these low budget op amps (£2 or less each).
To achieve greater levels of transparency, improved dynamics and better soundstage would mean using much more exotic (expensive) parts but it could be worth spending the extra on this basic design. These low cost boards are achieving remarkable results and are so much cleaner than the passive circuit at the Mid to Tweeter crossover region.

Qwin

06-07-2015, 15:11

http://www.jkwynn.co.uk/Project_Images/Yam/yam_28.jpg

I'm still at it and continuing with the op amp rolling, many of the ones I want to try are only available as SMD's so I bought some SOIC8 to Dip8 adaptor PCB's, warmed up the Iron and grabbed a magnifying glass to aid my getting on for 60 year old eyesight. :rolleyes: The op amps here are AD8066 which are supposed to have better deep bass than the OPA2134 being used at the moment, they are supposed to be good for the top end as well so two per channel required. They will be mated with LM4562 which are good for mids (band pass) and the input buffers so six of those needed. I have them running in at the moment, folks on various forums say run them in for a couple of days, which I can confirm seems to be about right for these devices to settle in.

I have done a lot of comparisons on tweeter phasing and decided the best results are with the tweeter in phase with the other two drivers. This is the opposite to the passive set up but seems to give a cleaner sound once the levels have been adjusted to suit, which I hadn't tried previously. I also found some documentation on line that said the circuit I am using was intended to work this way and phasing had been accounted for in the design. The difference was only small but nice to get it right. :)

Qwin

11-08-2015, 16:50

As you will have gathered, I get very fussy when it comes to the finishing details on my projects. From the start, I wanted to make this crossover case look as if it was part of the original speaker system. In trying to achieve this, I needed to apply some lettering and the Yamaha Logo on the front panel and to a high standard. The only way I could think of doing this, using my limited workspace and equipment, was to screen print the lettering onto the front panel.

Its 38 years since I did my last screen print at Hartlepool College of Art, so a bit of research was needed, YouTube was my friend here. The basic process is as I remembered, but the method of getting the design on to the screen has changed, so I had to pick up some new skills along the way.

http://www.jkwynn.co.uk/Project_Images/Yam/yam_29.jpg

This is the printing jig I made, I had to use quite fine 120T mesh on the screen due to the detail in the lettering and logo. I already have a UV light box from my efforts in making my own PCB’s so I just needed to apply the Photo Emulsion to the screen and crack on.
It was a bit of an extravagance going to all this trouble for a single pull, but I wanted to perfect the process for other reasons unrelated to hifi. I didn’t want to purchase expensive inks for this one off however, so I just used, Wilko non drip White gloss paint to print with.

http://www.jkwynn.co.uk/Project_Images/Yam/yam_30.jpg

I think the end result is quite good and I managed to get the registration right first time.
The front panel has been fixed in place and the shaft extension rods fitted to the attenuation pots. In terms of pure construction and finishing, the case is now completed and I only have the circuit to finalise.

I’ve spent a few days swapping op amps in and out. The crossover still has no cover fitted and is located out into the room, so access/swapping is simple. I power down the unit and pull the op amps out of the sockets with a removal tool and clip the new ones straight in. I use a suitable grounding wrist band connected to the chassis, just in case I pick up any static from the carpet. I have all the things needed, including the packets of various op amps, laid out ready next to the unit, so swapping a pair takes under a minute.

Op Amps tried (comments in brackets are how they have been reviewed by others)

OPA 2134 (Full rich bass if not the deepest, soft and warm mid/treble)
LME 49720 (Transparent, good mids not bad all round)
LM 4562 (Transparent, good mids not bad all round)
AD 8066 (Good deep bass and clean detailed top end but not as good for mids)

Low Pass
I tried the AD8066 in the LP filter and while I can agree it gives a deep and transparent bottom end, it seems to affect only a narrow band and has no impact on upper bass/lower mids. To me the OPA2134 is almost as deep but is slightly fatter and richer and has a broader influence on upper bass/lower mids which I liked.
Over all I preferred the OPA2134 for bass compared to any of the others, drum solos seem more realistic and the richness that carries over into the mids works well on acoustic stringed instruments.

Band Pass
(LP element <6090Hz passing) The LM4562 worked best, it is very transparent and clean sounding for the interface with the tweeter.
(HP element >501Hz passing) The LME49720 worked best here. I know some say the LME49720/LM4562 are the same chip renumbered, and the spec sheets are identical. I am sure they are fully interchangeable in all circuits, but I find as others do, that they sound different. The LM4562 is slightly cooler and sharper sounding, so for the interface with the rich bass the LME49720 and it’s slightly warmer presentation Dove tail better.

High Pass
The AD8066 does work very well in this filter, cymbals and triangle are very clean with a long decay and a nice shimmer on cymbals. I also found that the upper range in Piano, located more in the upper mid region, sounded more percussive and realistic using this op amp, compared to the others.

Input Buffer
This seems to have an overall effect across the range. The LM4562 was very transparent, the AD8066 even more so, but upper mids/ treble became too dominant and tonally un-realistic. The slightly warmer LME49720 worked very well here also and gave the best tonal balance when used with the choices made for the individual filters.

It’s worth pointing out that changing the fitted op amps, with similar dual channel, unity gain stable devices, does not affect the cross over points in this circuit, it only effect’s things like transparency, timing, soundstage and tonal colour. So similar to valve rolling, you can tune to your own preference, but also like with valve rolling, there is a madness that can set in when searching for that illusive ultimate combination.

The combination I have settled on for now, is very good and suits my preference for realism over musicality, but it is never the less, still very easy on the ear, so the unit can be fastened up and I will live with it for a while. :)

Firebottle

11-08-2015, 17:45

Excellent efforts Ken, really looking good :clapclapclap:

Jimbo

11-08-2015, 19:45

Hi Ken, are you bringing the yammys to NEBO6 ?

How you getting on with the Croft pre?

Qwin

11-08-2015, 20:47

RichB

11-08-2015, 21:17

Nice work as always Ken.

Having heard the Yamahas in relatively unmolested condition I'd be interested to see how far you've taken them.

Beautiful finishing on them crossovers btw. Mrs B complemented your screen prints, she's a textile graduate and had to do loads of it at college. I remember all her stuff around the house.

R

Jimbo

12-08-2015, 11:37

Firebottle - Thanks Alan.

Jimbo - Hi James, yes I intend taking a complete vinyl system, Yammys, Croft and Nakamichi Power. The Croft, prety much like the Naka just gets on with the job, I will get round to some valve rolling one day, but lots of other tasks and upgrades are more critical as it is very good as is and I like what the phono stage does as well. I need to make up lots of Klotz interconnects for the active connections and build a pair of speaker stands for NEBO so keeping busy.
It would be intresting to try your modified Croft in my set up, you going to Nebo6?

Hi Ken,

Yes I am coming to NEBO6 but will probably be bringing the Firebottle KIN preamp with me - definitely worth listening too! Think a few other folk may also want a listen - it is MM and MC so will hopefully find a system I can slip it into for everyone to get to hear.

Qwin

12-08-2015, 22:26

You are welcome to try it through my system if you would like James.

That offer is on the table, what ever else you try, be intersted in how that combo works myself.

Jimbo

14-08-2015, 18:09

You are welcome to try it through my system if you would like James.

That offer is on the table, what ever else you try, be intersted in how that combo works myself.

Cheers Ken. I can tell you the Firebottle KIN is indeed the best phono stage Alan has ever built. May well bring the Croft series 7 amp as well!:)

Qwin

02-10-2015, 19:39

Back to the Yammies

Tried several opamps in the bandpass filter, they all seem to lack transparancy to me, otherwise I am very happy with the sound, mnnn.......

Maybe the buffer is effecting the transparancy of everything else down stream and it aint up to the levels I expect? (LME49720NA fitted)

I'm considering the LME49720HA TO-99 (metal can) version but also taking a long hard look at the Burson V5 Discrete dual channel Opamps.

Too expensive to use throught, but I would consider them for this one location on each board if it helped.

I asked a few technical questions at Burson, as to their suitability for my application, the time difference is delaying a response as they are Aussie's.

I'm tired and rambling, just really jotting down my thought process.

Haven't even considered how this will effect rediness for NEBO. :eek:

sq225917

03-10-2015, 00:16

Ken it might just be that the actual crossover is just a bit ho-hum and that no amount of opamp rolling will fix things. There's a good chance that each different opamp will have a different optimum requirement for local decoupling. The lm4562 variants for example defo need different decoupling to the bb opamps.

Have you considered other crossover designs?

Qwin

03-10-2015, 10:30

Hi Simon,
you’re absolutely right of course, I am assuming the basic design is good and putting my faith in that. It is so nearly perfect for my requirements, a bit more transparency in the mids and I would be a very happy bunny.

I looked briefly at a few designs including the build on diyaudio and Rod Elliot’s design, before going for the KMTech boards.

The KMTech boards have a different layout, but the circuit design is the same as that used in Rod Elliot’s P09 project linked here: http://sound.westhost.com/project09.htm

In the Forum for Rod’s project, one user tries many opamps and concludes that for overall transparency the Burson discrete is by far the best solution, but at a high cost.
He instead uses pairs of OPA627 on adaptors for the buffer and to great effect. The thing is x4 OPA627 (single channel) and the dual adaptors needed, runs at about £65 and the Burson discrete at about £87 for a pair of dual channel. At those prices I would go straight for the Burson discrete.

The way the design is laid out means the tweeter signal passes through 2 opamps and the mid and bass signals pass through 3 opamps. The bass signal goes through the LP element of the midrange band pass filter first. If I were to replace the Buffer and LP component of the band pass with Burson discrete, then each filter element would have just one general purpose opamp in its path. I am gambling that this would be a much cleaner sound, especially for the midrange where transparency effects are more noticeable.

Overall, changing opamps has a large effect on the sound produced, it is not subtle. I have already tried four different opamps in the buffer location and the effects on transparency are very marked, hence my focus on this area.
The caps used on the boards are Polyester. I was also wondering whether these or the resistors (1% metal film) were up to the job?

As for individual opamp decoupling requirements, that is getting beyond my experience and comfort zone. Note there is a small cap (100nf) across the +/- supply pins (No 4 and 8) on each dual channel location.

I’ll think about things over the weekend.

Firebottle

03-10-2015, 11:01

The caps used on the boards are Polyester. I was also wondering .......

Ken, at the risk of teaching Granny to suck eggs it's usually accepted wisdom that Polyester can be bettered by Polypropylene or Polystyrene.

With the values I suspect you have in your circuit I would suggest going for polystyrene. :)

Qwin

03-10-2015, 11:16

Hi Alan

You are reflecting my thoughts exactly, trouble is there is damn little room on the boards, I would have to look carefully at sizing and lead spacing of components.

My current thinking is to use the existing boards with the opamp sockets for development purposses, then buy some empty boards and populate them with better quality parts and leave out the DIP8 sockets, soldering the opamps direct to the PCB for the final build.

EDIT:

Just looked at some cap options, the ones in the signal path and controlling the crossover points appear to be 33nF and 3.3nF.

The Polyester fitted are 5% tolerance.

Polyprops small enough eg. Wima MKP2 series are 10% or 20% dependant on value.

Polystyrene with axial leads and mounted vertically could fit, these are available at 1%.

I would favour the Polystyrene, due the much tighter tolerance, for holding the crossover points and giving better matching between filter elements and channels.

Qwin

05-10-2015, 11:34

Well I've gone and done it!

Just ordered x4 of these dual channel Burson V5 discrete opamps. :)

http://www.ssaudio.com.au/wp-content/uploads/2015/09/SS-Opamp-V5-S41.jpg

Have to wait and see how long they take to get here from down under.

I'm pretty sure I will be leaving the AD8066 in the HP filter for the tweeter, it seems to work well there.

The Burson discrete will drop into the buffer stage and BP (LP) positions, so that just leaves the BP (HP) and the LP filter locations.

I will try AD825 for the LP (bass) filter and see how that sounds, it has a reputation for deep articulate bass, we will see.

I have a few things to try in the remaining BP (HP) position so a bit of suck it and see, once the Burson have run in.
50/100 hours according to them. They also recomend putting a good quality MKP cap (10 to 100nF) accross the supply pins on their unit and not a ceramic cap which is often used, and only fit after running in? Don't get that, but they are the experts. ;)

Qwin

24-11-2015, 13:40

http://www.jkwynn.co.uk/Project_Images/Yam/yam_31.jpg

Second Build.
My active set up, for which I purchased the boards ready built, is much better than the stock passive crossover, that said, I have not been totally happy with the mid range, so its about time I looked at this. The caps fitted, 27 on each board, are mainly Polyester and the signal passes through 12 of these on each board. Polyester is not good for sound so they had to go.
Because of the number of changes involved, it was far easier just to start again, so I got a pack of 5 unpopulated PCB's from KMTech. I have replaced the 5% Polyester Caps with 1% Polystyrene's by LCR. I also replaced all of the resistors, which were of unknown origin, with PR9372 1% metal Film's by PRP. These were all supplied by HiFi Collective and can be seen assembled in the picture.

OK, so how does this effect the sound?

Initially course grainy top end with very little and I'll defined bass.
The top end smoothed out and the Bass was much fuller after about 10 hours and continued to improve after this to a lesser degree. The previous caps had focused on the top and bottom end with a slightly recessed mid range. The new caps give a much more even response across the range, with the mid range being more detailed and tonally rich than before. I am liking what they do a lot, small details I hadn't noticed are now highlighted, little breaths, additional backing singers, those kind of details and a blacker background.

I will re-evaluate the op amp rolling now to finalise the components.

AlfaGTV

24-11-2015, 16:49

May i ask what the cost was for the Burson OPAmps? I am aiming to upgrade my Karan KA L which uses a few opamps. A Karan kit of four goes for something like £800! Ithought perhaps the Burson would be a less costly but equally good alternative?
Mike

Qwin

24-11-2015, 17:14

Hi Mike,

I got four dual channel modules for a total of £156 including postage off the Bay.

They are even cheaper if you buy from Burson direct as the Bay takes a cut.

I have just ordered another pair this way, but I got a special deal, check out their new SS V5 which is what I am using, here:

http://www.bursonaudio.com/products/supreme-sound-opamp-v5/

Ali Tait

24-11-2015, 17:15

Nice one Ken, I replaced all the ceramic caps in my Denon head amp with polystyrenes, well worthwhile.

sq225917

24-11-2015, 19:23

Where suitable styrenes have always cleared things up for me, any small signal value under 100n and they are my goto part choice. Easy with the soldering iron though, the little blighters melt easily.

Qwin

04-12-2015, 15:25

Op Amp upgrade
The boards come with x5 OPA2134 fitted to each.
I found that with all five fitted, it sounded a little soft and lacked resolution, as the chips are all on dip8 sockets, I got into some Op Amp rolling.

The boards require dual channel unity gain stable devices, I tried the following in all kinds of combinations (Adaptors needed for SMD and/or single channel chips).

AD8066
AD825

Burson Supreme Sound V5 Discrete

LM 4562
LME49720 (NA)
LME49720 (HA – Metal Can)
LME49723

LT1358

OPA1612
OPA1642
OPA2134

I got some major improvements with *AD825 for Bass and AD8066 for Treble* (Also worked for Bass), with LME49720HA and LM4562 doing well for Mid Range Band Pass and Input Buffer. OPA1612 worked well for Mid Range, as did LT1358, which could also do well in the Bass filter.

Burson Discrete Op Amps
The whole thing moved to a different level when I tried the new Burson SS V5 discrete modules (Launched 22 Sep 15). The more of these I fitted, the better it sounded, especially when used for the buffer and band pass filter for the Mid Range (+ op amps at *). With this arrangement it produced a Blacker background and deeper well defined Bass. I got a smooth top end and a very detailed and natural tone to voices and instruments. These new V5 modules are very compact and when soldered direct to my PCB’s, which have 6mm mounting pillars, will fit in a 1U case with a few mm’s head room. Using the Burson modules has produced a seriously good sound with this circuit, which is now considerably better than the stock passive crossover, they gave a significant improvement over the other Op Amps tried.

http://www.jkwynn.co.uk/Project_Images/Yam/yam_32.jpg
Boards with Burson modules and new caps/resistors fitted.

KMTech also produce 2 way boards and versions with Balanced Input.
I suspect their boards won’t work with every type of speaker, but they are cheap enough to try it and see. If you have complex crossovers with notch filters and zobel networks, there may not be a good match, as they offer no form of correction, simple crossovers offer a better chance of compatibility. These active boards work extremely well with the NS-1000M and the Burson SS V5 Discrete Op Amps get the very best out of them.

Macca

04-12-2015, 15:50

Would these Burson modules work in a passive crossover, Ken?

Arkless Electronics

04-12-2015, 16:05

Nice work Ken. Glad to see you've had some worthwhile improvements from all that work.

Just one comment before folks start ripping all the caps out of their equipment (not aimed at you but a general N.B), Whilst polyester caps are not the best from a SQ point of view, any type of film cap is still much better than even the best electrolytic ;) Ceramic caps are great only if of NPO/COG type which are only available in small values, up to around 10n and usually pF values. Do not ever use the "high K" ceramic types (Z5U etc) in audio anywhere other than certain areas of a power supply!! Now they are worse than any electrolytic cap!

Arkless Electronics

04-12-2015, 16:06

Would these Burson modules work in a passive crossover, Ken?

Just checked the date..... wrong month and it's the 4th :D

Qwin

04-12-2015, 16:09

Hi Martin
No, the Burson discrete modules wont work in a passive circuit, they are replacements for Opamps which are active devices, meaning they require a power supply, as a simple explanation.

Thanks Jezz, fiddley job but it was worth it.

Just to clarify folks, the Polystyrene caps I used are good for small signal stuff like on this line level crossover or in a RIAA stage but no good for a passive crossover, they only go up to a few hundred nF for a start, so no large values available that would be needed in a passive crossover. :)

Macca

04-12-2015, 16:45

Just checked the date..... wrong month and it's the 4th :D

Hey if I knew I wouldn't ask and if I was making a joke it would be funny. Or funnier, anyway ;)

YNWaN

04-12-2015, 17:12

Made me smile :).

Arkless Electronics

04-12-2015, 17:32

Hey if I knew I wouldn't ask and if I was making a joke it would be funny. Or funnier, anyway ;)

wasn't sure but thought it was probably a wind up :)

sq225917

04-12-2015, 20:12

if they sound as good as they look you'll be right happy with them. Next step will be the fully active DSP route?

Ali Tait

04-12-2015, 20:31

How does it compare to what we heard at NEBO Ken?

Qwin

04-12-2015, 22:02

Simon - :rfl:, I'm trying to stay analogue, but when I get it fully sorted, Gordon (Halfway Tree) is going to bring a Mini DSP unit over and wire it in so we can compare the pros and cons of each method.

Ali - The mid range was a bit recessed at NEBO, its more forward in the mix now, there is generally a higher level of detail, a blacker background and a more natural tone to the mids. We're not talking Black and white, but still noticeably different and better in all the right places in my opinion. The third pair of Burson discrete modules are still running in but they allready sound better than the opamps they replaced. 10 hrs on them, with Burson reconing on 50/100hrs needed. The previous set seemed at their best after about 50 hrs and yes they improve a lot over that period so I have high hopes for when this latest set settle down.:)

Ali Tait

04-12-2015, 23:16

Cool. Look forward to your impressions to the comparison with the miniDSP units, that's what I use on my OB's.

Qwin

05-12-2015, 08:30

Don't hold your breath Ali, we probably wont get round to that till the New Year. ;)

Ali Tait

05-12-2015, 11:53

No rush. :-)

drSM

22-01-2016, 00:53

Hi Qwin
Super thread - from a base old pair of Ns1000Ms to passive X over upgrade to Electronic X over in a matter of months and quite superb analyses of the sonic benefits !
I am afraid u lost me once u got into the active electronic X over. Lovely work.
For the benefit of 'non electronic me' could u summarise what amplification u settled on.
Presumably 3 stereo power amps
I see a Nakamichi Power amp at the bottom of the stack on the floor.
What model is that?
While we are on Nakamichi do u have an opinion on the Nakamichi PA7 and its suitability to drive the NS1000(passive)?

btw i hve an unmolested pair of NS1000s that i love.
Have thought about upgrading the X over but dare not !
Ta
Shahrin

Qwin

22-01-2016, 09:59

Thanks for the kind words Shahrin.

The set up is like this:
The two outputs from my Croft Micro 25 Pre Amp plug in to my crossover box. Each channel is connected to a KMTech board, that splits the signal, without any further amplification, into Hi, Med and Low frequencies. So three line level outputs per channel making six outputs in total. These plug in to my Nakamichi AVP1 on which I use six of its seven channels. All channels have identical amplifier boards, each giving around 110w into 8 Ohms and the Croft controls the volume level, what ever the signal it outputs the crossover passes it through at more or less the same level to the Power Amp. The Tweeter and Mid driver on the Yamahas (or any other speaker) needs far less power than the bass driver and as there are no gain controls on the power amp I added a stereo volume pot to the Hi and Mid outputs inside the crossover case, to turn down the output levels for these. To give you some idea, the output from my CD player is about 2V, this passes through my Pre amp which is zero gain and the crossover which is also zero gain, so the bass channels on the power amp sees roughly the same 2V. When the volume pots in the crossover are set by ear to give the best balance between the drivers, the Mid and tweeter outputs sent to the power amp are around 0.75V for the same 2V sent from the CD player.

The Nakamichi AVP1 is a superb AMP and although designed as an AV amp handles audio in a very nuatral and matter of fact way. One of the best things I can say about it, is when folks have listened to my system and a couple of others I have heard, using these, no one comments about the amp performance, it just does its job without fuss and to a standard that does not draw attention to itself. Capable of tremendous dynamics and slam yet sweet and refined when needed. The only down side is a slight hum through the speakers that is more noticeable when using the unbalanced RCA inputs rather than the balanced XLR's. This is only small and not noticeable from the listening position.

As for the Nakamichi PA7 I only know these from reputation. A beast of an amp knocking out about 250w/channel of quality sound. A look at some of the reviews shows a well built product using quality parts and a sound that was highly regarded. I would think it would drive your Passive Yamaha's well.

Enjoy your speakers.

drSM

22-01-2016, 15:02

Ta Ken
That i could understand
The Nak AVP1 is a 7 channel AV amp huh.
Currently my refurbished AU111 drives them quite nicely
I swap in the 9500 sometimes and even the little 70 does a decent job
They are not to difficult to drive to modest levels
I ve wondered if a 200 wpc SS amp might exert better control.
Cheers
Shahrin

Qwin

22-01-2016, 16:12

I ve wondered if a 200 wpc SS amp might exert better control.

This was my thought when getting the Naka and it seems to work out in that respect. It's a debatable point as to weather the amps damping factor is as important. The Naka has quite high damping, but I have seen arguments that beyond a certain level it's irrelevant.
To me a good big'un usually sounds better than a good little'un but if your happy as you are, then don't change anything.

Mika K

22-01-2016, 17:51

Before you have tested the NS-1000M's with some really powerful power amp you do not know what it is capable of with that 12" bass driver. Thinking now something like monster Krells etc doubling their output to bigger loads from 8 -> 4 -> 2 -> 1ohms. The downside is that quite rare of these monster have the delicate nature that the Be-drivers again wish for, so it will always be compromise until bi-amping comes along.. :)

AlfaGTV

22-01-2016, 18:35

As Shahrin said; Good thread!

And i agree with Mika's thoughts, they do benefit from some major powerhouse up their rears! ;)
Mine are driven by a Karan KA S270, a dynamic bias Class A jobbie with one the highest damping factors ive heard of.
They sure sound fine that way, but also do a great job with a small EL34 pentode tube amp. A lot less bass though, which is a good thing as it is nowhere near as well controlled as with the big Karan.

YNWaN

22-01-2016, 19:10

Mika K

22-01-2016, 20:14

That should do the job.

If I remember correctly NS-1000M was one of the example speakers used at the study of amplifier measurements with the difficult speaker loads and there NS-1000M bass driver was stated to be very difficult because of the phase coherence. The measurements were done by Finnish professor Matti Otala who did some major research e.g. for Philips and Harman Kardon and is usually remembered because of the Transient Intermodulation Distortion..

drSM

22-01-2016, 23:27

I do not have the know how to go active.
Down the road i will prob get help to spruce up the passive X over and what Ken had done here in the earlier part of the thread is what i will prob do.
I ve heard of guys using the Quad II to drive these
I get lovely mids and highs with my AU111 - bass is there but is a wee bit untidy.
Overall the tone is v nice i feel.
I have read that one should not overpower these babies. they are rated 100 watts i believe.

drSM

01-02-2016, 00:38

Seems to me that if i were to do anything at all to my NS1000 s it would be to go active
I found this very impressive Aussie range of products http://deqx.com/
The opening video is q cool
I sent them a message and got a nice reply from Alan Langford its founder
The model he recommends is the Express II pre amp which has a DAC, lots of dig inputs, is an electronic X over with 3 sets of analogue outputs and is also a room correction device.
Alan Langford tells me that they have a number of DEQX users that have converted the NS-1000 to DEQX active and can also provide NS-1000 calibration filters to help
Building an active X over system from scratch as Ken has is impossible for me so this may well be the way forwards
All i need is about $3495 and 3 stereo power amps (or your Nak power amp) !
Any thoughts on DEQX ?

Ali Tait

01-02-2016, 05:35

Not familiar with that kit, however something similar can be had for a fraction of the cost from miniDSP.

drSM

01-02-2016, 09:38

Minidsp huh
Had a quick look
Do they do the active x over?

Qwin

01-02-2016, 12:19

Shahrin – I wouldn't say I had built my crossover from scratch, more like wiring pre built boards together and then doing some component upgrades, but I can fully understand if you don't feel comfortable working at this level. In which case the miniDSP is a very cost effective way of doing things, all be it in the digital domain. Their 4x10 HD unit is a digital signal processor (DSP) which is an active crossover. You connect it to a PC to set up the crossover points and levels and it runs stand alone once this is done. It even supports a pair of analogue inputs for a turntable etc. I considered this myself and the fact that Siegfried Linkwitz fully recommends it for his flagship active speaker is testimony to the quality of the sound it produces. They also produce less costly boards which are more DIY. From your point of view the beauty about the minidsp 4x10HD is it comes ready built in a nice case.

YNWaN – For me the weakest point in the stock passive crossover is the mid to tweeter crossover region, I find this on a lot of three way systems and feel it is the area where most improvements can be had from going active. That's not taking anything away from what you have done Mark, I heard your set up at Scalford and it sounded very nice, but for me the upper mids into the higher frequencies sounded exactly like mine in passive form (as you would expect). Maybe consider this for the future as I feel you can make things even better by going fully active.

As for the amount of power to drive them, it depends on the quality and control available as much as out and out Watts but I would agree that they like a bit of power on reserve.

The other thing to consider is that with a passive pair being driven by a 100w/channel stereo amp, the bass is probably grabbing about half of that (50w) and the remainder is being shared between the other two drivers, in very rough terms that is the maximum available. When going active with a 6 x 100W set up the bass gets all the 100w to itself so more like a 200W passive set up.

I am currently waiting on parts from Australia, my next build is having all of the opamps replaced with Burson discrete units (making 10 in total). The volume pots for driver matching are to be replaced by a simple voltage divider using fixed resistors, equal to the measured values from the pots and the boards with balanced input are to be used. I will be fitting JST or Molex multi pin connectors so the boards can easily be transferred into another project.

The build after that, which may be some way off, is to transfer the cross over boards into the case of a DIY 6 channel amp using Hypex UcD180HG HxR modules (120W/8ohm each). I will keep left and right speaker channels separated in the case, so two of the Hypex SMPS will be used, one for each set of amps/drivers. This will hopefully help avoid stereo crosstalk. Each power supply can handle up to 6 amp modules when used in an AV set up (multi drivers per amp module). So in my proposed set up, each supply is driving just three modules and those are each powering just a single driver. This should mean there is plenty in reserve for each bass module to draw on, should it be required.

I have worked out that the crossover boards can also be powered from the SMPS. The main output for the amp modules is 2x46Vdc but there is a 500mA auxiliary output on the supply that is half of this (2x23vDC). I have a couple of dual symmetrical and adjustable rectifier/regulator boards I got cheap from China, if I remove the rectifier stage and just use the regulator element, this will give me the 2x12Vdc required for the crossovers. These boards handle 1.5Amp max 1Amp continuous so the 100mA drawn by the crossover should not produce much heat and the small heat sinks fitted to the regs on the PCB should be enough.

All of this should fit into a stock 19” racking case which is 2U tall (roughly 88mm) and 300mm deep. I will fit a strip of 8mm thick Ally as a heat sink in the base of the enclosure and mount the hypex modules vertically. As you can tell I am well into the details of how all this might come together and when I get going will start a thread in “The Drawing Board”.
As a bit of a taster here is a rough layout of what I have in mind:

http://www.jkwynn.co.uk/Pics/6Channel_Hypex_Case.pdf

The dimensions are for the max internal space available in my chosen case.
The DC cables supplying the regulator boards will be tied up high on the back of the front panel and not interfere with the SMPS. The supply cables for each amp module are at the top of the vertical PCB's and the speaker outputs will run along the bottom of the case so there is a decent space between the cables. Other than that I think it is fairly self explanatory with the DC and signal cables being well separated and the AC supply kept well away from everything else.

drSM

02-02-2016, 01:53

The other thing to consider is that with a passive pair being driven by a 100w/channel stereo amp, the bass is probably grabbing about half of that (50w) and the remainder is being shared between the other two drivers, in very rough terms that is the maximum available. When going active with a 6 x 100W set up the bass gets all the 100w to itself so more like a 200W passive set up.e AC supply kept well away from everything else.

hmm the minidsp 4x10HD sounds very interesting.
Ken i am sure one can use different power amps.
From what i understnd from reading about Mark s set up he uses a very powerful Class D amp to drive the woofers and something else to do the highs and mids.
In your system its 100W x 6
Does/ can the miniDSP compensate for the different power ?

Qwin

02-02-2016, 08:14

MiniDSP has adjustment for levels in the software, but this is only for small amounts of adjustment. Any major differences in levels created through the use of different sized power amps or the requirements of different drivers (tweeters/mids need less) requires some form of attenuation. If the power amps do not have volume controls, which most don't, you need to add a control of some kind between it and the miniDSP on each channel that needs adjusting. As for using amps of different power ratings, that's up to you, but I prefer to use identical power amps, same size, same make. You have to consider sensitivity on the power amps as well, how much signal it needs to go loud, this varies between models. Also how linear the amps volume level is in relation to signal increase, if this slope is different between the amps chosen it is difficult if not impossible to balance the levels across the whole of the volume range as the amps increase at different rates. I'm not sure how much of a problem this is in reality, but it's something I took into account.

My Nakamichi is around 110w/channel and the Hypex DIY I am considering making would be 120w/channel, this seems enough to me, I get very deep and well controlled bass with the Naka, but I haven't tried anything bigger to compare.

drSM

02-02-2016, 13:36

Excellent advice there Ken regarding the power amplification.
I did note on another thread about the rare active Finnish NS 1000 That the amplifier feeding the tweeter produces only 3 or 5 W.
But there is just a specific Genelac design While this is something generic that we are adapting to the NS 1000.
I am convinced this is the way to go to take these wonderful spkrs up a notch .Gotta find some funds and the time !

whats the big deal about the Hypex - just 10 more wpc ?

Qwin

02-02-2016, 14:16

The hypex UcD (class D) modules are reported to sound superb and have been used by manufacturers in even two channel form costing several thousand pounds. Their nCore are even better but costly for a multi channel set up.
If I build an amp using these I can use the Hypex SMPS (Switch Mode Power Supply) and also get rid of the linear supply for my active x-over. I can drop it all into one small case weighing just a few Kilos. The Nakamichi is in a huge box, has a massive Toroidal Transformer and weighs 32 Kilos. With the crossover in the amp case it will simplify my external interconnects significantly.

Its as much about space saving and giving my aching back a rest. I love the sound the Naka produces and the only thing I can fault on it is the slight hum that can be heard close to. That on its own would not be enough to make me change, its a combination of all the benefits, oh, and the Hypex amp/smps are supposed to be dead silent.
Many studio guys have built Hypex based amps for their active monitors and their is much praise for them on Gearslutz.com etc, some saying they equal their Bryston power amps. :eek:

Ali Tait

02-02-2016, 16:34

I use a mini DSP unit to run a pair of Open Baffle speakers, which use a pair of Eminence Alpha 15" woofers per side, with a Visaton B200 driver. Alphas run up to 200hz, B200 from there up.

For the bass I use a Behringer Inuke 3000w amp, this has twin volume pots, so levels can easily be matched to the Monarchy Audio SM 70 power amp I use to power the B200's.

I find the Behringer excellent for bass duties, and it is not expensive.

Qwin

02-02-2016, 17:37

3000w:eek:

Do the street lights dip when you switch that big boy on :D

Ali Tait

02-02-2016, 17:49

:D Nah, it's a chip amp.

drSM

02-02-2016, 23:42

Hi guys u see from my gear that i use some old sansui integrateds
95 % of the time i have the au111 'flat' with tone defeat ON v occasionally on quiet listening i may turn on 'Prescence ' which bòosts bass a bit
Very occasionally i might use filters to compensate fir crappy vinyl cd etc
Not a purist
I like that flexibility

If i were to go active w the ns1000 using the minidsp mentioned can i 'fiddle ' ?
Could i set presets for a bit of bass boost etc
Thanks for indulging me exploring this which is entirely new to me

sq225917

02-02-2016, 23:43

I had a pair of Ncores here for a while. They sound great on the bass, but bring a somewhat steely edge to the mids and treble that sounds a touch synthetic. Not the best match with the Ns1000m

Ali Tait

03-02-2016, 00:08

Ali Tait

03-02-2016, 00:09

In any case, if you are running actively nd want a little more bass, just turn up the amp driving the woofers a little.

struth

03-02-2016, 00:15

drSM

03-02-2016, 00:48

Thats good news Thank you
This project to 'activate' the ole yammies could b fun

Ali Tait

03-02-2016, 07:17

how many drivers in the ob ali??

Two Eminence Alpha 15", a Crystal D500 midrange horn and an Isophon DKT tweeter per side Grant. I was also looking at that unit as it gives me the option to go balanced.

Qwin

03-02-2016, 09:33

I had a pair of Ncores here for a while. Not the best match with the Ns1000m

Interesting Simon, I would suspect from this that you are running your Yammies with the stock passive crossover.

One of the things I have found with my active set up is it has a warmer richer mid section and highs compared to the original.
If what you say is a trait of all the Hypex variants it might balance itself out, we will have to see, but thanks for the heads up.

Mika K

03-02-2016, 10:32

Based on my experiences the hypex UcD (class D) modules and many similar ones are good for bass only as with the Be-drivers for the mids and treble the modules bit clinical and harsh sound become too obvious. Personally couldn't handle that sonic signature at all.. :(

Qwin

03-02-2016, 11:11

Mika - What was your set up using UcD, was it 180/400/700 and which variant ST or HG, did you use the HxR regulators also?

Getting a bit concerned now.

Mika K

03-02-2016, 16:42

The modules were the UcD400, no more information available as the units were loaner from one local builder. Usually he goes for the best quality PSU that he can get but cannot confirm this at least for now. Have also tried Bel Canto smaller monos and the Apoll M-1.0 mono blocks with extra PSU having the B&O ICEpower modules and those have the same issues noted.

My opinion is that the be-drivers are at passive configuration very easily driven badly and leading to sharp, clinical and harsh sound. This has been true for all class D amps I have tried with these speakers. In fact, there are only few transistor amps that would use with my NS-1000M's for when it comes to listening music. The same applies with the cap updated XOs. About the active setup I have no experience :)

Qwin

03-02-2016, 18:16

Your experience with the passive crossover mirrors my own.
I have found a richer mid range through using an analogue active crossover, so the drivers are not holding things back, its the very basic passive crossover and I suspect the type of caps they used, which were better than Lytics but not up to modern film cap standards.

As for the UcD400 you mention, whether it is the ST version built for economy, or the HG version built for performance would make all the difference. If it was using the optional HxR regulators for the opamp buffer stage would also effect things. The type of power supply, Linear or SMPS and its suitability has to be factored in as well.
So there are quite a few things to consider before coming to any conclusions on the sound quality of the range per say.

There are literally dozens of articles praising these things, some from DIY builders, some from reviewers, some from Studio Engineers. Its hard to find anyone that has a bad word to say about them, so something isn't quite right. I read another 3 reviews today from studio guys that had built them and were now using them in preference to their Bryston 4BST amps.

I'm not doubting your opinion, just open to the possibility that there may have been a route cause for what you heard.

Wish I could get to hear one myself, to work out if I like what they do.

YNWaN

03-02-2016, 18:23

I also heard the nCores on Simon's Yammies and it wasn't so much that they were hard and bright, it's more that they were very dry sounding without much sense of atmosphere and the top end just lacked tonal variation - good bass though.

Qwin

03-02-2016, 18:36

Maybe they just don't suit the Yamaha's.

YNWaN

03-02-2016, 19:24

Well it certainly put me off from building a pair.

Mika K

03-02-2016, 22:09

All the UcDs, nCores, ICEpowers etc are just something I do not consider proper amps yet. I've tested them in several systems with open mind but haven't still heard music from them. My own systems, friends places, trade shows etc. The first version very horrible and the current ones are better but still I'm not convinced. Yes there is sheer power, and control, prime measurements and many things but..

The problem goes even deeper as the world is filled with excellent amps that just do not really suit NS-1000M's. And life is too short to have any of those in my system. This despite my main speakers are something else than these and NS-1000M's just visit at the spotlight every once in a while :)

sq225917

03-02-2016, 22:57

Qwin, my crossovers are rebuilt, all Teflon and polypropylene with the original coils. My RAAL tweeters use a modest first order crossover to pair with the yammy kids.

drSM

04-02-2016, 03:04

Ok you guys have lost me again...
I take it the discussion has moved towards the merits or lack thereof of Class D power amps, and some of you voice concern that sonically they my not measure up.
Or perhaps may not match the NS1000. ??

I am considering the miniDSP 4X10HD with three stereo power amps say three Exposure 3010S2 s , or if i can find a Nak AV 7 channel power amp here.
Would that be a reasonable plan ?

Qwin

04-02-2016, 09:56

The miniDSP 4x10HD will give you what you want in terms of an active crossover that is ready built and easy to adjust.
As for the Amps, as you can see from the previous few posts, its a very personnel thing, so what I like you may not.
They are both well respected products you mention, I think you mean Nak AVP1, which is what I have, the AP7 was a stereo amp.
The only thing I can say for certain, is that powering the original Yamaha drivers via an active analogue circuit, the AVP1 is punchy yet can dig deep, has a high level of detail and a nice rich midrange and crisp top end. I can not fault its output tonally, so that pairing works for me.

My gripes are merely its size, weight, the need for a separate x-over case and the small hum at the speakers.

drSM

05-02-2016, 10:22

Have my eye on one of these for my new Open Baffle project, this should have all the flexibility you want-

https://www.minidsp.com/products/minidsp-in-a-box/minidsp-4x10-hd

Thanks

The minidsp4x10hd has 2 or 1 pair of analogue inputs
The minidsp10x10hd has 8 or 4 pairs as far as i can make out
So to connect tts, tuner, cdp etc that would be preferable ?

Would 3 x Exposure 3012S2 amps be ok?

Ali Tait

05-02-2016, 10:29

Use a preamp. A simple passive would do. Connect your sources to it, output from the preamp goes to the miniDSP, then outputs from the miniDSP go to the amps you are using.

drSM

06-02-2016, 02:07

Use a preamp. A simple passive would do. Connect your sources to it, output from the preamp goes to the miniDSP, then outputs from the miniDSP go to the amps you are using.

Aha, Ta Don Corleone !

Ali Tait

06-02-2016, 13:30

You need something to act as a master volume control, otherwise you'd be adjusting three volume pots every time you change volume..

Qwin

06-02-2016, 15:49

Like Ali says, you need a master volume control on your pre amp even if it is just a simple passive.

The pots placed between the minidsp and amps are to balance the amount of power distributed to the drivers (call them trim pots), once adjusted to balance the levels the "Trim" pots wont need touching again, you will just use the master volume on the Pre amp. Once the "Trim" pot positions have been established you could even measure the resistance and replace them with a pair of fixed resistors, which would give you better channel balance (pots are never very accurate).

Note: The 4x10 has a built in volume control (big round knob on the left), so you may be able to run your sources straight into the dsp and control the master volume from there, it depends on how much the output from your various sources varies. To aid with this there is also a pair of dip switches inside the 4x10 unit which gives you a high/low option for input sensitivity of max 2.0v/0.9v. You can also set the output gain to 2.0v or 0.9v, using a combination of these settings gives you a lot of flexibility. - Check out the user manual on their site.

If you use identical amplifiers, you don't need attenuate the bass amp channels, let them run without any "Trim" pot after the minidsp.
The bass drivers are the most power hungry, the Mids and Tweeters use less so will require turning down using "Trim" pots till the drivers sound evenly balanced.

drSM

07-02-2016, 02:40

Thank you Ali and Ken
Ok pre amp up front
4 x10dsp allows some twiddling of input sensitivity and output gain

Some power amps have in built pots
The ones i will use probably wont.

2 of my Sansui integrateds can be used as power amps by removing links (pre out/power in)
For starters I could use the most powerful power amp to the woofers without any need for trim pots, right?
And amps feeding the mids and tweeters would need trimming pots to evenly balance the whole set up ( and may later be replaced by set resistors, as u mentioned Ken)
Is that correct?

Gents why do some people use such powerful (usually Class D) amps to drive woofers ?
I note Ali u are using a 3000 watter for yr open baffle 15 '' woofer and Mark uses some sorta big power for the woofers of his NS1000
Wouldnt a decent 100 wpc amp with headroom be fine ?
Furthermore the amps powering the mids and tweeters will need trimming and these are hardly breaking into much sweat at all
( i note that the Genelac Active NS1000 discussed in a thread here, the amp driving the tweeter on each side is pretty puny)
So trimming pots between minidsp and the amps driving the tweeters and mids are necessary for balancing, and does this not mean that the amps do not have to be identical ?

sq225917

07-02-2016, 09:18

Class d is cheap per watt, typically has a high damping factor and its perceived shortcomings tend to be at higher frequencies.

Macca

07-02-2016, 10:03

There is also the comfort of having far more power than you are ever likely to need. it's the hi-fi equivalent of nuking the site from orbit, just to be sure.

Ali Tait

07-02-2016, 10:08

What Simon said. I was previously using a Ruby Chameleon 200w power amp. The Behringer is cheaper, runs cooler and sounds better. Cheaper to run too.

But no reason not to use what you have already.

RichB

07-02-2016, 11:44

There is also the comfort of having far more power than you are ever likely to need. it's the hi-fi equivalent of nuking the site from orbit, just to be sure.

Hahaha, I like that.

RichB

07-02-2016, 11:49

Class d is cheap per watt, typically has a high damping factor and its perceived shortcomings tend to be at higher frequencies.

Not so much the higher frequencies I found, I loved my mini-t for the treble and mid but there's something about how it did bass just didn't sit right with me when compared to a SS amp with a nice big torroidal.

Qwin

07-02-2016, 12:17

Mega power is ok as long as you don't turn the volume too high, the Yammies are only rated (conservatively) at 100w as a three driver system. I once saw a DJ connect the bass channel to a tweeter on the PA system by mistake, it exploded across the room when the music kicked in.:doh:

Yes, use the most powerful amp for the bass drivers and you could use the power amp ellement from the Sansui's, they don't need be the same power wise, I just prefer identical amps, remember amp sensitivity is also going to dictate how loud it sounds. Different amps also sound well, different, I just think it is easier to dovetail drivers and crossover filter elements when the gain slope, tonality and sensitivity of all the amp channels are the same. There are however, lots of people who have used different size/make of amps that have been successful, so this is just my own personnel preference.

You are trying to create the right balance, a 3000w bass amp and a 20w mid would not work as the mid amp would be too low on output, so the bass would then need the trim pot to turn it down. As a very rough guide, the bass will use about twice the power of the Mid and the tweeter a little less, but it depends on driver impedance etc so this is only the case when all the drivers are the same impedance, like they are on the NS-1000.

sq225917

07-02-2016, 12:19

To be honest I was thinking of big dick multi hundred watters, not flea powered versions.

Qwin

07-02-2016, 12:31

Not so much the higher frequencies I found, I loved my mini-t for the treble and mid but there's something about how it did bass just didn't sit right with me when compared to a SS amp with a nice big torroidal.

The class "D" do bass very well, and a lot of folks share your view that class "T" do mids and highs well. I have seen more than one person try and combine the two, for the best of both. Trouble is, the mini-t hasn't enough umph for the mids, to allow a bass amp of sufficient power to be used and get the most out of the 12" drivers. IMHO.

I seriously considered 2 mini-t and a couple of hypex modules for the six channels. For the reason I gave and because of the different power requirements making it difficult to fit it all in one box, I backed away.

Firebottle

07-02-2016, 12:32

big dick

Not heard of that make before :scratch::lol:

King Dick - yes.

Ali Tait

07-02-2016, 13:06

sq225917

07-02-2016, 14:49

They sounded good with 200 watt of Krell KSA200 up em.

Qwin

07-02-2016, 14:56

drSM

08-02-2016, 00:10

Really Ali, I wouldn't have thought the 20w mids was enough to get to fairly loud levels.
Do you get to reasonable levels with this set up?

Mind you the NS-1000m are fairly sensitive at 90dB/1W/1M.

Damn it you just got me interested in the Mini-T amps again.

I did like what it did when Rich brought his to NEBO, like he stated, just seemed to lack the grunt for the bass as the wick was turned up. Other than that, it powered my Celestion 66's quite well.

Mmmmmm two mini-t plus power suplies and 2 Hypex modules plus SMTP and the crossover boards, can I fit that into one box? :scratch:

Ken are u thinking of using two Orb Mini T amps to power the mids and tweeters ie
http://www.amazon.com/Orb-Audio-Mini-Amplifier-Black/dp/B00GOK38OU
Class D, 50 wpc
3.5 mm jack inputs
Seriously?
and the bass handled by a more powerful class D amp ?

Ali, are u using the i NUKE NU 3000 Behringer amp ?
I believe there are the 1000, 3000 and 6000 watt models in the i NUKE range

Ali Tait

08-02-2016, 05:08

Yes Inuke 3000.

Ken, yes can get high levels, sensitivity is around 94db.

Qwin

08-02-2016, 08:31

Shahrin - I was actually thinking about the Amptastic mini-1 "T" amp but after re checking the spec on their web site, I see it is <20w RMS max @ 4 ohm.
Having heard it, I do like what this amp does with mid/high frequencies, just runs out of steam with the bass.
This may not be enough into 8ohms.

I am also considering the LM3886 (Gainclone) amps. The LM3886 is the warmest presentation of mids and highs in this series of chip amps (so folks say) I was particularly interested in this LM3886SMPS module by connexelectronic : http://connexelectronic.com/product_info.php/cPath/39_47/products_id/249?osCsid=c00773pfr9dcfkpf668rttleh4

Its a combined stereo amp and switch mode power supply on one 100mm square PCB.
It puts out around 40w/8ohm.

Like the "T" amps it lacks the ability to drive bass to good levels, so maybe a couple of these LM3886 modules paired with 2 Hypex for bass? The Hypex seems universally accepted as being excellent for bass duties. I have heard the terms sterile, neutral, bright, transparent, dry and dull applied to the mid/high end, but no one seems to really rave about the mids or top end, the focus is always on the bass. The reverse situation to the LM3886, so they might make a good pairing.

At the very keen price, I may just knock up a stereo amp with one of these LM3886SMPS modules and give it a listen.

Qwin

08-02-2016, 11:04

Interesting thread I came accross regarding an LM3886 build:

http://diyparadise.com/forum/index.php?PHPSESSID=0a3721095f09d8065d2e5beeb4165d 5c&topic=562.15

In particular from post #20:

"Just wanna share.. on May 30, 2008, a group of DIYers in my city Jakarta, sponsored by SAS (Sugeng Sarjadi Audio Syndicate), LangsungJadi Audio, Audio Video Mag (local Indo magazine) and other sponsors conducted a DIY Amplifier Blind Test Competition. Sort of like "free for all" class in sport game, the 33 (thirty three) participating amps range from Tube-based, Mosfet-based, Chipamp-based and Solid state, with participating DIYers from Jakarta city and its surrounding, Bandung, Surabaya, Medan and Samarinda. According to the participants, this event was first ever conducted within DIY community in Jakarta...

The jury was consists of several 'senior audio player' in the city, including one professional sound engineer, one musician ? also several of seasoned/veteran DIYers...

Thank God, surprisingly my Gainclone LM3886 won the second prize!! ?"

Note - First and third prizes went to valve amps.

These things must be pretty good to do so well in a blind test of 33 DIY amps. :)

Qwin

09-02-2016, 20:07

Here is something I had not considered before:

If I used 40w amps for mids and tweeters and say 300w for bass, the mid range would not need any attenuation.
The bass channels would have way to much power and need turning down, as would the tweeters.
This would leave the all important mids to come through unmolested.

There could be some positive benefits in doing things that way, this does have me re-thinking things a little.:rolleyes:

YNWaN

10-02-2016, 07:15

It's the gain structure of the amps that matters with regard to matching, not their outright wattage.

Qwin

10-02-2016, 07:53

True, but the small difference in sensitivity in this case is nothing compared to the fairly massive difference in power output. (40w v 300w).
As for structure, all drivers see the same gain from the pre amp, the filter elements are all zero gain so add nothing, so we are only talking about balancing the difference between the power amps for the drivers. I am fairly confident as to which amps would require attenuation in this instance. (He says with everything crossed) :)

walpurgis

10-02-2016, 10:59

40w v 300w is only about 8db difference in ultimate loudness.

Qwin

10-02-2016, 18:16

40w v 300w is only about 8db difference in ultimate loudness.

True, and in overall loudness it is not a great amount, but measured as the difference between mid/bass drivers output, when trying to match them, is huge.
1dB either side of the right amount will go from bass shy to bass heavy. Some say they can hear half a dB difference when matching DSP channels.

I am using identical amplifier channels at the moment, same watts/sensitivity/gain and drivers are all the same nominal impedance. The midrange drivers are attenuated and only receiving about 45% of available pre amp output compared to the Bass at 100%. At these levels they blend nicely.

All things being fairly equal, driver impedance and sensitivity/gain factor, then the bass driver would, from my experience with the Yamaha's, require about twice the wattage of the mid range in very broad terms, plus any headroom you want to build in. With this scenario the 300w/8R bass amp is going to be plenty big enough to work with the 40w/8R mid and tweeter amps I have in mind. The bass amp actually goes above 300w/8R but it has THD of 0.1% at this level, rising to 1% at 375w/8R.

If time permits, I might just build one of these 40w Gainclones to try it out, at replacing the mid channels on my 6x 110w Nakamichi and maybe bring it as a lash up to the next NEBO meet.

YNWaN

10-02-2016, 19:08

The power amp gain should be adjusted so they are all the same, then it doesn't matter if one is 5W and another 5KW. The power amp that drives my woofers has adjustable input gain. You put a test tone through the system and measure the output at the speaker terminals for the lowest gain amp and adjust the other(s) to match.

Juha

10-02-2016, 19:52

My Genelec active NS1000 are adjusted like this.

There are 3 trim pots (bass, middle, treble) on the active crossover.
These are adjusted at the factory probably for anechoic flat response.

Then 3 rotary switches for room adjustment. These are 5 1 db steps cut only.
Flat room response is done like this.

Additionally, there are 4 different bass cut slopes available to cure boomy bass.

The amps are 50w nominal for bass and middle 8w for treble.

Qwin

10-02-2016, 19:59

The power amp gain should be adjusted so they are all the same, .

How does this balance the outputs for bass mid and tweeter drivers that have vastly different requirements?

Am I missing something here?
As I understand things, Gain is just the multiplying factor of the amp, if all channels are multiplied by the same figure for a given input voltage the high and midrange content will be way too much and you wont reach a very high level of bass before the tweeters blow, all other things being equal. Attenuating the mid range and tweeters is necessary with amps of roughly equal sensitivity and gain to balance the output for each drivers needs. If amps of different sensitivity and or gain are used the criterion changes as to where the attenuation is needed. Sensitivity being the min voltage required for the amp to achieve its maximum gain.

Give me a clue if I'm wrong.

drSM

11-02-2016, 00:09

Following the discussion with interest.
Ken i looked up the Gainclone LM3886 and it seems to be a fav among diyers.
50 odd wpc i believe
Might even be good enough for the woofers
Is it a 'chip amp' ? (dont really understand what that means except that its something small!)

Qwin

11-02-2016, 08:22

Yes, LM3886 is a chip amp, one of a group of four chips who's circuits are loosely referred to as Gainclones, 68w/4R, 38w/8R for this one, but output is voltage dependant so these figures vary as does distortion levels. The harder you push it the more heat is generated and the bigger the heat sink you require.
Rod Elliot, who's opinion I have a lot of respect for, built a circuit to try these chips and described the sound as "very good, audiophile sound with a caveat" that being you shouldn't push the chip too close to its limits, where the "SPiKe" clipping protection kicks in, it gets noisy around this point.

Link to spec comparison of the four chips: http://home.pacific.net.au/~gnb/audio/lmcomp.html

The Connexeltronic link I gave a few posts back is for a stereo amp module using two of these LM3886 chips and a switch mode power supply all built on to a 100mm square PCB. This is what I am going to try out. You simply mount it on a heat sink, connect an AC supply and give it an input from your pre amp, some speaker leads to the output and your in business. the voltage from the smps gives an output of 40w/8R. These modules have DC detection/shut down for speaker protection, soft start to avoid switch on thumps, and over/under voltage protection.

It can be used for woofers, though it probably isn't going to get the best out of the 12" Yammies, which by reputation, are hard to drive. Rod said they were ideal for Mid/tweeter use in an active set up.

KMTech, who make my crossover boards (same circuit as Rods), are developing a board with 3 way crossovers, trimmer pots for driver matching and 6 channels of amplification using these chips as we speak. These are designed to go inside the speaker cabinet, or can be used external. They have a few substandard boards, suitable for DIYers for sale in their ebay shop.

Mika K

11-02-2016, 16:36

Old salt makes you thirsty, so inspired by this topic just had to try how these stock NS-1000M's without any other tweaks than modern binding posts would sound nowadays after living few years with the JBLs. Well, not bad at all being pretty transparent and accurate despite being little harsh and lacking some of sheer power I'm used to.. :)

https://c2.staticflickr.com/2/1530/24933785046_19788c0f64_c.jpg

sq225917

11-02-2016, 16:43

A little harsh, funny how things are perceived,

Mika K

11-02-2016, 16:55

Indeed. I think it is a subjective matter and more about what the adjective means to different people and compared to what.

This is my fourth pair of NS-1000M's during my journey and all together have spent years and years with them using different amps etc. Some also had updated xo, so I think I'm pretty familiar with the sound they are able to produce in passive configuration :)

Qwin

11-02-2016, 16:59

That's a fair description of how mine sounded in stock form, the recap helped clean up the mids/high frequencies and going active took it further.

Brass instruments were a bit harsh and slightly piercing, plenty of detail, and very transparent, just a bit too forward and grainy.

Being very picky with these comments, still sounded pretty good.

Mika K

11-02-2016, 17:03

Yes the compared to the cost of these speakers the sound quality you get is still astonishing. Despite of the development of the prices during the past years..

drSM

13-02-2016, 10:33

KMTech, who make my crossover boards (same circuit as Rods), are developing a board with 3 way crossovers, trimmer pots for driver matching and 6 channels of amplification using these chips as we speak. These are designed to go inside the speaker cabinet, or can be used external. They have a few substandard boards, suitable for DIYers for sale in their ebay shop.[/QUOTE]

Thats interesting.
Altho the Yammy woofers would need more grunt, yes?

Qwin

13-02-2016, 11:18

Yes, a bit more needed, that's my feeling and others said the same earlier.

I have been in conversation with Connexelectronic about using one of their class D amp/smps for bass duties and they recommended the IRS400SMPS.
With an input of 1.5v this would give 200w into 8ohm, it has gain of 26.67. Its a single channel unit so two will be required. I would team these up with a pair of their LM3886SMPS two channel units and that would give me my six channels. Either of these SMPS auxiliary outputs is suitable, via a regulator board, for powering the crossover boards as well. A plan is forming. :)

drSM

13-02-2016, 15:27

Excellent
Fingers crossed. Looking forwards to good news :)

YNWaN

13-02-2016, 23:09

How does this balance the outputs for bass mid and tweeter drivers that have vastly different requirements?

Am I missing something here?
As I understand things, Gain is just the multiplying factor of the amp, if all channels are multiplied by the same figure for a given input voltage the high and midrange content will be way too much and you wont reach a very high level of bass before the tweeters blow, all other things being equal. Attenuating the mid range and tweeters is necessary with amps of roughly equal sensitivity and gain to balance the output for each drivers needs. If amps of different sensitivity and or gain are used the criterion changes as to where the attenuation is needed. Sensitivity being the min voltage required for the amp to achieve its maximum gain.

Give me a clue if I'm wrong.

I'm sorry but I'm not sure what you are missing here, it may well be a communication/Internet issue. The active crossover sets the relative balance, the power amp input gain sets the relative level amplified - this has nothing to do with the maximum output available. This is well known system architecture and beyond this I don't understand what is being confused (sorry). The maximum output of the amps is not relevant though (or at least it shouldn't be in terms of frequency response).

Qwin

15-02-2016, 08:33

Mark - I think we may have been talking at cross purposes.
My analogue active crossovers have no variables, they are just fixed frequency crossover filters without any controls. Even with digital stuff like miniDSP you control driver matching by attenuating the voltage to the power amps, only very fine tuning should be done through their software. My Power amps have not got gain controls, as is the case with most power amps. It is best to try and match amplifier output to driver requirements so that the minimum attenuation is needed anyway. I will have to create whatever attenuation is required and I would rather attenuate the Bass than mids, hence the individual power amp sensitivity and gain behind each driver is relevant to me in achieving this. If all power amps are identical, as with my current set up, you end up attenuating mids and tweeter channels because the bass requires more, in the case of the Yamahas the midrange amps need less than half (45%) the input voltage of the bass amps, which have identical sensitivity/gain, to give the required output for a balanced sound.

This is the way things are set up in my case.

Note that analogue balance controls on an active crossover (where fitted) are simply volume pots on the output, which do exactly the same as the volume pot on the input of a power amp (where fitted). They are both just reducing the signal voltage fed to the power amp stage.

Also, and you may not have written what you meant, amplifiers do not have variable gain. Gain is the fixed multiplying factor of the amp x10 x15 etc, it is the input voltage which is altered to achieve the desired output. eg 1volt in x10 = 10 out. 2volt in x10 = 20 out. (leaving input sensitivity aside). This is more for others reading than yourself.

Just to clarify, I am using line level crossover filters with zero gain.

I hope that clears up where I am coming from in my approach :)

drSM

16-02-2016, 01:34

Another approach to tweaking the NS1000 would be Bi Amping
One could split the passive X over (during recap)
Pre amp with 2 sets of outputs (or Y splitters) feeding 2 stereo power amps (perhaps with pots between pre amp and power amps).
Can also do vertical bi amping.

Would perhaps retain the basic character of the speaker.

Qwin

16-02-2016, 08:08

Yes you can bi amp or even tri amp and use some or all the elements of the passive crossover by separating them on the PCB and using separate amps for each.

For me the biggest improvements are to be had by going fully active, a tighter well controlled and powerful bass, but more importantly a smoother and richer midrange and top end.

The one criticism people seem to levy on the NS-1000M is it can be a bit bright and forward. This is subtly changed for the better when going fully active.

sq225917

16-02-2016, 19:11

It doesn't matter if an amp has 10 or 100 watt ouput, if they both have the same voltage gain they will drive a speaker to equal volume levels unless the lower powered one reaches clipping before equal volume is reached.

An active crossover can either just be a simple frequency filter, or they can include an element of variable overall level attenuation as well. If the prior then you need to adjust amp gain to compensate for driver sensitivity, if they include attenuation then you can do it in the crossover and forget about setting amp gain per amp\ driver.

Decent active crossovers usually have attenuation built in, in the form of a level control. Thinking of amp power as a means to control level is a mistake as the only way an amps power can affect level is by clipping and you don't want to drive any amp into clipping as that'll burn your voice coils...

Oh and that brass, its meant to sound shrill and piercing, that's how trumpets and co actually sound

Arkless Electronics

16-02-2016, 19:49

Qwin

16-02-2016, 20:32

The discussion is going round in circles now, I don't think I have a problem with what I'm doing, never did, so I'll just get on with it and judge the results. ;)

Today I ordered one of the LM3886 amp modules with on board SMPS, this is to experiment with and check out the sound quality before getting into a build.
I constructed a lash up ready to accept it yesterday, this being a heat sink mounted on a piece of MDF and a piece of Ally sheet to mount the input/output sockets on.

I'll try it as a stand alone stereo power amp running a pair of Tannoy M20 Murcury and also try dropping it in to my active system for midrange duties, replacing two of the channels on my Nakamichi AVP1.

YNWaN

16-02-2016, 21:47

I only commented because I couldn't make sense of your earlier comment that a lower powered mid/treble amp would unbalance the system (I'm paraphrasing) - I don't understand because unless it's clipping it shouldn't matter at all.

Qwin

16-02-2016, 22:48

I will be using lower powered Mid and treble amps which tend to have less gain.
My goal is to use mid range amps without any attenuation (other than at the preamp volume control) and have additional attenuation on the bass amps to match the mids.
This is the reverse of my current set up where the mid range amps inputs are attenuated to match the bass amps which are not.
It's that simple.

What makes pairing these things up difficult, is that companies are not very good at specifying gain or sensitivity for kits/part built modules for DIY builds, sometimes you get part of the information in the spec, other times you don't get much other than watts into 8/4 ohms. So it ends up as suck it and see.

Qwin

18-02-2016, 17:38

The LM3886SMPS module arrived this morning from Bordeaux, pretty quick as I only ordered it Tuesday afternoon.

The supplier forgot to put the Input/Output/Aux Power leads in the box. :steam:

They posted them on today, so just have to wait for them to arrive to complete the lash up.

I drilled and tapped the heat sink this afternoon so I can bolt the chips down, run out of things to do now.

Qwin

20-02-2016, 12:16

Still waiting on the leads to connect the input/speaker sockets, otherwise the lash up is completed as you can see. Powered it up, the light comes on after several seconds, due to the built in soft start. Couldn't find any heat sink insulation kits big enough for these chips so ended up using a couple of layers of Kapton tape. I already had the insulating collars for the screw heads from a regulator kit. Checked there was no electrical connection between the chip and heat sink before turning on though. These bare chips are supposed to have better thermal transfer than the encapsulated ones, but it was what they supplied, so Hobson's choice rather than any grand design.

http://www.jkwynn.co.uk/Pics/LM3886_Lashup.JPG

Just to clarify, I'm looking at making a combined 6 channel Amp and active crossover for the Yamaha speakers in one 19” 2U case.
This module (x2) would be used for Mid and high frequency channels.
Rather than launch into the whole project (expense) I'm first trying just one module in this lash up, to see how it performs.

Question folks: Do you think I should carry on, with this as part of this speaker thread?
Or start a new DIY amplifier thread in “The Drawing Board”?

I suppose it is all related, as it's specifically for the NS-1000M and won't work with anything else due to the built in crossover.

I hate all this waiting around for parts to arrive, if they had shipped the leads with the module as they were supposed to have done, I would be listening to it right now. :(

Macca

20-02-2016, 13:25

I'd keep it all here Ken, save anyone who wants to reference it later having to jump across threads. But if you want to take it to DIY it is ultimately up to you.

Following with interest BTW.

Qwin

20-02-2016, 14:01

Most of my projects have a fair amount of DIY in them, I'm just conscious of the fact that this is in "Past Masters" and has deviated somewhat from my original Yamaha renovation thread.

I think your point about following the development and possibly having to jump threads is relevant Martin.

If folks have no objections, I will keep it all together.

Hoping to bring this lash up to NEBO7.

AlfaGTV

20-02-2016, 19:55

I agree with Martin, keep this thread intact! It an interesting read and provides lots of tips for modder of Yammies and other speakers alike.

Much appreciated!

drSM

20-02-2016, 22:35

Please keep it going Ken

Qwin

21-02-2016, 10:23

:thumbsup:

sq225917

21-02-2016, 10:41

Looking good ken, interested to see how fully active goes.

YNWaN

21-02-2016, 12:40

Ken, did you see/hear those fully active NS1000-M's that were at Scalford last year?

Qwin

21-02-2016, 15:15

Looking good ken, interested to see how fully active goes.

My present set up with the Nakamichi is fully active.
I love what it does, no character of its own colouring the sound, it just plays what ever is on the track. Huge detail and deep, controlled bass, so why the change......

Summary of how I got to this point:
What got me on my present quest is that when using the RCA inputs on the Naka it has a slight hum at the speakers, this is almost non existent when using balanced inputs however. It is a huge and heavy box, with an additional box for my active crossover, I struggle with the room for it all.

I thought I would make use of the size and weight advantages of Class D and SMPS and build it all into one reasonably proportioned case.

First stop was the highly regarded Hypex modules, the NCore were too expensive for me to consider as a six channel build, so I considered the UcD180 modules and Hypex's matching SMPS's. I spent some time on the layout and came up with a very tidy scheme, with AC, DC and signal all nicely separated with no close or crossing wire paths. I was all ready to go with this and even had a case picked out.

Several people that have heard Hypex modules and other class D raised the issue about the sound and whether I would like it. No question, everyone seems impressed with the bass they produce and they are very detailed, transparent and quite. The Pro guys seem to love these aspects for studio use, many using them in preference to their Bryston 4BST (check out Gearslutz threads). But for music reproduction in the home and for pleasure? Hifi users seem more reserved in their praise, other than for the bass and clarity. One or two have been brave enough to say they feel the all important mids are a little sterile, and somehow lack that illusive ingredient that lets you get emotionally involved with the music. I did a lot of reading and couldn't find much written about the mid range at all, generally people were not raving about it like they were for the bass, almost avoiding the subject, this started alarm bells ringing.

The "Gainclone" class AB chip amps have a keen following and are strong on mids and highs, are good for bass, but perhaps run out of steam a little when pushed. Rod Elliot is someone who's opinion I value, he has written many articles and has published many audio projects, he described them as "audiophile sound with a caveat" that being you don't push them so hard as to reach the point were the protection circuitry kicks in, they get noisy around this point. The legendary Siegfried Linkwitz used LM3886 to power his active Pluto and LXmini speaker designs, so they are regarded well by people that would know.

A plan started to form, where I might combine LM3886 class AB for mids and High frequency, with class D for bass.
I started to look at what was available in modules for DIY builds and came across Connexelectronic, they not only produce amplifier boards for both these classes, they also produce amplifier boards with on board SMPS. I am currently accessing these for suitability. Both the LM3886SMPS and the IRS400SMPS modules have an auxiliary voltage suitable for powering my active crossover boards, so their existing linear PSU will not be needed, saving a bit more space in the enclosure.

This process is what got me to my current way of thinking and the path I am now taking, but nothing is cast in stone.

Qwin

21-02-2016, 15:20

Ooops - duplicate

Qwin

21-02-2016, 15:25

Ken, did you see/hear those fully active NS1000-M's that were at Scalford last year?

Yes, I think so, but cranked up pretty loud if I remember so you couldn't really tell how good they were.

One set was definitely "Beobloke"

There was also a recapped passive set a few rooms up the corridor.

YNWaN

21-02-2016, 16:51

Qwin

21-02-2016, 18:21

They weren't turned up very loud either of the times I went in to the room.

I don't think Beobloke showed a pair either - could be wrong though.

I've listened to a pair of N-Cores at a friends house in a system I hear often. We also had the opportunity to compare them to his regular A/B amp. They immediately impress with a sense of precision, control and detail. Bass is well controlled too but there is a problem. The higher you go up the frequency range the less good they are with tonal character until at the top end they are quite white noise like, tish rather than sticks striking metal. This gives them rather a dry 'airless' quality. I use Class D myself to drive my woofers but I wouldn't use it to drive the mids or tweeters (that was my intention when I auditioned the N-Cores).

I could have sworn it said Beobloke on his name tag , but it may be my memory playing tricks.

Your description of the Hypex is similar to others I have read and is what started me putting the brakes on my 6 channel class D build.

I am hoping my pairing of "Gainclone" class AB for mids/highs with class D for bass will give a more engaging mid range and a smoother top end.
As I haven't heard a "Gainclone" the current lash up is to satisfy my curiosity as to whether they will suite my needs.

There are going to be a few Hypex based amps at Scalford this year so I will be checking those out to get a first hand impression.

Qwin

22-02-2016, 13:53

Mark, were these the Scalford Yamaha's you meant?

https://farm8.staticflickr.com/7632/16783520848_6f595d36c2_z.jpg

YNWaN

22-02-2016, 16:09

Yeah, they were the active ones. The box in the stand is actually the passive crossover. The active crossover was separate. A stereo Lab.gruppen amp powered them passively and three more powered them active.

Qwin

22-02-2016, 16:26

Yes, I remember those, I only heard them fully active (analogue filters), they sounded very nice and added weight to my leaning toward analogue myself.

They were in room 208, it was the ones in 210 that were a bit loud for the room size.

Qwin

23-02-2016, 18:20

The leads arrived this morning for the lash up, better late than never.:carrot:

Running it in at the moment using the Croft Micro25 Pre amp and listening through a pair of (passive) Tannoy M20 Mercury. In this guise it's working as a stereo power amp. Not bad from the off, tiny amount of hum through the speakers, a lot less than the Nakamichi. Compared to the Technics SU-V303 these speakers are usually paired with, the lash up wins hands down, even from first fire up. It should improve over the next day or two as things settle, so quite hopeful about how it will develop.

drSM

23-02-2016, 23:39

Hi Ken, so is that a single stereo Gainclone that you are testing ?

Qwin

24-02-2016, 08:06

Yes, a single stereo board (x2 LM3886) with switch mode power supply built in.

If I decide to use it for active, it will power the mid and tweeter on one speaker with a single IRS400SMPS (Mono) for the bass on that speaker.
All doubled up for the second speaker.

After I have run it for a while as a stereo amp, I will drop it into my current active set up, replacing the Nakamichi channels for both mid range drivers on the Yamaha's. This will give a better impression of how it will perform for my intended set up.

I must say I am quite impressed with it as a stand alone stereo power amp so far.
It needs a fair bit of volume on my Croft Micro25 Pre to get it to a reasonable level, but my Croft is not Standard and has zero gain on the line stage, so it is only seeing the output from my CD/DAC. My intended Pre for the final build has higher gain plus I will be using Balanced XLR from this, so it should work out fine if I go this route.

Qwin

25-02-2016, 14:06

The "Gainclone" module has about 20 hours on it and is sounding rather good, no harshness in the upper region and nice midrange tone for voices etc.

OK, I'm convinced enough to give it a whirl in the active Yamaha set up.
I unplugged the two midrange filter channels feeding the Nakamichi power amp and connected them to the Gainclone instead.
I had to re-wire the crossover balancing pots, moving one from the mids to the bass.(only two used)
The mid range is now playing through the active crossover to the Gainclone without any reduction.
The bass coming from the crossover is now turned down very slightly before going to the Naka.
The tweeter set up is the same as I had it previously through the Naka but required turning down a bit more to match the Gainclone mids.

So the set up, level wise, is working pretty much as expected.

The levels are fairly close to optimum, only playing more music will confirm this, but it's sounding very good.

I will continue to put more hours on the Gainclone to check for any further improvements but I think it's pretty much there.

With the wiring configured for the active set up, the hum on the Gainclone channels has increased to the level of the Naka. I can see that if I get as far as the full build, that grounding, signal paths and cable shielding will need careful attention. I noticed that the Naka being used for just tweeter and bass is now almost silent. There maybe some noise being generated from the simple linear power supply feeding the crossover boards and effecting the midrange, I will look into all of this as things develop.

sq225917

25-02-2016, 15:57

If you serious about quality and chip amps, Tom's modulus range are pretty good, better specs than all other. A modulus for tweeter and mid and a parallel for the bass would be hard to beat.

http://www.neurochrome.com/modulus-86/

Arkless Electronics

25-02-2016, 16:39

I have some point to point wired "gainclone" using LM3886 which I built years ago and have worked great ever since (I brought it to a NEBO some time ago). They are a very good solution to cheap high quality power.....

roob

25-02-2016, 18:07

Ken, did you see/hear those fully active NS1000-M's that were at Scalford last year?

They were the ones being demo'ed by Colin from Chevron Audio and belong to Gary Hargreaves.
They defiantly were not playing loud, there was a passive pair in another room which sounded nowhere near as good as the active ones.
The Lab Gruppen amps have now been replaced by Hypex ones.
Colin will be at Scalford again this year with some active NS10M's.

YNWaN

25-02-2016, 18:12

Yes, I assumed there was a strong link with Chevron Audio.

Qwin

25-02-2016, 18:37

Modulus - These boards are single channel without power supplies (90x70mm), so you need lots of them and suitable power supply/s for all of these plus my pair of crossover boards, which use a different voltage, it sounds as though it would end up a bit bulky.
I'm not sure you can parallel two of these boards.

It's getting too far away from my one box active filter/Amplifier objective for this project.

The board I have on test is 100x100mm 2 channel with built in SMPS.

Qwin

25-02-2016, 18:41

Jez - I remember that mass of bits you assembled at one of the early NEBO meets, it sounded a lot better than it looked.:D
Another reason Gainclones were in my sights, though I wasn't sure which chip you had used.

EDIT: I realise implementation is everything, but this little LM3886 with on board SMPS is sounding pretty damn good, listening to Malia/Boris Blank at present.

Qwin

27-02-2016, 15:24

Well, the Gainclone has continued to improve and now has 35 hours on it, I had to increase the bass and tweeter levels as the Gainclone seemed to get louder as it ran in, if that is possible. I have no complaints with the sound quality when used for the mid range amplification, in the active set up it is as good, if not better, than the Nakamichi. There are only very subtle differences in presentation between the two, that comes down to personal taste, but I could live with either. I like the vocal reproduction with the Gainclone a lot and brass instruments on some Mexican music I sampled was very clean, bright without being harsh or spity. Just listened to Hugh Masekela blowing his horn and it was sweet.

Overall very pleased with how it is performing and I'm confident enough to use these modules in my build, as amplification for mids and tweeters.

Received a pack of resistors and Caps this morning for the next crossover build, this is going to have balanced XLR input and all opamps will be discrete devices, I have the PCB's and I'm trying to source Header/Housing/terminals to suit the pin centres for the input/outputs. Although the crossover is currently free standing, the wiring on this build is going to be socketed so the boards can easily be transferred into the final combined filter/amp project. It's a nightmare trying to marry up the connectors from the basic spec that Farnell or RS gives, so many inaccurate details, so you have to check the manufacturers data sheets. Just for the female crimp terminals there is a choice of Brass/Copper/Bronze substrate and plated in either Tin/Nickel/Gold or a combination of several of these. I'm a bit bog eyed with it all but managed to get a full list of part numbers required.

Why the XLR balanced input, you might ask?

My intended Pre Amp for the final set up is the Pro-Ject Pre box RS, which has ECC88 (6922) tube output and has balanced output connections only. It has an RCA unbalanced output, but this is for taping and has no volume control on it, so no good for controlling my active set up. The most effective and straight forward solution, is to build the crossover board variant designed for balanced input and use stock balanced connection leads (Klotz mc5000).

There are a lot of stages to go through before this one is put to bed, but I'm getting there slowly, even when accommodating a few changed ideas, due to my pre conceptions, along the way. :)

Arkless Electronics

27-02-2016, 16:38

Shockingly good for a "car radio chip" are those LM3886 ;) They are a prime example of how sometimes in this game one can get lets say 85% of the performance that it's even possible to get for little money.... but then to get another 10% you need to add at least a nought to the cost :eek:

Qwin

28-02-2016, 10:29

LM3886 - Yes I believe it has been used quite often for car radios, flat screen TV's, etc.

It certainly does extremely well with price v performance and in a very compact package.

drSM

30-03-2016, 03:37

Wassup Ken, any progress?

Qwin

30-03-2016, 13:34

Hi Shahrin

Progress is slow, I'm either strapped for cash or parts. I just bought a Pro-Ject Pre Box RS, this Pre Amp has the balanced outputs I was after to feed the 3rd generation x-over boards. I'm now waiting on the Burson discrete opamps which have not been available since I tried to order them first week in January. They have told me they are working on the board layout, so either there is a problem with the design or they are simply repositioning things to make production easier. Either way I am prepared to wait for them to be happy with everything. I could buy some more of the first issue V5 modules from their distributor in France who still has stock, but I'll wait.
Meanwhile..............

http://www.jkwynn.co.uk/Project_Images/Yam/yam_33.JPG

I've populated the PCB's for my 3rd generation build. These are designed to accept balanced input and the input/outputs are socketed to allow easy transfer between projects. It took a bit of head scratching, but I managed to find some stock sockets that could be modified to fit the hole pattern on the PCB's. I had to use Takman 1% metal film resistors as the PRP I liked are not being restocked at HiFi Collective. The terminal blocks are to allow a quick change of resistors when I come to balance the output levels for the amplifiers/drivers requirements. I will start with pots then measure the values and use fixed resistors, this way I can do some final tweaking of the resistor values, if required, without having to unsolder anything.
I will initially use these boards in the existing separate x-over enclosure which feeds the multi channel Nakamichi Amp. Once I'm happy with them using the balanced inputs and a full complement of discrete opamps, I'll transfer them into my combined one box Amplifier/x-over build, hence the sockets, for easy change over. I'm taking my time with the changes and only progressing after each step has been optimised and proven to work.

drSM

06-04-2016, 08:12

Slow and steady Ken
Look forward to yr final active set up.
Me, i am gonna try clean up the L pads on my NS - they sound pretty scratchy
Best
S

Qwin

06-04-2016, 09:41

Thanks Shahrin - At last, the final batch of discrete opamps are on route from Burson (SS Audio) in Australia.
This will allow me to populate all 10 opamp sockets with discrete modules on this latest balanced input build.
I have acquired a Pre amp with balanced outputs and have made up some leads, I also have the XLR sockets ready to fit into the back of the case.
Things are starting to come together.

My first test is to fit the existing (mkII) board with a full set of Burson modules, this is just to confirm that this is the best way forward. I have experienced with ordinary opamps, a situation where one or two of the same type sounded nice, but fit 3 or more and it didn't sound as good. I had to mix and match using various types and make the most of their individual talents, some are better at bass, or very transparent for mids etc. I found this to work better than fitting one type throughout, no matter what the spec or price of the opamp. Every filter section and the buffers improved when they had a discrete module fitted, but this was in isolation. Things might be different when fitted throughout, if I have to mix some opamps in their, then that's what I will do.

I've done as much prep. as I can, its back to the waiting game now and listening out for the Postman. :)

Qwin

22-04-2016, 18:28

http://www.jkwynn.co.uk/Project_Images/Yam/yam_34.JPG

Parts arrived from Burson, so no op amps at all on the boards now, all 10 are discrete modules. Based on the previous sets, they will take 50 hrs to run in, but even after 5 hours they are sounding much better than the op amps.

I will definitely be using this configuration on the balanced input build.

drSM

18-06-2016, 03:38

Thats a cool Yamaha box
Whats it sound like?

Qwin

18-06-2016, 07:02

Not sure what you mean Shahrin?

The case for the crossover is one I made/adapted, I put the Yamaha Logo on it if that's what you mean.

I've been doing extensive listening tests and have my final views on which op amp combination works best (for me) in this Yamaha application.
I will post my findings soon.

337alant

19-06-2016, 08:41

Looking good Ken :eyebrows:
I don't see any voltage regulation though supplying the board's just a rectified and filtered supply, do they have onboard regs:scratch:

Alan

Qwin

19-06-2016, 09:40

Hi Alan
No, no regulator, the supply was made and recommended by the maker of the x-over boards (KMTech) and I have also seen it written elsewhere that unregulated works fine with this circuit, which is basically the same, electrically, as Rod Elliott's (ESP) boards. The supply is off a pre amp design of KMTech's and is way over specified for the x-overs, as the pair of boards only draws < 150mA. It seems to work fine, but I have thought about building an alternative.

I'm keeping busy and currently making very early concept designs for an active "FAST" speaker, using a wide band 5" and a 12" sub/Woofer, in a sealed cabinet. I will be using the same crossover circuit in 2way form and using a regulated supply on that one. I've got one of the 5" drivers mounted in the cardboard box they were shipped in and it sounds pretty good, even in this crude set up. I'm well enough convinced to knock up some rough 7L sealed cabinets to give them a proper evaluation. I will start a thread for this if it becomes a viable project.

Bye the way, I caught a couple of your YouTube clips on the R2R's, very informative and well made :thumbsup:

Cheers

337alant

21-06-2016, 08:52

Hi Ken,
Ah I see strange cant see why it wouldn't benefit from even a 317 / 337 reg supplying each board but I haven't worked with active Xovers before so maybe there is a good reason for it, or the op amps have enough PSRR ?.

Sounds like a great way to experiment with speakers though, Might have a go with them myself when I get though all the other things on the bench :D

Thanks re the Videos think I do some more when I get time

Alan

Qwin

21-06-2016, 10:57

Re; PSRR, the op amps seem very tolerant of voltage variations so regulation is not specified for this application, for the very small additional cost I will be using it in future though. I have a few of these lying around, £5.64 including postage, I change the Chinese caps for some basic Panasonic FC's.

http://i.ebayimg.com/00/s/MTAwMFgxMDAw/z/M~oAAOSwGIRXaLzb/$_57.JPG

Yeh, active gives you a lot of scope for experimenting.
Because you use separate amp channels for each driver, mixing 4ohm and 8ohm drivers is not a problem, as each amp just sees a single load. If the drivers have different sensitivities, just turn the amp up driving the inefficient one to match. You can also use a muscle amp for the woofers and a lower powered high quality amp for the mids and tweeters. These are the things that sucked me into going active. With passive circuits you quite often have a situation where you address one problem and create one elsewhere due to it all interacting, so you really have to know what your doing. I think active is much more DIY tweak-able, as you can tackle a problem on one individual driver without effecting anything else. Slight over simplification there, as you still have to watch out for phasing issues etc but there are far fewer variables to deal with as a whole.

For complete flexibility DSP is a great tool, and allows easy insertion of filters to flatten the response, or for room correction, but I still shy away from putting my analogue sources through an ADC then back through a DAC in a DSP. I would probably develop using DSP then transfer the knowledge to an analogue circuit. So much easier that way, you can move the x-over point live and listen for the changes using DSP, for my analogue boards that would involve changing 80% of the components every time I changed the frequency.

If you get the time Alan have a go, I started by getting some ready made boards from KMTech, I specified the stock x-over frequencies for the Yamahas and straight out of the box, regardless of their fairly cheap components, they outperformed the stock passive circuit for clarity, detail and dynamics. The bass really improves having a dedicated amp channel per driver. I don't think I'd ever go back to passive on a speaker project. ;)

Qwin

29-07-2016, 14:21

Well I've lived with the active crossover for many months now and tried all kinds of variations to the components. Preferences are very much a case of individual taste so YMMV but here are my findings, which only apply to this application.

Capacitor Choice.
With this KMTech circuit, I tried the stock supplied Polyester caps which are 5% tolerance.
LCR Polystyrene caps rated at 1% tolerance.
Vishay MKP 1837 (PP) rated at 1%.

The Polyester aren't bad at all, reasonably transparent, a bit edgy with the treble, still a big improvement over the passive circuit.

The Polystyrene are slightly more transparent, have a smoother top end and a much better defined bass, both in richness and depth, dynamics are also improved.

The PP are the most transparent of the three, detail is also a step up from the others. Bass has a level of detail and texture the others can't match, overall a richer mix tonally and lovely handling of vocals, easily my favourite and half the price of the Polystyrenes.

Resistors
I tried the stock items, some PRP 9372 and Takman, all were 1% metal Film.
If there was a difference it was very very subtle. They are all good quality parts of a similar type so differences would be small anyway, too small for me to pick up on.

Op Amps
I tried a full complement of the Burson discrete units, this initially impressed with the transparency, but gradually I grew less enamoured by it. It was all a bit too HiFi and lost some of the dirt, grime and rough edge character that brings things to life. I found better results by mixing and matching, this applied to other op amps as well, a full compliment of five, of any one type, never sounded quite right. The area where the Burson shone best was in the buffer stage, the neutrality and transparency having a knock on effect to all the downstream filter stages, LM4562 ran a close second. For the low pass Bass, AD825 was by far the best of the bunch I tried, great detail and slam in the upper bass and goes deep, Burson units were my second choice here being a touch polite by comparison. AD8066 works very well for the high pass on the tweeter, great tone and clarity which carries over into the upper mids. For the mid range band pass, it really does come down to personal preference, for me LM4562 worked well in both locations (Highpass/Lowpass) and sits so well dovetailed to the AD825 for added slam which can literally make you jump. The Burson worked in the lowpass part of the bandpass marrying well to the treble. The LM49720 (metal can) also worked well in this location and I still have it in there.
To sum up, use no more than three of the Burson units in the five locations, but make sure one of those is the buffer. It's all good after that but starts to get subtle in the changes and depends very much on your voicing preferences. I won't be moving the AD825 from its bass duties for instance, but that's just me.

I probably have the Yamaha's as good as they can be sonically, from my point of view, with my preferences.

I've had great fun experimenting with this circuit and learned a lot in the process, that will carry over to other projects. I can really recommend it, for anyone wanting to try a simple analogue active solution.

I just adapted a pair of my old boards to 2-way, crossing at 310Hz and have a pair of two way bookshelf speakers acting for mid/tweeter with their passive circuit still in use and the Yamaha 12" bass drivers connected for bottom end duties below the 310Hz. I've had this lash up running for almost a week and have had no desire to change it. It sounds so right that it has me thinking of a new project combining active/passive and reducing the number of amplifiers required. :D

Macca

29-07-2016, 19:32

You've blazed a bit of a trail there Ken. Certainly be less work for anyone following you in activating these speakers.

Qwin

31-07-2016, 09:25

Thanks Martin, folks may as well benefit from my experience and the numerous things I tried.

Note: In this application, the stock set of OPA2134 supplied, just sound too smooth, rounded off and a bit shut in, lacks air, bite and attack.

Qwin

25-10-2016, 11:01

I've been quite on this project for some time.
It's taken the back seat as I do some experimenting purely on the active crossover circuit.

I have a two way circuit I'm working on, everything I learn from this will eventually carry over to the Yamaha's three way circuit.

There have been way too many things to cover in detail, some of the things I have tried/incorporated:

Converted the single ended input to balanced by summing the two halves of the balanced input through the buffer op amp.
This involved cutting PCB tracks and soldering point to point alternatives adding a couple of resistors and changing the value of two existing resistors.
To do this I also had to modify the 2 terminal phono socket to a three terminal screw block.

Changed the basic Lytics de-coupling the power rails to Elna Cerafines, also tried Panasonic FC and FM before settling on the Cerafines.

Used Vishey MKP1837 (PP) caps instead of Polystyrenes for the filter elements.

Removed the basic 0.1uF ceramic bypass caps on each op amp that were across the power pins and replaced with 4.7uF Tantalum/0.1uF ceramic in parallel from each power pin to the ground plane. Added a 0.1uF Wima FKP2 film/Foil across the power pins on each op amp.

The board designer used dual op amps throughout so he could supply one type of op amp. This made it idiot proof for assembly, any chip would work in any socket, providing you fit them the right way round in the socket. Trouble is the buffer only uses one channel, op amps being fairly unstable devices, I have considered whether this could be a problem, the unused side is powered up but has no input or output tracks to the pins, it may be an issue, but I'm only guessing. To rule this possible source of distortion/noise out, I have changed the pin out on the buffers dip8 socket to accept a single channel op amp, which has different +V and Output pin allocation.

I have just ordered some of the SPARKOS LABS single channel discrete op amps to try in the newly configured pin out. The other reason for changing to single channel is they are a lot cheaper than duals, especially when you get to discrete prices.

I changed the basic power supply that KMTech offered to a Chinese regulated supply board, I swapped out the Chinese smoothing caps on the board for Panasonic FC's.

In changing the power supply board this also required a new transformer as I increased the voltage at the rails from an unregulated +/-9.2vdc to a regulated +/-11vdc. The Burson v5 discrete handle an absolute maximum of +/-15v but they were running a bit hot at even +/-13.5v so I backed it off to +/-11v. They run a lot cooler at this voltage and would probably be OK up to +/- 12v.

I had to increase the PCB pillar height to accommodate the mods I did to the underside of the PCB and give good clearance to the metal enclosure, also wired in some XLR sockets using Klotz MC5000 cable.

So I've been quite busy.

When I get to a final solution, I will post details and pictures of the mods I used for the Yamaha's and start a new thread for the 2-way project.

Qwin

08-01-2017, 11:39

With all my dealings with the NS-1000M it is clear, to me any way, that the mid range driver is the star of the set up, world class. The tweeter blends well with it, but has only average performance and drops off a cliff at 15kHz. With the high crossover point and gentle slope the mid driver is extending well up the range, switch the tweeter off and you barely miss it. The bass does its job, is fast, packs a punch and again blends well, but rolls off fairly early.

I sometimes wonder if this mid dome could be matched with a modern Tweeter and bass driver to give even greater ability across the whole range?
When I get my Scanspeak 12" bass drivers, it is very tempting to try these and my Fountek ribbon tweeters with the Yamaha mid dome, just to see!

I looked for spare mid domes on the bay, the only thing I found was a single driver and they wanted £284 for it.:(

If any one is interested a guy in Greece is selling a pair of brand new (NOS) bass drivers and a pair of tweeters, all still in the original boxes (on the Bay), he wants over £900 for the pair of bass drivers though.

Firebottle

08-01-2017, 12:46

Trouble is the buffer only uses one channel, op amps being fairly unstable devices, I have considered whether this could be a problem, the unused side is powered up but has no input or output tracks to the pins......

If you (or anyone else) is in this situation again then the best thing is to just earth both input pins of the unused op-amp.

Keep the updates coming Ken :)

Qwin

08-01-2017, 14:44

Thanks for the input Alan, I will remember that one.
Of course the other reason I changed to the single channel is the monetary saving £40/opamp so £80 less for the pair of single channels.

I'm spread a bit thin at the moment, with my semi active three way project taking most of my time and funds.

Hope to get back to the NS-1000M soon. I'm considering going over to the dark side and looking at DSP. After hearing Ali's 401/slate turntable and Benz Gullwing cartridge through a miniDSP, my concerns about degrading the analogue sound through an ADC/DAC have evaporated.

Qwin

12-01-2017, 13:33

Apart from making small tweaks to the components in the analogue active crossover, which will make only small changes to the sound, the NS-1000M are about as good as I will get them with the stock crossover points.

I have been experimenting with other speakers, but I keep coming back to the Yamaha's and their exceptional mid range clarity and imaging.

They are not without their faults, as I have mentioned before, the bass falls off early, with an F3 above 50Hz. It's all about that fabulous 3 1/2” Beryllium Mid Dome, it certainly dominates the voicing. It has a high upper crossover point of 6kHz and with the shallow roll off from the simple passive second order circuit, it will still be contributing at 10kHz and beyond. The tweeter has a bit of a ragged response at its lower end and drops off quickly at the top (15kHz). The dominance of the mid dome can be verified by switching the tweeter off. I can do this easily on my active system with the flick of a switch. It really doesn't contribute much, other than a sparkle to symbols etc. Having said all that, I keep coming back for more, so there is something fundamentally right about the sound.

I've been asking myself the question, as to what would happen if the tweeter crossover point was lowered and/or a steeper slope was introduced. This and maybe replacing the tweeter with a modern ribbon tweeter, I have been very impressed with the Fountek NeoX 1.0 I've been working with. Then there is the bass, keep it as a sealed box, but use a driver that goes a bit deeper?

There are lots of variables, I certainly wouldn't want to loose the signature sound, but maybe loosing some of the steely sharpness at the top might be a good thing. This might make them more Amp friendly, as they do need a bit of care when matching?

They obviously would no longer be an NS-1000M, but maybe a modern take on the theme? This is an image of the kind of thing I have at the back of my mind. “NS-1000 Revisited”

http://www.jkwynn.co.uk/Project_Images/Yam/yam_35.JPG

It would need a lot of experimenting with crossover points and slopes, so a minidsp would be used to help achieve this. I have had concerns in the past about using a DSP with my Turntable, but after hearing Ali's very good TT recently, through just such a set up, I am less worried, as the results were excellent. I would love to at least try this at some point and see what can be done, we all aim for that audio Nirvana, I think mine is based somewhere around the NS-1000M Mid Dome.

AlfaGTV

12-01-2017, 19:00

This looks really intriguing Ken! Looking forward to reading more on your adventures in NS-land! ;)
And i might add, my bog standard NS1000M were used as discotheque speakers on new years eve! I had my doubts wheter they would survive the heavy pounding they took from several hundred watts and disturbingly loud SPL's!
But they sailed through, with several compliments on sound quality from the dancing guests!
Cheers /Mike

Qwin

13-01-2017, 11:24

I once took some big wharfedale speakers to a party, they were played so loud no one could hear one of the tweeters had failed, blew a cap in the crossover, so an easy fix.

I've got lots of ideas floating round in my head for the Yamahas, I am getting the Scanspeak bass drivers for another project, and I want to experiment with a DSP set up, so I might get to try a lash up similar to the one in my picture.

Don't ask me when though :)

I will keep adding my findings to the thread.

Qwin

14-01-2017, 16:28

I didn't take this picture, but it's of the Scanspeak 30W/4558T00 12" Subwoofer.
It's featured in my sketch and I think it would make a good replacement for the Yamaha 12" Woofer and add some missing depth to the bass.
It's a fairly substantial driver as can be seen.

http://www.jkwynn.co.uk/Project_Images/Yam/yam_36.JPG

I think, because it has a low Fs of 17Hz, a reasonably large XMax and power handling, Scanspeak decided to place it in the Subwoofer category, it certainly gets used a lot for that application. But DIY builders including Troels Gravesen have used it as a straight Woofer in a three way. Below is Scanspeaks own frequency sweep for this driver.

Edit: It's worth pointing out this would be best implemented in an active set up, as this driver is 4ohm and also 4dB less sensitive than the 8ohm mid dome.

http://www.jkwynn.co.uk/Project_Images/Yam/yam_36A.JPG

Don't be too alarmed by the peaks and troughs in the lower section, Scanspeak plots always look like this. Where most manufacturers use smoothing on the plot, or include nearfield response at the lower end, to make it look better, Scanspeak prefer to give you the true results warts and all, they have to be applauded for doing so.

The thing that struck me about this plot, was how flat it is in the upper bass and how far it extends up the range, definitely a response more like a Woofer than a true Sub, with its typical hump back curve.

This driver has an Efficiency Bandwidth Product (EBP) of 49 making it best suited for a sealed enclosure.

Taking the standard Qtc figure for a flat response of 0.707 this computes as an F3 of 38Hz in an ideal 55L (Net) sealed box.

So, by keeping the same footprint for the cabinet, it would only be around 100mm taller than the 50L NS-1000M. It would probably perform well, if not ideal, in the Yamaha cabinet. Sealed cabinet pairings are less fussy about exact volumes, compared with tuned ported cabinets.

It certainly ticks a lot of boxes for the Yamaha's, so not surprisingly, the spec grabbed my attention when I was checking out drivers for my other project.

sq225917

14-01-2017, 23:45

Ribbons all the way...
https://c2.staticflickr.com/8/7371/9622892030_b6dc10f548_b.jpg

walpurgis

14-01-2017, 23:51

Why? The NS1000 tweeter extends up to around 40kHz.

What impedance are the ribbons and what filter are you using?

Qwin

15-01-2017, 10:39

The tweeter is claimed to go up to 40kHz but at what levels?
Troels Gravesen and others have had consistent measurements, showing it drop off a cliff at 15kHz with no significant output above this.
I will get round to measuring it myself at some point, but I have no reason to doubt the results which were identical for both tweeters averaged over 10 readings and measured on axis. Yamaha's own plot shows it dropping off rapidly from about 13kHz, being 10dB down by 20kHz, from its average of 93dB. (L-Pad set to normal)

That looks like a Raal Ribbon Simon?
If its like the Fountek ribbons they are listed as nominally 8ohm, but have an impedance matching transformer built in. Because of this the NeoX3.0 weighs over a kilo.

My efforts, if I get that far, will be in an active set up, so individual driver impedance is not critical, as long as its not extreme for the amps.

I read your nice build thread a long time back Simon, but can't remember whether you are passive or active? and I could have sworn they were gloss Black?