DIY phonostage.

[sq225917]

I've just finished the upgrades to the amp channels. I refinished the paintwork on their cases as well, it was looking grubby and marked. I'll pull the transformers out tonight and fit the new ones tomorrow, then i'll tidy up your psu cases as well. I'll give it all a listen and a good soak Thursday and probably drop them off Friday if you're around?

I'll stick you a set of Mark's acrylic front panels on as well, it irks me to see a Paradise without correct faceplates.

[Audio Advent]

Looks sexy!

Some lead a sheltered life...

[Wakefield Turntables]

Some lead a sheltered life...

:rfl:Yep, your probably correct. I'll be thinking of you whilst listening to the Paradise transcribes my vinyl and taking a sip of my favourite 18 year old single malt.....

[sq225917]

Andy, i'm not going to get these finished tonight, i need to pick up some extra 10 watt resistors tomorrow. So i'll come over some time on Saturday with them if that suits you. Let me know when is good for you.

https://c1.staticflickr.com/1/628/22203177308_0b848d42e5_b.jpg
New toroids

https://c2.staticflickr.com/6/5642/22203212580_f9c4261301_b.jpg
New jfets, organic polymer caps, new loading resistor

https://c2.staticflickr.com/6/5806/22378083422_ca6bab5da1_b.jpg

new shiny cases

[Wakefield Turntables]

Si, sent you a PM.

[sq225917]

How's that settling in then Andy?

[Wakefield Turntables]

It's not needed any settling in. I can hear a massive difference already. I really do need some time to adjust myself. The valve system has never sounded so good. I can't wait to do some low level listening tonight with some Jazz and my favourite single malt :eyebrows:. It just sounds so effortless, I have no idea what you plonked in their but it's been worth the money.

[sq225917]

http://www.diyaudio.com/forums/group-buys/282634-paradise-phonostage-pre-regboard-psu-group-buy-v2.html

New group buy thread, if you've got a paradise already without the full spec psu or are thinking of building one of hen do he group buy will offer you about £100 saving over getting the parts and boards on their own.

[dowser]

Will I just be able to order the parts kit Si (already have the boards :))? Thanks, Richard

[sq225917]

The group buy is just for the psu parts and boards, minus mains transformers. I hadn't thought to offer just PCBs. If that's what you want add your name and a note to that effect.

[Wakefield Turntables]

Finally joined the Charcroft club! Ordered and awaiting delivery, nice to see how they affect the paradise.

[usernam]

sq; clever.

[sq225917]

That group buy doesn't look like it will reach critical mass, which is a shame. There's not much to touch a Paradise with the full bore psu.

[YNWaN]

https://c2.staticflickr.com/6/5806/22378083422_ca6bab5da1_b.jpg

new shiny cases

Nice front panels...

[Wakefield Turntables]

Yes they look v.nice on my hifi rack, thanks.

[martinjohn308]

I played around with 3 sets of resistors for over a period of 6 weeks on my Paradise phono stage before ordering the Charcrofts. Expensive but well worth it in my opinion.

Regrds,

Martin

[Wakefield Turntables]

I played around with 3 sets of resistors for over a period of 6 weeks on my Paradise phono stage before ordering the Charcrofts. Expensive but well worth it in my opinion.

Regrds,

Martin

Glad to hear, I plumped for 150ohm what about you?

[YNWaN]

150 Ohms is pretty low.

[Wakefield Turntables]

150 Ohms is pretty low.

What do you suggest?

[sq225917]

Not everyone uses a Troika Mark...

[YNWaN]

Well they should ;); but I know that loading is both taste and cartridge dependant. I guess my somewhat oblique point was that stating ones choice of loading is meaningless without relating it to the cartridge.

150 Ohms may well be optimal in Andrew's situation - it's unlikely to be the best for Martin though.

[martinjohn308]

I finally settled on 220 ohms i found that hit the sweet spot for me.

Regards,

Martin

[sq225917]

I've just put an order in for RIAA and bridge resistors for my own rebuild. £110 for ten resistors from the US!!!

[Wakefield Turntables]

Well the Charcroft's have arrived. Plonk them in tonight and have a listen.


I've just put an order in for RIAA and bridge resistors for my own rebuild. £110 for ten resistors from the US!!!

Why go to the US? Which brand are you using?

[sq225917]

They are Vishay Naked Z-foils 0.05% tolerance. Pretty much the same part as Charcrofts but these are in a range of values and tolerances that you can't buy in the UK, 18k2 and 147k. Vishay licence the fabrication of these parts out to a few manufacturers, it's not the sort of 'volume' business that Vishay are interested in themselves, they just bought the company who came up with them.

I'm ambivalent about making the RIAA section up with such high precision, low noise parts. Experience tells me that i think I can hear a difference using them in the cart loading position, so replacing the bridge and all the RIAA resistors should be more of the same. Frankly it's the last Paradise I'm ever going to build for myself so it's getting everything possible thrown at it, no expense spared, no matching too time consuming.

[Wakefield Turntables]

NICE! Once you've finished you'll have to tell me whats under the hood ;). The Charcrofts are sounding pretty nice incidently.

[sq225917]

You're up to date mate. If these make a difference I'll let you know.

[Wakefield Turntables]

You're up to date mate. If these make a difference I'll let you know.

ok thanks.

[sq225917]

Here's my new shiny waiting for the RIAA parts.

https://c1.staticflickr.com/1/602/22987230863_c1c7f940cc_b.jpg

[Audio Advent]

Out of interest, what is the "EXT RIAA" connection? Does it bypass the RIAA on the board (for other curves I guess) or does the RIAA need to be unpopulated for its use (assuming it's for external RIAA)?

[sq225917]

It bypasses the RIAA section on the board and allows you to take the input, after the gain stage, out to an external RIAA board.

Here's number 17, just balancing the Calvin buffer after winding out the input offset.

https://c1.staticflickr.com/1/725/23755666385_d0db856b5b_b.jpg

[Audio Advent]

It bypasses the RIAA section on the board and allows you to take the input, after the gain stage, out to an external RIAA board.




Cheers for that.

[sq225917]

No worries, are you thinking of building one?

[sq225917]

Early Xmas presents to myself.

https://c2.staticflickr.com/6/5673/23828569371_99bab9a3c9_b.jpg

Fitted.



https://c1.staticflickr.com/1/663/23829085531_13e5f32452_b.jpg

[337alant]

Very Nice
Let is know how they sound?.

Alan

[sq225917]

Hard to say as I've built into an all new build, with different values in the Calvins, swapped back to my old psu, and refitted my cart all at the same time. I can certainly say it sounds no worse. In fact I think this Paradise sounds exceptional, but I don't know where attribution should lie.

[sq225917]

Paradise pre-reg psu group buy, kits or pcb only (http://www.diyaudio.com/forums/group-buys/282634-paradise-phonostage-pre-regboard-psu-group-buy-v2-2.html#post4578022)


Just a note to anyone who's interested, all of you who may have a half finished Paradise but are waiting on the suitable PSU solution. Have no fear PSU pre-reg group buy V2 is here.

[danilo]

Ohhh Blingy resistors there :-)
Curious such devotion to resistors when the Transistors, seemingly outnumber the resistors.
Note that transistors, not being created even remotely 'equal' ... should be a whole new world of worry :D.

[YNWaN]

The transistors are matched to closer than 1% tolerance so at least as much, if not more, worry has been devoted to them.

[Audio Advent]

No worries, are you thinking of building one?

Ah, no.. I bought someone else's own build already! But err... my system is in a bit of flux and I haven't used it other than checking it worked. So at some point I will need to go through the diyaudio thread and work out what loading etc might work best as a general setting.

[sq225917]

Back from the dead.

I've built another two since this thread was last active. The group buy died and some German asshat ran off with 6500 euros of our money...

Any way the group buy for the PSU pcbs has been reinstated and details are to be found here.


Paradise PSU group buy- its alive (http://www.diyaudio.com/forums/group-buys/282634-paradise-phonostage-pre-regboard-psu-group-buy-v2-22.html#post4797682)

[337alant]

Thanks Simon I have just signed up for 2 sets
Excellnt job on the circuit boards and your doing a good deed for everyone so well done you.

Alan

[sq225917]

Power supply boards are here, I've shipped a few out to early responders, yours are in the post Alan.

I think I'll end up with a couple of sets spare for anyone who missed out, £24+ p and p for a set of four boards. Well worth it, as they build to a spectacularly quiet choke regulated psu.

[337alant]

Great news Simon and thanks very much for you efforts

Alan

[Ali Tait]

Simon, I'll take a spare set if there's still one going please.

PM payment details and I'll send my address.

[sq225917]

Here's the BOM for anyone who's interested


2x Talema 70065k transformers: https://www.electronic-shop.lu/EN/products/133830 RS doesn't have them on stock until January !
4x Epcos TDK 82uH common mode chokes: http://www.mouser.co.uk/Search/ProductDetail.aspx?R=B82724J2501N1virtualkey592500 00virtualkey871-B82724J2501N1
8x LQA10T300 high speed zero recovery diodes: http://www.mouser.co.uk/Search/ProductDetail.aspx?R=LQA10T300virtualkey59040000vi rtualkey903-LQA10T300
16x 0.1uf (100n) wima caps: http://www.mouser.co.uk/ProductDetail/WIMA/MKP2D031001F00JO00/?qs=sGAEpiMZZMv1cc3ydrPrF7l45uRd9dVhK2iviYXLQ3E%3d
4x Caddock, 20 ohm, 30 watt thick film resistors: http://www.mouser.co.uk/Search/ProductDetail.aspx?R=MP930-20.0-1%25virtualkey68400000virtualkey684-MP930-20
4x Hammond 158T 300mA chokes: https://www.jukebox-revival.eu/staging/smoorspoel-hammond-158t.html

[sq225917]

I've been tinkering with a new toy.

https://c2.staticflickr.com/6/5527/30067134050_242b2e396c_b.jpg


Tis a little square wave generator that can be used for stimulating ringing in transformers and chokes to help set optimum snubbing values.

https://c2.staticflickr.com/6/5537/30343801546_ec7d929cc7_o.jpg

Works well the numbers you want for the quietest paradise psu are given above

[karma67]

No worries, are you thinking of building one?
hi,is this project suitable for a 1st time project?
im ok with the iron but all the technical terms go over my head,whats the ball park cost involved in getting one up and running please?

[YNWaN]

No, I wouldn't say it is suited as a first time project - not even a second one - maybe an ambitious third one. It's not exactly a budget build either.

[karma67]

cheers,do you have anything in mind i could have a go at?
ive done a few crossovers and interconnects,thats my experiance.

[Arkless Electronics]

http://theartofsound.net/forum/showthread.php?46403-Anyone-interested-in-a-DIY-valve-line-stage

Not a phono stage but it doesnt get any simpler as a 1st project..

[YNWaN]

I would listen to Jez in this. With the Paradise the only bit of it that is a kit is the circuit board (if you can find a pair) is made! You have to buy all the components (and there are loads of different types and options), also source and build a power supply (not any old one) - you need two of everything of course and a whole bunch of the transistors have to be matched - this needs measuring equipment and you need to buy a couple of hundred to select the half a dozen you need.

:-).

[sq225917]

The difficulty with the Paradise is that the transistors require matching and that requires good circuit building skills of specialist kit, not costly, but you need to know what you're looking for, matching them for current gain and noise if possible. The arse is that you'll ned about 500 parts to get the match you require for a complete paradise and that takes time, lots of time. And the total parts cost included cases in nigh on £900 for a well specced version with competent power supply.

On that basis alone it's not really a first project, as Ynwoan says maybe a 3rd project once you've learnt you way around.

[danilo]

You could ? try this;
It's working well beyond my expectations.
MM Only! though, and the parts specs.. Are.. important. And yes they are easy(ish) to source.. even today
A couple of 9 volt batteries as power..draw is Tiny so they live surprisingly long
Can one do better? Always :eyebrows:
But this is far closer to Really Damned good than you would think, certainly more than good 'nuff for MY lps.
Simple to build as bonus

[Arkless Electronics]

You could ? try this;
It's working well beyond my expectations.
MM Only! though, and the parts specs.. Are.. important. And yes they are easy(ish) to source.. even today
A couple of 9 volt batteries as power..draw is Tiny so they live surprisingly long
Can one do better? Always :eyebrows:
But this is far closer to Really Damned good than you would think, certainly more than good 'nuff for MY lps.
Simple to build as bonus

Hmmm... could sound pretty good but the lack of source degeneration means FET's may need to be selected for it to work properly.

[RothwellAudio]

Hmmm... could sound pretty good but the lack of source degeneration means FET's may need to be selected for it to work properly.
I think you're being quite kind - it's actually a very poor circuit and the performance you get will be a pot-luck thing depending on the characteristics of the FETs you happen to get. Even the DC bias is left to luck.

[Arkless Electronics]

I think you're being quite kind - it's actually a very poor circuit and the performance you get will be a pot-luck thing depending on the characteristics of the FETs you happen to get. Even the DC bias is left to luck.

Hence my saying it would need selected FET's to work.... meaning at all! I would expect bad distortion from it as well but some may like its sound...

[Marra]

I can recommend the Salas Folded Simplistic phonostage which can be configured for either MM or MC cartridges.

http://www.diyaudio.com/forums/group-buys/241736-gb-salas-folded-simplistic-phono-pcb.html

[danilo]

Hmmm... could sound pretty good but the lack of source degeneration means FET's may need to be selected for it to work properly.

Absolutely True!
This being a slightly simpler variant of a 'Le Pacific' circuit.

Parts selection IS critical .. substitutions may seriously mess it
Note that it states that.. Very Clearly.. on the schematic.
Happily there are Very Few parts involved... KISS .. clearly has dominion.
Good news is one can Buy a such a precisely matched Jfet Quartet for ~6$ plus shippings.
Even today... I bought a few sets.
Those Toshiba Jfets are basic Pass bldg block items.
The Siemens 1% caps are also easily found.

MY cobble up works V well actually.. surprised by how well TBH .
My Vinyl Playback [within the agreed limitations of using a MM cart .
sounds as Good as I've heard in ~45 years of Fooling with Audio.
I find it more to my liking than the Booze Hound circuit
I shall Not be upgrading this...I have no reason to dick about yet more.

Don't subscribe ? Hey.. not trying for converts here .
Just trying to help/offer an alternate path to the asked question.

Marra:
Yessss The Salas Circuit is better.
But the build I heard offered only a ..small..(arguablly any actually) improvement on MM and as I have no want of MC capability ?.
I wasn't willing to bother with the added complexity for... My needs.
This one took about an hour.. from start to listening

[danilo]

I think you're being quite kind - it's actually a very poor circuit and the performance you get will be a pot-luck thing depending on the characteristics of the FETs you happen to get. Even the DC bias is left to luck.

Surprised Mate..that reads as Amateur ? Beginner talk

[walpurgis]

Surprised Mate..that reads as Amateur ? Beginner talk

To whom is the "amateur" bit addressed and who is the "beginner" danilo?

[Arkless Electronics]

:rfl: Oh stop it! The irony is beginning to hurt:D Bet Danilo's looking forward to a Pass Labs belt buckle for Xmas.... and "The Ladybird Book of Electronics".... Illustrated.

Andrew, it's been refreshing to see another EE on here with actual opinions... Danilo would do well to listen to them, as you seem to talk sense:)

[RothwellAudio]

I'm surprised at the "beginner" and "amateur" comments, but not particularly upset. Though I was having a pop at that circuit, no insult to any forum member was intended.
Being a bit sad I read technical papers for an evening's entertainment. Ironically, I was reading a paper from Siliconix only a few days ago all about biasing JFETs. I wonder if they're bebinners too? :scratch:

[sq225917]

Back on track with the Paradise, a not piece of guess work pot luck. 160+ carefully selected components per channel.

https://c2.staticflickr.com/6/5780/30895901792_cc6a9be8c5_b.jpg

Destined to end up in Manchester.

Sounding double fine in my house at the minute. makes me realise I really do have to upgrade to the dual box psu.

[337alant]

Very Nice Simon
Any good reason for the red Leds ?

Alan

[sq225917]

They're lower current so an extra one is needed either side for the ccs to draw the correct current, and that extra one can be carried to the back of the front panel as an indicator that each side of the shunt reg is functioning. I've just left em on the edge of the board in this build but I wanted to try it as an option. The lack of a visible 'on' has always bugged me.

[sq225917]

It's been a while. I've had a part-built sat waiting for cases for several months. It's probably the last Paradise I'll scratch build and the demise of Andrew P's Naim shoebox cases has left me high and dry. So... I put some money down and ordered some new full width cases. These things are lovely, turns the Paradise into a real visual statement.

I'll order up cables and sockets and get cracking with this over the next few weeks. I've never done a big two box build with my psu boards so this should be a cracker.

https://farm3.staticflickr.com/2844/32863432814_8cc9a9e0da_k.jpg

I've also got final set to build, for none other than Joachim, the designer...

[Bigman80]

Looks fantastic !!!!

Sent from my EVA-L09 using Tapatalk

[martinjohn308]

Hi Simon,

Looks a great build as usual, mentioned the Paradise on the Linn Forum and not one comment I suppose because it's not a Linn phono stage but it I did mention it on the " 3rd party product ".

Regards,

Martin

[sq225917]

I think I built Paradises for four linnies, two replace uphorik, one an ulrika and one Linto.

[sq225917]

This one's going to a chap in Wales.

[Bksabath]

Just found and read this tread top to bottom ...
Better keep the tech discussion here
Simon what HFE you use for the transistors?
I got several batches going from 340 to 620 Hfe
The last 1000 of each arrived yesterday from Mouser those at Fairchild on tape I have measured a few and the 337 are around 420 t0 490 the 327 are around 360 so pretty useless.

BTW I would be interested in a set of boards for the PSU do you have a spare set?

[Bksabath]

I am not going to moan about Paradise PSU whiteout doing sumtink about it
Consider that hammer and Mig are my tools of choice but ... If you willing to help ...
I am working on those (second version) once finalised I am planning on a small professional run

[Bksabath]

BTW if you need same... I am more than happy to help any one

[sq225917]

Sandro if you want I have a spare set of the psu shunt boards for the paradise.

[Bksabath]

Just sent you Pay pall money :D

[Bksabath]

Simon
Boards arrived today

Just had quick look
Awesome

Many tanks

[sq225917]

Excellent, the snubber value was measured with the Quasimodo for that exact transformer. You'll need a 10-15r dropper resistor for each rail, depending on your mains voltage.

In the meantime I've been struggling on with No21.

https://c1.staticflickr.com/3/2905/33994834906_00e33dee68_b.jpg
Buffer boards looking tidy with the new machine wound inductors.

https://c1.staticflickr.com/3/2897/33223320853_0bb51ea71e_b.jpg
You could get a bus under there.

https://c1.staticflickr.com/3/2941/33994763626_a691066ca1_b.jpg

Should be done in a week or so. Still a few parts to arrive. Then I've got to crack one with one for Joachim and that's the last one ever...

[Bksabath]

I've got to crack one with one for Joachim and that's the last one ever...
:rfl::rfl::rfl::rfl::rfl:

There was a mention of a super version One suppose is kept secret...

[sq225917]

Yeh, he wants one as a benchmark to use for his new phono stage development. It's a simplified paradise with balanced input stage using monolithic rather than all discrete devices. It makes sense to do a simplified version as the paradise takes too long to match and build to be commercially viable.

[Bksabath]

I am struggling whit remote PSU positive side is oscillating like mad
I just don't have the skill to cure that Hfe 385 +- 0.5 % 470P 2.7 N and 47 P + 2SK 170 8.5 mA and 365R
All caps in place 470 uF Panasonic FM and 100 uF 10 uF silmic and crapy 100 uF whit 15 R tracks are so much shorter than original PCB still :scratch:

[sq225917]

https://c1.staticflickr.com/3/2840/33365758423_f15fcce9b6_b.jpg

I boxed this one up today, it's going to Wales.

I can crack on and finish Jochim's.

Sandro, what freq is it oscillating at? Have you tried putting the decouplers on the heatsink? 0.47uF film cap direct onto the underside of the board, from the emitter to ground on the PNP and the NPN on the outside of the heatsink. Essentially they do the same job as the electrolytics on the psu input (which knocks put cabel inductance effects) but they're local, and sized to mop up HF in mHz range.

https://c2.staticflickr.com/2/1519/26057002053_6b800fbc37_b.jpg

Show me what you've got.

[Bksabath]

It is on the remote home brew boards main thing I try to do is to put the heat sink's outside the box
Will get scope/ pictures later on.
I have 47P on C B of Q105

you suggest 47nF on emitter of Q103 and Q203 ?

[danilo]

Damn.. but that thing seems to include More parts than a TV set.
Clearly KISS wasn't a consideration.

[sq225917]

Sandro, yup 47n on emitters. If you've got oscillation coming in around 4-7mhz this should mop it up.

Danilo, I think it's fair to say that sound quality was the only consdieration with the Paradise. ;-)

[Bksabath]

Sorry no pictures
I messed up the boards ...
I am not going to give up but loads of things going on Moving home soon
That 47n on emitters would point out problems whit the long tracks on the PCB maybe?
Agree whit Simon the R2 I built was populated in a rush before all the little tweaks where put in place and it did sound really good I am taking 10 times more building a second one
As mentioned I got a fair bit going on so slight change of plan for now I am going to stick to building a standard R3 version whit all the little tweaks
Master plan has not changed it just going to take a little bit more time
The Paradise is better than most already but I am sure that it could be even better

[Bksabath]

20111Got same presents today (no not really about £140) just a few resistors for the RIAA
Well worth it for R2 Paradise IMO
R3 is going to be super
May be not comercialy viable certainly I would like to listen to the Luxman £4K plus and compare
Any fellow nutter In Liverpool
I know there must be one or 2 considering that Shieffield got probably more than 30 :lol:

[Bksabath]

Slow progress on the R3 boards
PSU's working
RIAA fitted and matched to +-2 Homes 5 Pf
Will need to solder the transistors those matched to +- 1% or better
But moving home in the next few weeks so is going to be a while (probably the best improvement to my FiFi
My SME V is going to be late Been waiting over 3 months for it .

Simon could you let me know where You get the transistors from please
Can any one make any sense out of this
>>>>>>>>>>>>>>>>>>>


Quotation Number: QP23492

Customer Reference: RE Micross Quote Request Form Submission

Thank you for your recent enquiry, in response to which we have pleasurein submitting our quotation as follows:



Line Item

Device Type

Qty

Unit Price £

Delivery



1

Description
• Electrical screening of BC327-40
• Micross to Procure parts-see note 2
• Parameter to be screened at 25°C
Hfe with a Gain of 400 ±10%, as per note 3
• Manuf. ON SEMI
• Date code: 1308
• Lead Free/ RoHS

100

500

£9.52 EA
£6.52 EA

4-5 Weeks



2

Description
• Electrical screening of BC337-40
• Micross to Procure parts-see note 2
• Parameter to be screened at 25°C
Hfe with a Gain of 400 ±10%, as per note 4
• Manuf. ON SEMI
• Date code: 0521
• Leaded

100

500

£9.52 EA
£6.52 EA

4-5 Weeks



3

Part mark passed parts (optional)

Per

£0.25 EA





Notes
1. This quotation is for procurement ofparts and Hfe testing only and parts are to be screened for a Gain of 400 ±10%.
2. Micross will supply best yield from150 devices for a requirement of 100 finished parts or, Micross will supplybest yield from 750 if 500 finished parts are required. If no devices fall within the customer requirement there will be a MLV charge of £550 for each line item 1 and2.
3. Parametersto be tested for BC327-40:


4. Parametersto be tested for BC337-40:












5. Micross is unable to deliver partswith the original manufacturers C of C
6. Pricing quoted is for the totalquantity of any part placed by itself at one time on one purchase order, withdelivery not to exceed twelve months from the date of order placement.
7. Purchase orders placed for thisrequirement are NCNR
8. Please note for customers own productMicross limits it's liability to the value of the wafer processing servicesprovided only.


As I understand they will charge £1900 for 100 BC337-40 and 100 BC327-40 but if they don't manage to match them they will only charge £1100
Got a few thousands of the little bastard that don't :mental:
???????????????????????????????????????????

[RothwellAudio]

It all sounds a bit mad. BC327s are common transistors that cost 4 pence each when you buy 100 of them. These guys will select ones with Hfe of 400 and charge you £9.52 each for them. However, the usual design approach is to use negative feedback to eliminate component variations so the particular Hfe of any one transistor doesn't really matter.

[YNWaN]

You are rather missing the point.

[RothwellAudio]

You are rather missing the point.

Maybe I am. What is the point?

[Arkless Electronics]

It all sounds a bit mad. BC327s are common transistors that cost 4 pence each when you buy 100 of them. These guys will select ones with Hfe of 400 and charge you £9.52 each for them. However, the usual design approach is to use negative feedback to eliminate component variations so the particular Hfe of any one transistor doesn't really matter.

It has no negative feedback and is DC coupled...

[RothwellAudio]

It has no negative feedback and is DC coupled...

I'm not sure I could even get anything useful out of a BC327 without negative feedback :scratch:
I certainly don't want to pay £952 for the 4(?) I want plus 96 I don't want.

[Arkless Electronics]

I'm not sure I could even get anything useful out of a BC327 without negative feedback :scratch:
I certainly don't want to pay £952 for the 4(?) I want plus 96 I don't want.

They are a perfectly normal BJT and fairly quiet at the bias used. A lot are used in this design. To match properly you need to do it at the conditions seen in the circuit. Matching them with the Hfe setting on a multimeter will not do! Potentially, several matching jigs or one adjustable one will be needed. I can't be arsed to count them but there's probably something like 30 transistors per channel in this design and most need matching. You wouldn't need to match for example all the BC327 to each other, more a case of pnp and npn opposites to each other. To get enough matched pairs with close enough matching you may well need to buy say 500 to build one unit and it will take hours of work to do.... Even though I would agree that the price of the pre-matched ones above is well steep!

[RothwellAudio]

I haven't seen a circuit diagram from this design but when I'm designing something I prefer not to rely on transistors having any specific performance. I thought that's what good design was all about.

[Bksabath]

Well one thing is they will charge 550 even if they can not match them :lol:
On semi sez that they can screen the PNP but not the NPN but advertise them as complementary Case for advertising standard ombudsman ?
Other quoted that to get screened transistor the minimum Purchase Quantity would be 250000 units Maybe around £16000 (possibly much less )
Group buy any one?

I had the R2 built and believe me it is one of the best around
Simon did ear it for about an hour or so and ended up building 23 of those

One thing that I found was that Salas Simplistic had much more accurate staging and I am sure that this is not a Paradise related problem but due to my build whit poor transistor matching

I have the DCA pro (best money ever spent) now and I am sure I can do much better than before but trust me when I tell you testing 5000 transistor will give you sore back

[Bksabath]

I haven't seen a circuit diagram from this design but when I'm designing something I prefer not to rely on transistors having any specific performance. I thought that's what good design was all about.

Quite agree whit you I wish I could design circuits I am much better at hammering and welding
That you have not seen the circuit is lets say not good just a few resistor values to change and use Toshibas complementary pairs may be the answer but this is way above what I can do

[Arkless Electronics]

I haven't seen a circuit diagram from this design but when I'm designing something I prefer not to rely on transistors having any specific performance. I thought that's what good design was all about.

Yes it makes things vastly cheaper, easier and more reproducible to do things that way... but not always better;)

[mikeyb]

I'd like to try one of these stages but it would need to be built for me or are there any demo ones available? I'm assuming not but I live in hope.

[Wakefield Turntables]

I'd like to try one of these stages but it would need to be built for me or are there any demo ones available? I'm assuming not but I in hope.

You need to buy one.

[mikeyb]

You need to buy one.

Ah ok, are the boards still available, sorry I'm still reading through the thread and noticed that it was a group buy that ended ages ago, but I'm only up to page 15 of 36 ( I use Tapatalk ) lol.

[YNWaN]

The boards are long sold out and can now only be picked up from people who bought them and are selling them on. There used to be quite a few available this way but they are now becoming rare and don't come up often.

Simon used to let potential customers try his but I don't think he's planning on building any more.

[mikeyb]

The boards are long sold out and can now only be picked up from people who bought them and are selling them on. There used to be quite a few available this way but they are now becoming rare and don't come up often.

Simon used to let potential customers try his but I don't think he's planning on building any more.

ok, no worries plan Z it is then :(

[Bksabath]

2111521116211172111821119
mikeyb

Do not despair options are :
1)Ask on Diyaudio
2)ask Simon
3)wait a long time for some one to design and make new boards :)

I did a first sketch and print of the boards a couple of years ago posted on Diyaudio Hessener comment was that is a very tight input stage
I am still working on it and made quite same progress

1) get rid of GF . Done
2)improve listening room . 95% need to isolate the wall whit neighbours but that can be done later
3)get new TT and arm worthy of paradise
4) sort out end of chain to be worthy of new TT and arm ( this is constant work in progress just got a pair of Daly Rubicon on loan from Dogodaudio Liverpool
5) obtain a decent number of BC337-40 to finish populating the R3 boards I had on the bench from 2 years ago and that where started before point 1 of this list became the main priority as remaining point 2 to 5 would not have been possible .
6) once R3 boards whit Calvin Buffer are completed dig out old drawings and restart

As you understand it is a slow process for example the next stage point 6 will mean design and build a few prototypes and once that is done and working I will find a decent professional studio to make decent boards
7) Get decent boards made
8) make a lot of peoples happy and a further lot not so ...

It is going to take same time work and money, SEE 1 and 2 where 2 is buy new house whit 15 ft X 16 ft living room, but once U listen to Paradise for the first time I am sure you would agree that it is well worthy :D

[Arkless Electronics]

Another option is to wait until I begin making a phono stage which works on the same almost unique operating principle as the Paradise in a few months time.

[YNWaN]

Are you going to use the alternative output stage?

Also, interesting choice of capacitors in the RIAA - any particular reasons for that combo?

Edit: oh sorry, just read your post properly and I see you are going to add the Calvin buffer boards :).

[mikeyb]

Another option is to wait until I begin making a phono stage which works on the same almost unique operating principle as the Paradise in a few months time.That would be good, always keen to try new stuff.

[howlindawg]

Simon,

Is that a HiFi-2000 / Modu case?
I was looking for a full size Alu case to house the dual PSU but can't find one exactly like I want.
Can either find the size I want in steel or smaller than I'd like in Alu.



https://c1.staticflickr.com/3/2840/33365758423_f15fcce9b6_b.jpg

I boxed this one up today, it's going to Wales.

[sq225917]

Yeh it's a modushop, vented case, 10 mm front panel.

I have one complete Paradise left, with original psu and Calvins. Just needs boxing up and a buyer. It'll defo be the last one I do.

[Bksabath]

Another option is to wait until I begin making a phono stage which works on the same almost unique operating principle as the Paradise in a few months time.
Far above what I can do, my knowledge of electronics is limited to welding and bashing
I was thinking about using complementary Toshibas for the mirrors as they do not have such a wide HFE span as the not so complementary BC337-40
I am sure the PSU would also benefit from a re- design maybe use Zen Mod Shunty

[Arkless Electronics]

Far above what I can do, my knowledge of electronics is limited to welding and bashing
I was thinking about using complementary Toshibas for the mirrors as they do not have such a wide HFE span as the not so complementary BC337-40
I am sure the PSU would also benefit from a re- design maybe use Zen Mod Shunty

The unit I intend to build will have nothing in common with the Paradise other than the transconductance operating principle. All circuitry is of my own design. I don't discuss circuit details as I'm running a business and consider this proprietary ;)

[RothwellAudio]

The unit I intend to build will have nothing in common with the Paradise other than the transconductance operating principle. All circuitry is of my own design. I don't discuss circuit details as I'm running a business and consider this proprietary ;)

Without going into details, just sticking to general principles, what's the attraction of transconductance?

[Arkless Electronics]

Without going into details, just sticking to general principles, what's the attraction of transconductance?

I'm afraid I don't want to discuss even basic principles of the technique... It's no secret but it's not that intuitive and not widely known either...
Its down side is that it's not suitable for budget equipment as without negative feedback the characteristics of the components, especially the semiconductors, become critical and closely matched parts are essential for proper operation. There may be only £2 worth of transistors used but it takes hours going through bags of the things to select ones that match to the required tolerance!

[Macca]

. There may be only £2 worth of transistors used but it takes hours going through bags of the things to select ones that match to the required tolerance!

Send them to Simon SQ to do it for you, I think he secretly enjoys it.

[RothwellAudio]

Two quid's worth of transistors? Nah mate, won't sound any good - you need to be spending at least sixteen quid per transistor and getting them from a guru who sells audiophile transistors. It's all about quality components, see? The stuff that the rest of the world relies on everyday to keep us safe, keep the wheels of industry turning etc. just won't cut it for sound quality. Got to get transistors with the right tone.
:D

[Arkless Electronics]

Two quid's worth of transistors? Nah mate, won't sound any good - you need to be spending at least sixteen quid per transistor and getting them from a guru who sells audiophile transistors. It's all about quality components, see? The stuff that the rest of the world relies on everyday to keep us safe, keep the wheels of industry turning etc. just won't cut it for sound quality. Got to get transistors with the right tone.
:D

:D

[Bksabath]

Without going into details, just sticking to general principles, what's the attraction of transconductance?

All I can say (for lack of knowledge not for mercenary intention) the speed of the Paradise is unbelievable.
If a sound is on the record, the T before the finger completely closes the holes on a flute for example, will be there to ear.

Maybe it would be wise to listen to one before much posting ;)

[Bksabath]

Martin Tanks for advice
Simon I have sent you 4000 337-40 :lol:

No not really
The problem whit the transistor selection and from the paradise design spec. the tolerance was 10% is that the 40 grades of the PNP and NPN pairs ranges from 250 to 650 so far I had no problem whit finding the PNP for the NPN parts it has been much more of a strain most of the batches I have range over the 460 mark
I have a lot on the 500 to 540, Is just pot luck to get the right bag of transistors from the suppliers

I have built the R2 in a rush it was one of the very first and can not even remember what Hfe I used ,still it sounded pretty good , for the next one on the R3 boards I want to use 1% matching and transistors whit HFE of 400 for the input stage, this may be border-line Paranoid skizzo or whatever not sure which as my Mum had me tested at the age of 5 and it was along time ago, and purely because at the same time as the Paradise R2 I had Sallas Simplistic running and the Simplistic sounded better for staging
The 400 HFE came up due to same reports of oscillation, that I did not have.

Once the R3 is up and running I will re design the PCB on single layer boards and whit the shortest track length possible this may sort out the oscillation problems it would be then the time to use hi HFE samples that should in theory sound even better.

I have no mercenary intention nor a business to run (do we really need to go there ;))and I am doing this just because I want to learn new things....

So do not let the transistor matching issue blow out of proportion or be put off by commercial second ends...

[Wakefield Turntables]

I've got 1% matched in mine and they sound excellent although this is soon to be changed to matched 0.01% tolerance in some areas :eek:

[Bksabath]

I've got 1% matched in mine and they sound excellent although this is soon to be changed to matched 0.01% tolerance in some areas :eek:

Cool
where did you get the transistors from ?

[howlindawg]

Cool
where did you get the transistors from ?

From the 4000 you sent to Simon! :ner:

[Bksabath]

:lol:
Cool

[Wakefield Turntables]

USA

[RothwellAudio]

All I can say (for lack of knowledge not for mercenary intention) the speed of the Paradise is unbelievable.
If a sound is on the record, the T before the finger completely closes the holes on a flute for example, will be there to ear.

Maybe it would be wise to listen to one before much posting ;)

I asked Jez a question - why you seem to take offence at that I don't know. How transconductance is related to speed, I have no idea.
Transconductance is an unusual amplifying principle (voltage in = current out) so I was asking what the attraction is. Why would I need to hear any particular phonostage to ask that question?

[Bksabath]

I asked Jez a question - why you seem to take offence at that I don't know. How transconductance is related to speed, I have no idea.
Transconductance is an unusual amplifying principle (voltage in = current out) so I was asking what the attraction is. Why would I need to hear any particular phonostage to ask that question?
No offence taken and certainly not my intention to give, why should I ?
Just stating fact that, probably more in Hi Fi than in any other sector, the proof of the pudding is in listening.
Transconductance ?
I don't know as well, to me is just magic, I have heard the Paradise and that is the impression I had, did Jez tried one ?
For the same reason that I will be going to Doug Brady for a demo next week so I can hear what a £16 each transistor can do ;)

[sq225917]

Resistors Andy, not transistors.

Your transistors are matched to better than 1% for hfe and matched for low noise.

He's has heard one of mine running a single box psu and slightly low rails, 25v. It wasn't a great match to his cart to be honest.


The benefits of transconductance with respect to low level signals are well covered in the 3000+ posts long MPP thread on diy audio.

[Wakefield Turntables]

Simon,

The photo that you requested, I could not link via a PM

21143

[sq225917]

I've got it.

Looks like we'll be adding them to the front panel.

I need to see inside now...

[RothwellAudio]

The benefits of transconductance with respect to low level signals are well covered in the 3000+ posts long MPP thread on diy audio.

Any chance of a summary to save me wading through 3000 posts?

[YNWaN]

Si, no rest for the wicked - you must be most fatigued then ;).

[Wakefield Turntables]

Simon,

The photo that you requested, I could not link via a PM

21143


I've got it.

Looks like we'll be adding them to the front panel.

I need to see inside now...

OH SHIT! Too many beers and just got this post. Will try and cobble something tomorrow after hangover recovery!!!

[Wakefield Turntables]

Si,

As promised pics of inside of Paradise.

211732117421175

A

[sq225917]

Mmm ok. They might squeeze between the rcas

[howlindawg]

So...
What do we do with all these thousands of BC-337 and BC-327 with hFE < 400 or hFE > 450? :scratch:

[sq225917]

I just bin them

[Wakefield Turntables]

Mmm ok. They might squeeze between the rcas

Here's hoping!

[RothwellAudio]

Any chance of a summary to save me wading through 3000 posts?

I take it that's a "no" then.

[howlindawg]

Back at it.
Finishing up the first pass through the BC-337.
So that's 1000 each BC-337 and BC-327 filtered for 400-450 range.

https://farm5.staticflickr.com/4343/36693418695_2a83d17f5f_c.jpg (https://flic.kr/p/XUtmZV)
Bench (https://flic.kr/p/XUtmZV) by Martin Fay (https://www.flickr.com/photos/howlindawg/), on Flickr


https://farm5.staticflickr.com/4350/36693418595_e130b50e8e_c.jpg (https://flic.kr/p/XUtmYc)
Measured (https://flic.kr/p/XUtmYc) by Martin Fay (https://www.flickr.com/photos/howlindawg/), on Flickr

And second pass of the BC-337 now complete - looking good for a better than 1% match in the 430 - 450 range with about 150 pieces in this range to choose from.

Second pass on the BC-327 yet to go. :drugs:

[mac72]

Back at it.
Finishing up the first pass through the BC-337.
So that's 1000 each BC-337 and BC-327 filtered for 400-450 range.

https://farm5.staticflickr.com/4343/36693418695_2a83d17f5f_c.jpg (https://flic.kr/p/XUtmZV)
Bench (https://flic.kr/p/XUtmZV) by Martin Fay (https://www.flickr.com/photos/howlindawg/), on Flickr


https://farm5.staticflickr.com/4350/36693418595_e130b50e8e_c.jpg (https://flic.kr/p/XUtmYc)
Measured (https://flic.kr/p/XUtmYc) by Martin Fay (https://www.flickr.com/photos/howlindawg/), on Flickr

And second pass of the BC-337 now complete - looking good for a better than 1% match in the 430 - 450 range with about 150 pieces in this range to choose from.

Second pass on the BC-327 yet to go. :drugs:

Let me know if you decide to get rid of leftovers 400< >450 hfe

[howlindawg]

Let me know if you decide to get rid of leftovers 400< >450 hfe

Sure.
It might be a while though.
I'm not renowned for my blistering pace. ;) :hotrod:

[Bksabath]

Sure.
It might be a while though.
I'm not renowned for my blistering pace. ;) :hotrod:
Blisters on finger tips :)

I would be also interested in the low range 337 and have quite a few 327 on that range that once I am done I am willing to swap /resell

Will you also do a third pass ,this time using pliers and same room temperature, before they go on the board ? ;)


Where did you get them from? can you also give me supplier part number ,as I said before it is pot luck to get the right batch, and this time look like you got the right ones.

[RothwellAudio]

Back at it.
Finishing up the first pass through the BC-337.
So that's 1000 each BC-337 and BC-327 filtered for 400-450 range.

https://farm5.staticflickr.com/4343/36693418695_2a83d17f5f_c.jpg (https://flic.kr/p/XUtmZV)
Bench (https://flic.kr/p/XUtmZV) by Martin Fay (https://www.flickr.com/photos/howlindawg/), on Flickr


https://farm5.staticflickr.com/4350/36693418595_e130b50e8e_c.jpg (https://flic.kr/p/XUtmYc)
Measured (https://flic.kr/p/XUtmYc) by Martin Fay (https://www.flickr.com/photos/howlindawg/), on Flickr

And second pass of the BC-337 now complete - looking good for a better than 1% match in the 430 - 450 range with about 150 pieces in this range to choose from.

Second pass on the BC-327 yet to go. :drugs:

Well done on having the stamina to do so many measurements. However, I should point out that the meter you have seems to be measuring at just one current drain. That's fine and will give some level of matching, but the next level up in terms of matching would be to measure the Hfe at the current drain they'll see in your circuit. Better than that would be to also make a measurement above and a measurement below the operating current to get a crude measure of the shape of the curve. The ultimate in matching would be to measure the actual curve rather than just one point on the curve.

[mac72]

Well done on having the stamina to do so many measurements. However, I should point out that the meter you have seems to be measuring at just one current drain. That's fine and will give some level of matching, but the next level up in terms of matching would be to measure the Hfe at the current drain they'll see in your circuit. Better than that would be to also make a measurement above and a measurement below the operating current to get a crude measure of the shape of the curve. The ultimate in matching would be to measure the actual curve rather than just one point on the curve.
DCA75 Pro is capable of curve tracing as far as I know

[mac72]

Sure.
It might be a while though.
I'm not renowned for my blistering pace. ;) :hotrod:
Time is not an issue, after few months I managed to built PSU , at this pace I should be done just before I retire .
I have 500 of each transistors but haven't opened Mouser cardboard box yet, couple months since receiving .
Obviously I'd need find a way to compensate you for time you've spent testing them .
Keep in touch

[howlindawg]

Better than that would be to also make a measurement above and a measurement below the operating current to get a crude measure of the shape of the curve. The ultimate in matching would be to measure the actual curve rather than just one point on the curve.

I think I'll wait for the final selection before going to those lengths.
2000 curves might be a bit too much, even for my OCD. ;)

https://farm5.staticflickr.com/4414/36580258861_9f8e84571b_b.jpg (https://flic.kr/p/XJtowi)
Peak-337 (https://flic.kr/p/XJtowi) by Martin Fay (https://www.flickr.com/photos/howlindawg/), on Flickr




Will you also do a third pass ,this time using pliers and same room temperature, before they go on the board ? ;)


The second pass was already at a reasonably consistent room temp (mid twenties) and without touching so should be pretty close.



Where did you get them from? can you also give me supplier part number ,as I said before it is pot luck to get the right batch, and this time look like you got the right ones.

Mouser but it really is pot luck batch to batch.
I just got lucky that the outliers of my samples overlapped in the usable range.
327s are clustered between 390 and 450 - mostly around 400.
337s are clustered between 450 and 490 - mostly 460-480.
Fingers crossed I get matching 327s on the second pass.

Mouser No: 512-BC32740TA
Mfr. No: BC32740TA
Desc.: Bipolar Transistors - BJT PNP -45V -800mA HFE/630

Mouser No: 512-BC33740TA
Mfr. No: BC33740TA
Desc.: Bipolar Transistors - BJT NPN 45V 800mA HFE/630

I'd definitely be up for swaps within the usable ranges.


Time is not an issue, after few months I managed to built PSU , at this pace I should be done just before I retire .


I've had my psu built and tested years ago (literally!) but have been too unwell until recently to continue with the build.
Brain Tumors suck!

[Bksabath]

21197Hi Martin
same results here
The PNP are in range
The NPN are a pain in the ...
I have a batch from Rapid those are low from 300 to 350
batches from mouser are hi they start from 460 and go all the way up to 600

I have ordered today a further 1000 from Farnell instead of the tape (TA) I am going to try the BU there may be a bit more span
Same as you
I am using the DCA Pro on the first page (component testing) that seems to be a bit faster for the first pass
I will do the last pass before they go on the board whit the curve tracer function

[sq225917]

Howling, you've gotta do curves for the input stage. Trust me.

[sq225917]

Before you ask, 8mA

[howlindawg]

Howling, you've gotta do curves for the input stage. Trust me.

I do Simon and I will.
I should have all the 430-450 whittled out now so that's the easy part done then. ;)

Results not as good as expected.
150 337 in the 430-450 range
50 327 in the 430-450 range (hundreds in the 400-430 range!)

looks like I'll being ordering a few hundred more 337s then. :lol:

[RothwellAudio]

DCA75 Pro is capable of curve tracing as far as I know
I stand corrected, thank you. I that case, curves would probably be worth doing, at least after narrowing down the selection by single point measurements.

[Avinunca1]

I have a DCA55 and it looks like I will need to upgrade to a DCA Pro. I am wondering how people on here are matching gain to 1%
From http://www.peakelec.co.uk/resources/dca75.html
BJT hFE accuracy (hFE<2000) ±3% ±5 hFE
I assume the figures are relative rather than absolute.

[Bksabath]

211992120021201
Just got home
Postman dropped off another batch of 337 going to have a bite and a beer then start testing them
It has been a busy morning as I just came back from same place in Warrington :D

Off topic :
Quite impressive Andrew !!!!
Home demo soon ;)

[Bksabath]

I have a DCA55 and it looks like I will need to upgrade to a DCA Pro. I am wondering how people on here are matching gain to 1%
From http://www.peakelec.co.uk/resources/dca75.html
BJT hFE accuracy (hFE<2000) ±3% ±5 hFE
I assume the figures are relative rather than absolute.
I have built the R2 version using the old DCA
It was pretty good but not as good as Salas Simplistic for staging IMO
For the R3 I am going over the top so yes DCA pro is worth it and upgrade is must ,also consider that you be saving few thousand pounds worth of stupid batteries that the old DCA use to drink dry in minutes
On the third hand Build guide whit 10% matching work for most

[Bksabath]

I take it that's a "no" then.
The first mention of Paradise was on the MPP tread so it is going to be more than 3000 posts
Still there is very little explanation about how the circuit work
IMO quite easy to get loughed at in DIYaudio if one make mistakes

[RothwellAudio]

The first mention of Paradise was on the MPP tread so it is going to be more than 3000 posts
Still there is very little explanation about how the circuit work
IMO quite easy to get loughed at in DIYaudio if one make mistakes
Is there a circuit diagram somewhere?

[Bksabath]

All the information is on the Diyaudio treads

http://www.diyaudio.com/forums/group-buys/216891-paradise-phono-stage.html

I be happy to email you what I have if you can not get there

I can not post PDF here as attachments

[Bksabath]

I have just measured 50 or so random samples from farnel
Part number 2453782 BC337-40 batch H21 they start at 460 and go all the way up to 520
NO GOOD

[sq225917]

Avinunca,
The 3% is a spot measure, if you batch them all at once you get much better than that and you'll end up measuring the best parts a handful of times to dial in a match.

[RothwellAudio]

All the information is on the Diyaudio treads

http://www.diyaudio.com/forums/group-buys/216891-paradise-phono-stage.html

I be happy to email you what I have if you can not get there

I can not post PDF here as attachments

Thanks for the link. Unfortunately, I'm not a member of that forum and cannot download files from it. If you could let me have a copy of the circuit diagram it would be much appreciated.

[Avinunca1]

Bearing in mind the new price for a DCA Pro is north of £100 I have just bought a reconditioned one on EBay supplied by Peak for £78 inc postage. They had another three available. They may have some marks or scratches but of course are fully functional.
I think at that price you might hope to recoup most of your costs if you decided to sell after the matching.

[Bksabath]

Thanks for the link. Unfortunately, I'm not a member of that forum and cannot download files from it. If you could let me have a copy of the circuit diagram it would be much appreciated.
sent you email to the office

[RothwellAudio]

sent you email to the office

Received, thanks :)

[Avinunca1]

Bearing in mind the new price for a DCA Pro is north of £100 I have just bought a reconditioned one on EBay supplied by Peak for £78 inc postage. They had another three available. They may have some marks or scratches but of course are fully functional.
I think at that price you might hope to recoup most of your costs if you decided to sell after the matching.

My DCA Pro has arrived. Looks like new except very slight marks where the case halves were separated. I need to find my pc laptop now. Wish they would make the tester Mac compatible.

[Arkless Electronics]

Now whet yer really need is a Tektronix 576 :eyebrows:

[Bksabath]

Yup a 576 be nice but there is an alternative ... If one need to measure hi current output devices
The DCA Pro is really nice clever gadget only snag for me is that it does not recognise Jfet pins
It beats the old 75 as no more really expensive batteries that is enough to justify the upgrade

[Arkless Electronics]

Yup a 576 be nice but there is an alternative ... If one need to measure hi current output devices
The DCA Pro is really nice clever gadget only snag for me is that it does not recognise Jfet pins
It beats the old 75 as no more really expensive batteries that is enough to justify the upgrade

I make my own matching jigs here;)

[Bksabath]

Willing to share this time ?
If you show me yours I show you mine :eyebrows:

One may goggle up Zhoufang (is that right?) curve tracer

[Arkless Electronics]

Willing to share this time ?
If you show me yours I show you mine :eyebrows:

One may goggle up Zhoufang (is that right?) curve tracer

Not sure if that's aimed at me but the ones I use are not curve tracers, just quickly knocked up on vero board and measure at one current only. That being the current at which I will operate the device at in the intended circuit. Being a manufacturer and wanting to earn a living from this (:lol:) no I do not share anything really interesting/original/of use to rivals... as one would expect!

[Bksabath]

21219Well then I am not going to show you this one

For all the nice peoples that do this for passion or just to stay out of troubles :lol:

A constant temperature (50C) rig to measure up Moss Fets
Set up the pot (gate voltage ) to get a wanted current flowing through source and drain

A further Mossfet is used to warm up the plate and the Fan (CPU cooler)speed is regulated by a thermistor

[337alant]

My DCA75 pro also just arrived, very nice
Ill have to dig out my bags of transistors and start testing again, I seem to remember hfe of 415 was the match I got for my first build but it took a few thousand transistors to achieve that

Alan

[sq225917]

Rapidonline have the diotec 327 and 337 in stock. They have been the best parts I've found. That said, they are loose and not ammo tape. Soselectronics have them on tape but the 50euro MOQ and lack of any other interesting stock makes that a tall order. 5000 transistors anyone?

Alan did one of your mates just pick up a Paradise?

[Bksabath]

The last batch of 327 from Rapid where hi a pretty consistent batch whit not a single one under 440 mostly between 470 and 530 they have 16,000 ,if I remember, left all from the same batch

I have ordered same (5 different part numbers 327 and 337) from Soselectronic, they should be in in 4 days will let you know ...

[sq225917]

My order of Diotec 337-40 and 327-40 from Rapidonline arrived today, all on ammo tape. From a quick sample of 10 of each the 327 are 320 +/- 10 and the 337 are 330 +/- 10, all at room temp 22 degrees. I'd imagine there will be some outliers but I think it'll be easy to get a truly stellar match for HFe and Vbe from these.


These really are the last ones Im ever buying.... no really they are...honestly.

[337alant]

Thanks for the tip Si I have just ordered 500 of each :eyebrows::D

Alan

[Bksabath]

I 21298212992130021301have tested about 2 of the new batches Just the BC337-40BA from SOS
Most are on the on the low 380 or less but managed to get enoughfor 1 set of board’s with HFE around 400
When I say low it may well be usable provided you have a similarbatch of PNP
I had problem so far getting low NPN from Mouser RS Farnelland Digi key and still do not have more than a few on the 440Mark (some one is been going around and picking off my transistors :scratch:)
As I have quite a few PNP on the range of 400 I have beenselecting for this
I have mentioned before I am going over the top this timeconsidering that the recommended is +- 10 % any transistor from the lowest (360)to the Highest (440) would in theory match
As you see from the DCA pro graph I pick the transistors 2by 2 one for left and the other for the right Channel
IMO the most critical is the quad on the input the mirrorsmay not need matching as closely if there is no difference between channels
Am I missing something

[Bksabath]

I 21298212992130021301have tested about 2 of the new batches Just the BC337-40BA from SOS
Most are on the on the low 380 or less but managed to get enoughfor 1 set of board’s with HFE around 400
When I say low it may well be usable provided you have a similarbatch of PNP
I had problem so far getting low NPN from Mouser RS Farnelland Digi key
As I have quite a few PNP on the range of 400 I have beenselecting for this
I have mentioned before I am going over the top this timeconsidering that the recommended is +- 10 % any transistor from the lowest tothe Highest would in theory match
As you see from the DCA pro graph I pick the transistors 2by 2 one for left and the other for the right Channel
IMO the most critical is the quad on the input the mirrorsmay not need matching as closely if there is no difference between channels
Am I missing something

[sq225917]

What current are you matching at, looks high...

[Bksabath]

21302213032130421305
I am using the DCA Pro component test for the firstand second selection where second is made whit using pliers and constanttemperature (22C) as a rough guide for the first selection handling atransistors for 10 seconds give a higher HFE by about 15
The Graphs have Ib of 25 uA and just one trace for each at 12 V

As you know my knowledge of electronics is very limited so I would very much appreciate any guidance

I was going to ask you as well from the previous post are you going to use Hfeon the range of 330 being that the case I have quite a few hundred that spanfrom 350 to 370 as you can see on my first pass box for the NPN I have so farmanaged to find few between 380 and 440 the top row is made up of any above 480Left most and the remaining are all above 500 further to those in the box Ihave a couple more thousands all over 500 in bags due to limited space
I did much better whit the PNP as you can see in the last of the 4 pictures

[sq225917]

Sandro once I'm done with my current build ill send you a matched set, I tested a few today at 30 degrees, all 410-450. You'll get a great set.

[Bksabath]

Tank you Simon that would be great
From the previous post You mention that possibly I am using hi current while testing.
What settings and graph do you use ?
Also is it possible to use lower HFE than the ball park 400 and I may be wrong but effects of this in the current mirrors, from what I have been reading and maybe less than understood, should be limited what is your experience on this regard ?

[sq225917]

I've never used under 400.

You can set the curve trace range on the pc 1-14ua is ideal.

[mac72]

What's the best/latest transistor choice for Calvins buffer, manuals states :
Q3,Q6 - 2SA1381
Q1,Q4 - PF5102
Q2,Q5 - J107
thanks

[sq225917]

That's it. The j107 and pf5102 need to be matched though.

[mac72]

Thanks , where did you get 2SA1381 from , Mouser don't seem to stock these

[sq225917]

They don't. KSA1381 is a replacement part, or any of these with adjustments to these caps.

BD 140 (Suffix -16) ( C1,C2 = 180-220pF)
MJE/MJD 350 ( C1,C2 = 100-47pF)
2SB 649/A (Suffix C oder D)
2SA 1930 ( C1,C2 = 180-220pF)
MJE 253 ( C1,C2 = 47-22pF)

[mac72]

Thank you Simon

[mac72]

What's the best place to get matched pairs of 2SK170, please?

[sq225917]

http://stores.ebay.co.uk/Himmels-Rohren-Shop?_trksid=p2047675.l2563

There's a few others with legit parts but 99% 0n eBay are fake. This guys are legit and well matched, I've used him for years.

[sq225917]

Get the BL grade

[mac72]

Get the BL grade

Thanks , I will

[mac72]

Q107,207 what's better 2SK170 or J310 , going through DIY Audio thread again would be painfull, thanks

[sq225917]

J113 with 1k resistors replacing the 3k3. While your doing the shunt put a pair of 33uf 50v organic polymer caps on the input from the psu board, ie right on the -0+ on the pcb

[mac72]

J113 with 1k resistors replacing the 3k3. While your doing the shunt put a pair of 33uf 50v organic polymer caps on the input from the psu board, ie right on the -0+ on the pcb

Thanks Simon
I have spare quad of 2SK170 but I'll order J113 instead

[sq225917]

Yeh, save the 2sk170 for signal duties. They're wasted in the psu.

[Bksabath]

Yeh, save the 2sk170 for signal duties. They're wasted in the psu.

Yup 100%

[mac72]

One more q , 10uF inductor for Calvins why not to use fixed inductor fromt the Mouser or elsewhere ?

[sq225917]

You could use any air core you desired. I roll my own at 8uH, works fine, rock solid stability. They do need to be air cored though.

https://c1.staticflickr.com/5/4342/37234278052_53ce18bc37_b.jpg

[mac72]

Nice , what diam . of core and wire , no. of turns?

[sq225917]

Pm me your address.

[mac72]

PM on it's way ,
Many thanks

[mac72]

Simon , couple more questions , in the Calvins do you use same value R5 R6 R7 R11 (6R8 or 4R7) or as per schematic
Did you change R104 R204 from 10R to lower value due to increased current draw

[Bksabath]

21478
I have completed the boards
all BC 337 and BC327 are matched between the 2 channels and the groups whit Hfe from 397 to 404
EG the left and right channels have all 16 transistors matched at Hfe 401
I have a question in regards to the Calvin buffers just to make sure

Q1 =2SK170 8.2 mA
Q2 =PF4391
Q3=2SA1930

C1 = C2 =220 Pf

I have used those as they are what I have at present but I can wait and get others

What are the best ?
And quote from guide :

For the input transistors, also these can be used: 2SK170 (IDSS=8...8.5mA) / cascode: PN4391
BF244B / cascode: BF245C (C1,C2 = 22pF)
For the output transistors, a lot of choice is possible: BD 140 (Suffix -16) ( C1,C2 = 180-220pF)
MJE/MJD 350 ( C1,C2 = 100-47pF)
2SB 649/A (Suffix C oder D)
2SA 1930 ( C1,C2 = 180-220pF)
MJE 253 ( C1,C2 = 47-22pF)

Should C1 C2 be chosen for the input transistors as stated for the BF244 /BF245 = 22Pf
or should it be chosen for the output transistors ?

[Bksabath]

Simon , couple more questions , in the Calvins do you use same value R5 R6 R7 R11 (6R8 or 4R7) or as per schematic
Did you change R104 R204 from 10R to lower value due to increased current draw

Sorry Slav
You got me confused should not be R5 and R7 be adjusted to get the idle current on the output transistor Q3 and Q6 and the idle current is ? I can not find this on the Guide
and R6 and R11 are 2.2 Homes

[mac72]

I think Calvin on Diy thread mentioned something and came up with some simulations for same values of R5 R6 R7 &R11 but cannot find it now , I try to work my way through one more time
What power resistor in the current sources do you use 8R2 ?

[Bksabath]

I think Calvin on Diy thread mentioned something and came up with some simulations for same values of R5 R6 R7 &R11 but cannot find it now , I try to work my way through one more time
What power resistor in the current sources do you use 8R2 ?

You mean 8Ra and 8Rb 220 r and 2K2 pot ?

[mac72]

I meant 10ohm 3w resistor on Paradise board which need to be dropped in value to provide enough current for RIAA and Calvin, I think it should be 8.2 ohm, am I right?

[Bksabath]

Yes I have dropped that to 7.6 Homes and fitted same slightly larger heat sinks

[mac72]

Yes I have dropped that to 7.6 Homes and fitted same slightly larger heat sinks

Thanks Sandro , I must find Calvins simulations , is there a particular reason to use 2KSA170 ?

[Bksabath]

Thanks Sandro , I must find Calvins simulations , is there a particular reason to use 2KSA170 ?

If you mean 2SK170 not that I know off I have used them (reluctantly as they are precious things to pass on to my gran children's eventualy) because I have BF245 but not the BF244
same for other recommended pair

[sq225917]

I use the values from the notes, not the schematic and have always built with j107 and 5102 or whatever the match is. I use mje350 and 220pf caps. On the shunt I run 8.2 ohm 5 watt non inductive wire wound five watts and heat output is modest if you use a 300mA choke regulated power supply/ pre regulator and stick to 25.5- 26v under load into the shunt.

Slav I think I have a spare quad of those 8.2 ohm resistors. Ill include them with your coils.

[sq225917]

I use the values from the notes, not the schematic and have always built with j107 and 5102 or whatever the match is. I use mje350 and 220pf caps. On the shunt I run 8.2 ohm 5 watt non inductive wire wound five watts and heat output is modest if you use a 300mA choke regulated power supply/ pre regulator and stick to 25.5- 26v under load into the shunt.

Slav I think I have a spare quad of those 8.2 ohm resistors. Ill include them with your coils if I can find them.

Don't forget to lift that 100k before you plug them into your paradise. That 27r need only be 250mw, use a good metal film.

[mac72]

Simon , don't worry about 8R2 resistors I have some 3W in the cupboard, appreciate your help with inductors

[sq225917]

https://c1.staticflickr.com/5/4474/37345674622_077ea60c8a_b.jpg

Theyre in the box, use then if you want. the extra dissipation keeps things cool.

[mac72]

https://c1.staticflickr.com/5/4474/37345674622_077ea60c8a_b.jpg

Theyre in the box, use then if you want. the extra dissipation keeps things cool.

Very kind of you , thank you Simon

[sq225917]

For a change I've been doing some metalwork for myself.


https://c1.staticflickr.com/5/4498/36723437394_ab7fe26240_b.jpg


What's under all those holes?


https://c1.staticflickr.com/5/4464/36758747673_ef4b731246_b.jpg

Of course my soldering iron died today, so It's all on hold until a new iron arrives. Leaving me without a PSU for my Paradise. Time to turn on that dac....

[Marra]

Nice work as always Simon.

I see you recommend the addition of 33uf 50V Organic Polymer caps at the p/s input on the Paradise boards; from where do you obtain these as I cannot find them at either Farnell or RS.

[sq225917]

Farnell, wurth make em

[Marra]

Thanks Simon; found em.

[Bksabath]

Well got one Channel running
And just spent 3/4 hours testing the second one which is not working
All voltages are ok current throug the supply resistors is fine and I have made sure all transistor are the right way around and not mixed up the 337 whit the 327 about 12 times but no way I could find the blipping fault, then I just noticed that I was using the wrong channel plug from my passive RIAA :rfl::wanker:
Feel free to take the piss out of me ...

I may get it to play tomorrow and start fine tuning

[mac72]

Powered up one channel , I have fairly big voltage fluctuation +/- 80mV on the output ( with 10k on input) and 1.6V on pin 6 of the OPA .
I covered input stage transistors with little cardboard house , board sits in the box , all transistors matched within 2-3 Hfe , I guess I leave it powered for some time .

[mac72]

What's optimum current should be running through Calvin power transistors , I have 19mA only , shouldn't be close to 30mA?

[Bksabath]

Not sure there is no information on the guide best I can say is If you can have a look at the Paradise Builders tread on Diyaudio post 3053 (I can not and just have little info as I got banned for posing 5 or 6 X :eyebrows:)

I am going to have a play whit the buffers tomorrow (messed one up whit same heavy welding)
I am thinking about putting a square wave from the scope trough passive inverse RIAA and paradise then have a look at the picture Will this work?

[Bksabath]

On my build whit 2SK170 and PN4391 output transistors2SA1930
I have set r5 and r7 whit a pot in parallel to the 22Homesresistors
The current through the combination is given by
V across the resistor over the resistance I=V/R
I have the following r5=15R 483 mV 32.2 mA
R7=15R 497mV =33.12
Changing the resistors to 6.8Rgive me the following
R5=6.8 297 mV 43.6 mA r7=6.8 294 mV 43.2

Changing the setting on the offset pot R8 changes thecurrent through bot resistors
The V drop for R5 R7=6.8 R at 30 mA should be around 200 mV
I now I have 210 mv on R5 and 202 mV on R7
A small adjustment of R7 increase current on Q6 and reducethe current on Q3
I have now R5 =208.7 mV and R7 208.7 mV
Small tweak ¼ turn on R8 and now volts drops are 201 and 200mV
One thig I have not done so far is fit a trim pot on R2 andtry that
Obviously the gains and Ids of the transistors /Jfet has aplay in the results I got
Still much more should be said on the Calvin Build guide

[sq225917]

Sandro the original Calvin buffer thread has more info in it.

[sq225917]

http://www.diyaudio.com/forums/analog-line-level/226099-preamp-buffers-simple-idea-17.html

[mac72]

On my build whit 2SK170 and PN4391 output transistors2SA1930
I have set r5 and r7 whit a pot in parallel to the 22Homesresistors
The current through the combination is given by
V across the resistor over the resistance I=V/R
I have the following r5=15R 483 mV 32.2 mA
R7=15R 497mV =33.12
Changing the resistors to 6.8Rgive me the following
R5=6.8 297 mV 43.6 mA r7=6.8 294 mV 43.2

Changing the setting on the offset pot R8 changes thecurrent through bot resistors
The V drop for R5 R7=6.8 R at 30 mA should be around 200 mV
I now I have 210 mv on R5 and 202 mV on R7
A small adjustment of R7 increase current on Q6 and reducethe current on Q3
I have now R5 =208.7 mV and R7 208.7 mV
Small tweak ¼ turn on R8 and now volts drops are 201 and 200mV
One thig I have not done so far is fit a trim pot on R2 andtry that
Obviously the gains and Ids of the transistors /Jfet has aplay in the results I got
Still much more should be said on the Calvin Build guide


I like the idea of trimpots ,
how do you match bipolars and jfets , DCA55 is not much of a help , any quick to build circuit ?

[mac72]

I use the values from the notes, not the schematic and have always built with j107 and 5102 or whatever the match is. I use mje350 and 220pf caps. On the shunt I run 8.2 ohm 5 watt non inductive wire wound five watts and heat output is modest if you use a 300mA choke regulated power supply/ pre regulator and stick to 25.5- 26v under load into the shunt.

.
Simon
where did you get those MJE350 bipolars , for some reason I can't get 2SA1381 to work , what's the best way to match them? Same applies to J107 ?
You've mentioned on DIYAudio testing Calvins as standalone unit , do you power them from Paradise boards with disconnected In ,Out, Gnd , just shortened with 100k?

[sq225917]

yeh, just short the input with 100k. I'll be doing a Mouser order tonight/tomorrow would you like me to add a quad of MJE350's for you?

[mac72]

yeh, just short the input with 100k. I'll be doing a Mouser order tonight/tomorrow would you like me to add a quad of MJE350's for you?

It would be great , can you order 8 so there is something to choose from and if you don't mind 8 of J107's , I'd like to start fresh ( ordered matched Pf5102 earlier today) , PM me how much I owe you (I'll participate towards shipping charge if applies) .
How do you match these Simon?
Regards

[mac72]

Simon
What values of C1&C2 you use , Calvin manual says MJE/MJD 350 ( C1,C2 = 100-47pF)

[sq225917]

100pf

[mac72]

Just to make sure all buffer parts are as follow:
Q3,Q6 MJE350 PNP power transistor
Q1,Q4 PF5102 n-JFET, IDSS=6mA, VTH = -1V
Q2,Q5 J107 n-JFET, IDSS=100mA, VTH=-3V
R6,R11 2.2 Ohm
R5,R7 6.8 Ohm (on PCB as 22 Ohm)
R2 75 Ohm (on PCB as 51 Ohm)
R1,R12,R16 100 Ohm
R4,R10 1k Ohm (on PCB as 100 Ohm)
R13 1 MOhm
R14 27 Ohm
R3,R9 10 Ohm
R8a 220 Ohm
L1 10uH Inductor
R8b 2k Trimpot
C1,C2 100pF Foilcap 63V
C3 0.22uF Foilcap 63V

[mac72]

and one more thing , on the main reg board replaced J310 with J113 and resistors 3k3 with 1k , they seem to have same pinout but I remember swapping gate and source legs over , is it correct

[sq225917]

No leg swapping for J113, they just drop straight in.

Those parts are all correct.

[Bksabath]

I am really confused now VTH is same as Vgs ?

[mac72]

I think it's exactly the same thing , gate threshold volatge

[mac72]

Seems like I kind of sorted both Calvins , it took bit of tweaking of R5 R6 R7 R11 , now I run one set with all resistors @2R2 and the other one with all @6R8 , on one board which was running very low current it's input transistors biased very close to Idss , gate-source close to 0V , the other one I have no idea what causes the offset @30mV ,to increase regulation I soldered temporarily another 220R in series with 220R and I can set offset to 0V .
Once I get new matched PF5102 and all others I give it another go , hopefully it would work without too much teaking

[mac72]

Guys , how you've sorted grounding arrangement ,
I think something like mains earth to the case star ground , from there to paradise board ground and to the tonearm ground post ( C+R?)
Do you use DDRC ?

[sq225917]

Link to my Flickr acct by one of my previous pics, plenty if examples

[RothwellAudio]

Guys , how you've sorted grounding arrangement ,
I think something like mains earth to the case star ground , from there to paradise board ground and to the tonearm ground post ( C+R?)
Do you use DDRC ?

That sounds like a very bad idea.
What's DDRC?

[mac72]

DDCR is the way of decoupling mains ground from star/chassis with network diode-diode-cap-resistor ,

[RothwellAudio]

The chassis needs to be connected to mains earth for safety. That's always true if there's mains voltage inside. The chassis should not be decoupled from mains earth.
The circuit's 0V is another matter. That could be left completely unconnected to the chassis, but it's convenient to use the chassis for screening the circuit, in which case something like your "DDCR" would be useful, but it's there to decouple 0V from mains earth and the chassis, not to decouple mains earth from the chassis!

[mac72]

That's correct , please disregard decoupling mains ground bit

[sq225917]

https://c1.staticflickr.com/5/4483/36990352403_82e6d33c41_b.jpg
Sadly they're spoken for.

I run a four wire connection between the psu chassis and the amp chassis. 3 wires are +/0/- and the 4th (Grey) is the chassis ground which comes in from the amp chassis and meets the star earth in the PSU chassis. The star earth is also the PE. In the amp chassis all sockets are isolated from the chassis. Signal 0v does not join the chassis or mains/PE at any point. My tonearm body is earthed to the chassis gnd on one of the amp channels. So signal is floating from mains gnd, always seemed better, lower noise than running a DDRC as a lift.


https://c1.staticflickr.com/3/2916/33920896391_b2cd7059f2_b.jpg

I don't run the common mode choke 0v node back to the star earth, this only ever coupled noise.

[Bksabath]

Simon sorry
On the Calvins no 27 homes resistor under the inductance ?

[sq225917]

It's under the pcb

[337alant]

It's under the pcb

Like this

https://farm1.staticflickr.com/528/32189934201_e0c1e0cd06_c.jpg (https://flic.kr/p/R3vPXx)IMG_2910 (https://flic.kr/p/R3vPXx) by Alan Towell (https://www.flickr.com/photos/69508926@N05/), on Flickr

Alan

[sq225917]

I think he means on the Calvins, 27r.

I have caddock 10r in the psu, after the choke. You can just see them.

[sq225917]

....resistors after the choke in the psu...not before as shown in both of our psu's.

[sq225917]

I need to put a transformer order in with Farnell so I'll order 50 j107 as well. Should be here Tuesday I'll send them out with your PF5102 boys....