PDA

View Full Version : ESS ES9008/9018 Sabre dac



leo
02-10-2008, 09:23
I've had my eye on these chips for a while, only place I know that offered these for diy was twisted pair in the Buffalo dac modules, they have been pretty hard to get hold of usually selling out as soon as they are available
Anyway I managed to get one from the latest batch and it just arrived.
Be a bit busy to try it just yet though:doh:

http://i273.photobucket.com/albums/jj220/kingbusoms/001.jpg

Filterlab
02-10-2008, 10:08
Looks interesting, nicely laid out anyway. Get soldering man!

Mr. C
02-10-2008, 22:08
Yes these are a most acceptable piece of equipment, for the diy guys done properly this dac genuinely compete with a lot of quality dac's upto £5K.
A worth while project

fordgtlover
03-10-2008, 10:33
That's great Leo.

I'd love to read your listening impressions. For such a popular DAC and chip, there are very few reviews that I've been able to find.

leo
03-10-2008, 15:07
I've quickly lashed it up, good god:eek: this thing hints at huge potential! its already shouting feed me nice supplys:lol:

Nobody else played with one yet? I don't think the supplied output stage and supplys is going to do it justice tbh

Mr. C
04-10-2008, 15:44
Leo,

You maybe on to something, the true potential of this dac is good, it actually does produce performance that it can deliver, real nos dac killer lol!
Have fun

themystical
04-10-2008, 20:53
I've quickly lashed it up, good god:eek: this thing hints at huge potential! its already shouting feed me nice supplys:lol:

Nobody else played with one yet? I don't think the supplied output stage and supplys is going to do it justice tbh

So go on Leo....give it to us straight! Does it live up to its hype? How does it compare with your other DAC's?

Marco
04-10-2008, 21:03
If it betters his TDA1541 I'll be shocked and stunned ;)

NOS DACs rule, Tony! :ner:

Leo's the DAC man, though, so he'll tell it as it is...

Marco.

leo
04-10-2008, 22:41
My TDA1541S2 (to me) still beats it with the human voice and sounds more realistic in that department BUT the dynamics , detail and sheer depth of the Sabre is infront , instruments on some well recorded music sound incredible!
Also this dacs regulated psu's and output stage are not optimal IMHO, once I've found some nice low output impedance regs and a discrete differential I/V this should make things much more interesting

The flaw with 90% of other dacs to my ears is that they sound artificial , they excell in certain area's but just don't sound real enough in the vocals.
Lets see how this one goes after a bit of experimenting;)

Just need to find some nice regulation, any idea's?

Marco
05-10-2008, 07:00
Sounds really interesting, Leo. I totally agree with you about the "artificiality" of other DACs compared to the TDA1541, and also for me the lack of weight, depth and scale, and like you say, the ability to convey a natural tonality. Everything else sounds 'digitally compressed' compared to the more 'analogue' sound of TDA 1541s.

*BUT* once you've sorted out the regulation, if the Sabre matches or betters the 1541 in the areas mentioned above then you'll be on to something really exciting. It will mean that at last modern CDPs or off-board DACs will offer the type of performance CD was designed to produce 25 years ago! ;)

Marco.

themystical
05-10-2008, 07:31
My TDA1541S2 (to me) still beats it with the human voice and sounds more realistic in that department BUT the dynamics , detail and sheer depth of the Sabre is infront , instruments on some well recorded music sound incredible!
Also this dacs regulated psu's and output stage are not optimal IMHO, once I've found some nice low output impedance regs and a discrete differential I/V this should make things much more interesting

The flaw with 90% of other dacs to my ears is that they sound artificial , they excell in certain area's but just don't sound real enough in the vocals.
Lets see how this one goes after a bit of experimenting;)

Just need to find some nice regulation, any idea's?

Not sure about regulation, this one has a 3V supply as well I believe. A nice balanced valve output stage ala lampizator should help with the realism on vocals if everything else is almost perfect.
How does it compare with the new Wolfson 8741?
How much did you pay for this? I quite fancy playing with this one as well but the idea of queing up ready to press the buy button when Twisted Pear Audio releases the next batch is putting me off. There is no other way of getting hold of one these is there?

leo
05-10-2008, 10:56
Sounds really interesting, Leo. I totally agree with you about the "artificiality" of other DACs compared to the TDA1541, and also for me the lack of weight, depth and scale, and like you say, the ability to convey a natural tonality. Everything else sounds 'digitally compressed' compared to the more 'analogue' sound of TDA 1541s.

*BUT* once you've sorted out the regulation, if the Sabre matches or betters the 1541 in the areas mentioned above then you'll be on to something really exciting. It will mean that at last modern CDPs or off-board DACs will offer the type of performance CD was designed to produce 25 years ago! ;)

Marco.

Hi Marco,

The TDA will always be a reference for me, its the one all new units are compared against here;)

Once I've got the Sabre as best as I can and if it still falls behind the TDA in the vocals then the old classic will still remain as my favourite chip.
My goal is try as many units as possible both commercial and diy, going by a lot of things I've heard so far I totally understand why CD gets a bad name

Anyway I'll add updates as soon as their available:)

leo
05-10-2008, 11:19
Not sure about regulation, this one has a 3V supply as well I believe. A nice balanced valve output stage ala lampizator should help with the realism on vocals if everything else is almost perfect.
How does it compare with the new Wolfson 8741?
How much did you pay for this? I quite fancy playing with this one as well but the idea of queing up ready to press the buy button when Twisted Pear Audio releases the next batch is putting me off. There is no other way of getting hold of one these is there?

It uses 3.3v and 1,2v for the core, my plan is to separate the XO supply too and separate as much other parts as possible
The onboard regulation isn't too bad tbh but shares its voltage with various bits, theres LM4562 after the LDO feeding the chip, theres always an area for improvements though with this stuff

The Wolfson chips are very good, their not as good as the Sabre though IMO
The 8741 is as far as I know their best chip

I'm afraid TP is the only place I know that offers these as a kit, the other option is one of the eva boards from ESS but tbh it didn't offer me the easy options of adding my own regulation , I/V stages etc , the Evaluation board also costs a lot more than the TP kits which isn't bad if you just want a plug and play unit

I doubt I'll be using a valve stage with this dac, they work nice with say the voltage output chips but they do add colouration, its something I'd like to avoid with the Sabre
We'll see how things go:eyebrows:

Mike
05-10-2008, 18:10
I'm watching this one with great interest! :)

I've registered with Twisted Pear and I'm hoping to be able to grab one from the next batch. Seems you need to be bloody quick off the mark though, that last lot sold out in under four minutes!!! :smoking:

Marco
06-10-2008, 07:58
Hi Marco,

The TDA will always be a reference for me, its the one all new units are compared against here

Once I've got the Sabre as best as I can and if it still falls behind the TDA in the vocals then the old classic will still remain as my favourite chip.
My goal is try as many units as possible both commercial and diy, going by a lot of things I've heard so far I totally understand why CD gets a bad name

Anyway I'll add updates as soon as their available

Definitely keep me posted, Leo. My suspicion is that in the final analysis the "old classic" will still edge it ;)

For me, the best DAC will always be that which makes hi-fi sound like real music; not the other way round. This is why as far as I'm concerned 'specs' or 'latest new technology' often mean jack-shit other than to impress a technology freak. It is far from being a guarantor of realistic and believable musical performance from any component or piece of equipment.

Marco.

purite audio
06-10-2008, 08:51
Leo Hi, is the implementation for both chips exactly the same, ( is it posssible to have identical implementation for both chips ) ? Keith.

leo
06-10-2008, 11:07
Leo Hi, is the implementation for both chips exactly the same, ( is it posssible to have identical implementation for both chips ) ? Keith.

Hi Keith,

I'm not sure what you mean mate, you mean is it possible to have the Sabre implementation same as the TDA1541S2?
IF so unfortunately the answer is no, their circuit requirements are a lot different

Leo

themystical
16-10-2008, 06:33
Hi Leo
Now that you have had (hopefully) a longer listen and play with this DAC, whats your latest view on the sound quality aspects? I believe they might become available again at TPA next week.
Avinash

Paul Hynes
16-10-2008, 09:01
Hi Leo,

I have only just joined the forum and, as yet, have not worked my way through the forum threads due to lack of time.

Your assessment of the Sabre DAC has grabbed my interest. I am looking for a DAC that will help close the gap between the enjoyment I get with my analogue system and the lack of enjoyment I have generally found with digital sources. The closest so far with overall presentation has been the Altmann DAC which I have to say is very good musically.

As you already know I have recently upgraded u72's Altmann DAC by replacing the on-board surface mount regulators with wideband shunt regulators. This has taken the Altmann DAC to a higher level of performance and the improvements were not subtle. If you wish to try this with the Sabre I have some regulator modules that have been previously used in regulator trials. I am prepared to let you have these free of charge, so you can see how well the Sabre responds to improved regulation.

I can provide shunt regulators for the 3v3 and +-15 volt supplies. Unfortunately the error amplifier I use in the shunt circuits will not operate on 1.2 volt supplies. However it should operate at 1.2 volts in a wideband series regulator circuit that I use (I will have to build one to verify this before I can be certain of correct operation). The series regulator will need between 4 and 7 volts input to operate. I will need to know more about the current draw of the various loads the regulators will see, if you have such information, as this will enable me to chose the correct module size for the job.

Regards
Paul

leo
16-10-2008, 14:35
Hi Guys,

The more I listen to this dac the more I like it,
The thing I'm finding with the Sabre dac is the lack of distortion compared to a lot of other dacs I've tried, its just so clear and detailed without the nasties, the hardest thing is getting the best out of it without adding signature
Twisted pear have also announced a new discrete output stage so I'll be hopefully trying one of those later , it'll be a straightforward swap from the IVY board

Hi Paul,
Great to see you on here! I did read u72's thread with interest, the TDA1541, TDA1543 and TDA1545's are much liked chips, it seems your shunt regs brought out the best in the Altmann

First I'd like to say thank you for the generous offer, I'd love to try your regs and post the results
I've since tried a couple of spare ALW super regs supplying the IVY output stage set at +/-15v
The sound improved in some area's but wasn't as good in others, theres more detail but with the expense of adding a sharpness and a mechanical quality which is not something I wanted so I reverted back to the standard regs, the improvements the ALW's brought are now lost but so are the bits I didn't like:)
I have checked the regs output on the scope for any instability, unless theres anything further up my scope can't pick up it seems ok

The standard regulated psu modules for the Buffalo and IVY board are these (obviously IVY needs LM317/337 and Buffalo needs 2 x LM317) http://www.twistedpearaudio.com/images/power/lcdps_schematic.jpg , To start with I built the twisted pear version on vero from that schematic, a transformer with 9v secondarys was used for the dac supply and a 18v secondary one used for the IVY, caps are Panny FC's, diodes are UF4002.
The +/-15v supplies the IVY's op-amps direct which are a pair of differential OPA1632's , these go to both the balanced out sockets and also a LM4562 which converts the differential to line out, decoupling caps are 100uf panny FM's bypassed with 0.1uf films ( I may play with those later)
The Buffalo's regulated psu is set to output 2 x +6.5v ,one supplies the Buffalo's onboard 3.3v LDO reg and the other supplies the onboard 1.2v LDO reg + a LM4562

I have measured the current on the outputs of the 6.5v external regulated supplys
153.6mA's for the 3.3v LDO reg (VD)
61.3mA for the 1.2v LDO reg(VA)
35mA for the +/- 15v IVY (+v side measured)

It seems the 3.3v supplies the Sabres 3.3v VD and onboard 80mhz crystek XO, comparator for spdif and a pic chip

On the input socket for the 1.2v LDO is also a LM4562 with the outputs supplying the dacs left and right AVCC http://www.twistedpearaudio.com/docs/digital/buffalo_schematic.pdf the LM462 has 3.3v on its output so I guess the 1.2v LDO feeds the dacs core and the LM4562 supplies the dacs VA left and right section

I'm trying to find a more detailed spec sheet for the Sabre, theres not a lot out there at the moment:doh:

Cheers,
Leo

Paul Hynes
16-10-2008, 23:05
Hi Leo,

Thanks for the welcome. Pleased to help the project.

The 3v3 and +-15v shunt regs will be the same type I fitted to the Altmann.

You will need to increase the input voltage of the 3v3 reg from 6v5 to 8v5 to allow for the regulator current source dropout voltage. I will set the available current to 250ma to give a little headroom.

The +-15v regs will need to see an input voltage of 20 volts and I can set the available current to 100ma.

Can you let me know the VA ratings of your mains transformers so I can assess their current capability?

The 1v2 series reg (three terminal drop-in replacement) would run pretty hot (too close to ratings for comfort) in this position so I will have to devise a different solution here. I do have a high performance discrete component series reg circuit that will that will operate down to 1v2, but will need to update the circuit board layout before you can have one of these. This will probably take a week or two to get ready.

I will need to know your postal address which you can e-mail to me at paul@paulhynesdesign.com

Regards
Paul

leo
17-10-2008, 08:15
Hi Paul,

Well I hope you can stick around when you get time

Altering the pre regs is no problem

The transformers specs are

2 x 0-18v 2 x 2.22A
2 x 0-9v 2 x 555mA

Before the pre regs the voltage drop supplying the 3.3v (160mA ) LDO's is 8.9v (regulated down to 6.5v) and the voltage drop before the pre regs supplying the 1.2v 56mA LDO's is 11v
These voltages are measured after a 10R resistor on the end of CLRC filtering before the pre regs

If the transformers aren't suitable I can see what others I have, also I presume using pre-regs before your shunts are ok and won't effect their performance ?

No problem regarding the 1.2v, I'd have thought the 3.3v is as important as any seeing as though it supplys the XO, comparator and dacs VD

Mail sent:)

Cheers,
Leo

Marco
17-10-2008, 12:29
Hi Paul,

Welcome to AOS :)

I have upgraded your account to trade member (Manufacturer) status to reflect your business position. Would you like your website http://www.paulhynesdesign.com/ featured in our Links Box shown below?

http://theartofsound.net/forum/showthread.php?p=123#post123

Marco.

Paul Hynes
17-10-2008, 15:10
Hi Leo,

Your power supply will be fine powering the shunt regs.

Pre regulation is more likely to be of benefit as less interference will be heading for the post regulation.

Hi Marco,

Thanks for the welcome. I would be happy if you include my web address in the Links Box. I do not have a logo at this time.

Regards
Paul

Marco
17-10-2008, 15:28
No problem, Paul. I'll organise it later :)

Marco.

leo
17-10-2008, 16:43
Thanks Paul, I'll adjust the pre regs output when needed

Leo

leo
24-10-2008, 16:51
I've re-drawn out the schematic for part of this dac module to give a better idea, it seems one socket supplies the 3.3v LDO reg, the LDO reg then supplies the XO,comparator, the 1.2v LDO reg for the core and the non inverted input of a LM4562, the 3.3v outputs of the LM4562 supply the dacs VREF left and right

It seems the VA socket supplying the onboard LM4562 supply rail has quite an influence , I tried a simple zener,cap multiplier set for 7v output to feed the VA socket for the onboard LM4562, this sounded worse than the standard LM317 based reg.
I'd imagine the output impedance of the cap multiplier was too high

C8 1uf SMD cap is right next to the +v LM4562 pin

http://i273.photobucket.com/albums/jj220/kingbusoms/sabrepic.jpg

leo
24-10-2008, 16:57
The LM4562 is quite a decent op-amp in analogue circuits , its crap at driving capacitive loads so needs a resistor on the outputs and seems to require a very good supply to stop it from sounding nasty

The question is, is it the best type of op-amp to be used in the above circuit for a voltage supply?

Paul Hynes
25-10-2008, 10:58
Hi Leo,

I think that the LM4562 could be improved upon however, as this part is essentially providing a buffered 3v3 for the Sabre left and right reference pins, it may be worth trying a pair of mini shunt regs in place of the LM4562. I will send you two to try with the other regs later this week. I will fit bendy wire leads so you can manipulate them into place. The noise performance should be slightly better than the LM4562 but the more important parameters in this application, like slew rate, settling time and low output impedance over a much wider bandwidth will greatly improve on the LM4562 specs.

Regards
Paul

leo
25-10-2008, 13:30
Hi Paul,

Thanks mate, it seems the VA socket supplys the onboard LM4562 only, I was surprised how much effect that supply has.

Rather than remove the actual SMD LM4562 I was thinking it would be ok to remove R4 going to the non inverted inputs of the LM4562 and also not connecting any supply to the VA socket?
I could then run the shunts outputs to the 1uf SMD ceramics going to the Vref L & R

Cheers,
Leo

BTW, theres an update for the IVY input filter which Twisted pair posted, I tried it the other day and its slightly improved the sound mainly the upper midrange.
Heres a link to the updated circuit, obviously the 100k resistor across the differential op-amps is for the sim so those are not fitted to the actual circuit http://www.diyaudio.com/forums/attachment.php?s=&postid=1637932&stamp=1224629138

Paul Hynes
25-10-2008, 13:53
Hi Leo,

You will need to remove the LM4562 as I cannot predict what will happen to the internal circuitry of the LM4562 chip if you apply power to the output terminals with the non-inverting inputs floating.

Regards
Paul

leo
25-10-2008, 14:11
Hi Leo,

You will need to remove the LM4562 as I cannot predict what will happen to the internal circuitry of the LM4562 chip if you apply power to the output terminals with the non-inverting inputs floating.

Regards
Paul

Thats fine Paul thanks, think I'll do it the cut pins and remove the remaining bits on the pads technique, don't you just love ROHS solder:eyebrows:



Cheers,
Leo

Mike
25-10-2008, 15:05
don't you just love ROHS solder:eyebrows:

Try this: http://www.wbtusa.com/pages/0800.html

There are UK sellers on ebay who supply it. :)

Of course, it's no help if you are de-soldering stuff thats been assembled with ROHS stuff. :(

leo
25-10-2008, 15:18
Try this: http://www.wbtusa.com/pages/0800.html

There are UK sellers on ebay who supply it. :)

Of course, it's no help if you are de-soldering stuff thats been assembled with ROHS stuff. :(

Hi Mike,

The Sabre uses 90% SMD parts, the Solder used looks to be ROHS, previous experience tells me removing SMD parts soldered with lead free solder can be a pain in the arse depending on track quality, rather than the risk of any track lifting I'll cut the op-amp out up to its body and carefully remove each remaining leg on the pads with solder wick

I've no problem working with SMD parts, especially soldering with leaded or lead free solder:)

Cheers,
Leo

leo
25-10-2008, 15:37
The separate IVY output stage board will be very easy , the +/-v regs are separate and supply the input socket

We'll have to see best place to run the grounding wires, the onboard regs and the input socket are connected to the same top ground plane
Could maybe run them to unused VA socket GND or to the existing pads of the removed parts
http://i273.photobucket.com/albums/jj220/kingbusoms/buffalo_layout.jpg

leo
08-11-2008, 01:25
I've now changed the input comparator for the SPDIF (marked as IC2 above)
, the fitted one was a LMV7239, these are too slow for 192k so I've changed it to LMV7219 which is a faster part.
Surprising that SPDIF even from my modded Squeezebox3 sounded better with the faster comparator in the Buffalo

This thing is really starting to sound damn good!

Just a note of caution, the comparator is bloody tiny so should only be replaced if your confident working with very small SMD components

leo
11-12-2008, 12:27
Heres an update after two weeks of upgrading the dacs regs to Paul Hynes shunts and series

The ones fitted are

1 x Z1A-3v3 positive shunt regulator set to 250mA which supplies main 3.3v and onboard Crystek 80mhz XO
1 x PR3G-1v2 positive series regulator supplying Sabres core and VDD1.2v
2 x Z1703v3 50mA positive shunt regulator supplying the Sabres left and right AVCC (VREF)
1 x Z1A-15 150mA positive shunt regulator set to 150mA supplying the +15v of the IVY I/V stage
1 x Z1N-15 150mA negative shunt regulator set to 150mA supplying the -15v of the IVY I/V stage

http://i273.photobucket.com/albums/jj220/kingbusoms/032.jpg

Fitting the new regs for the IVY was quite easy as the outputs are wired straight to the IVY pcb input sockets.
The ones replacing the dac pcb's oboard LDO regs required desoldering the SMD LDO's and wiring the output of the new regs to the pcb pads
The AVCC (VREF) was more tricky as it required removing the onboard LM4562 SMD used as a voltage buffer, I wired the output of the Z1703v3 shunts across the local SMD decoupling caps which are right next to the dac pin, I tapped their input from the now removed LM4562 supply, I did manage to damage one of the onboard capacitors so had to replace it, I'd advise keeping these caps to help with any inductance from the wiring, the AVCC supplies are critical ! Just be very careful if working with SMD parts

http://i273.photobucket.com/albums/jj220/kingbusoms/mods1.jpg

After powering up and having a quick sound check revealed much more detail, it sounded a little light though so I resisted temptation to listen too much at this early stage so left it alone to burn in, a few hours later the sound was far better, bass had much more depth but most of all was the improvement in the midrange !!!
Over the next few days it continued to improve to a point I was struggling to pull my self away from it, focus,detail with extremely clean and clear highs, unlike a lot of digital sources I've heard ,this unit had a total lack of HF distortion and sibilance, no grain!
The biggest plus for me was now the midrange purity which I felt was originally lacking with the standard Sabre, its actually infront of my TDA1541 dac at the moment, I now need to do some more work on the TDA dac and see just how much further it can be pushed, I'll add any updates as soon as I do on the other thread to see if it can be improved to regain top rank

Conclusion is compared to standard three pin regulators, the Paul Hynes discrete regs have FAR better performance, their impedance is lower over a much wider bandwidth , quality of the regulation in any audio equipment is important and will have an influence on the sound, these regs are much more effective than relying on boutique capacitor changes IMO

My advice, when first fitting these regs you have to be patient , give them a few days/ a week running in time as I felt their performance greatly improves as they settle down

Big thanks to Paul for giving me the opportunity to evaluate these regs, they get the big thumbs up from me

Yomanze
11-12-2008, 12:51
The biggest plus for me was now the midrange purity which I felt was originally lacking with the standard Sabre, its actually infront of my TDA1541 dac at the moment, I now need to do some more work on the TDA dac and see just how much further it can be pushed, I'll add any updates as soon as I do on the other thread to see if it can be improved to regain top rank


I have three DEM reclocking boards arriving soon if you'd like to tinker with one?

translated link (http://translate.google.co.uk/translate?hl=en&sl=zh-TW&u=http://www.pureaudio.idv.tw/re-dataa.htm&sa=X&oi=translate&resnum=6&ct=result&prev=/search%3Fq%3DDEM%2Breclocking%26hl%3Den%26client%3 Dfirefox-a%26rls%3Dorg.mozilla:en-GB:official%26hs%3DaaQ).

leo
24-12-2008, 16:34
I have three DEM reclocking boards arriving soon if you'd like to tinker with one?

translated link (http://translate.google.co.uk/translate?hl=en&sl=zh-TW&u=http://www.pureaudio.idv.tw/re-dataa.htm&sa=X&oi=translate&resnum=6&ct=result&prev=/search%3Fq%3DDEM%2Breclocking%26hl%3Den%26client%3 Dfirefox-a%26rls%3Dorg.mozilla:en-GB:official%26hs%3DaaQ).

Sorry Yomanze, I totally missed this:o
I once tried DEM reclocking, the results was very promising but I had some HF noise ,probably caused by having it built on a piece of vero board or maybe the choice of chip.
If you have one spare I'd love to revisit the DEM reclocking, I think in this case a pcb would be better

Cheers,
Leo

Telstar
26-12-2008, 13:40
1 x Z1A-3v3 positive shunt regulator set to 250mA which supplies main 3.3v and onboard Crystek 80mhz XO

Hey Leo,

Did you replace the XO? The supplied one is of very low quality. With 20-50$ you should have already something v. good.

leo
26-12-2008, 15:20
Hey Leo,

Did you replace the XO? The supplied one is of very low quality. With 20-50$ you should have already something v. good.


The 80mhz crystek XO isn't bad, problem is finding a suitable SMD replacement type.
Using a through hole XO on a separate board sounded worse even when it was wired as close as possible to the dac, I have tried it;)

I've compared XO's from Tent and Trichord before when modding cdp's , using a high quality low noise regulator the different XO's only made a small difference , much smaller than the differences between regulation

Telstar
28-12-2008, 21:49
I dont have first hand experience, but replacing the XO in the buffalo was one of the top mods suggested by people in the huge thread over diyaudio boards.

Personally i think the psu is more critical.

leo
01-02-2009, 03:41
Just a few updates

I've changed the wiring to the regs for the AVCC slightly, I now have the input and ground wires twisted and running to VA and G socket

I also refitted IC3 (LM4562) and added small track cuts on the op-amps output, doing this saves having to desolder and remove the op-amp as it isolates its output from the regs, I also removed the SMD resistor supplying the op-amps input, the track can also be linked back up if needed

Since the mods and after plenty of run in this remains one of the best dacs I've had in my system, just need to find a case now :scratch:

leo
01-02-2009, 22:37
http://i273.photobucket.com/albums/jj220/kingbusoms/044.jpg

fault151
23-03-2009, 20:15
Id love to hear this compared to my stock version of the buffalo. Are there any audio meet ups or events going on any time soon?

leo
10-04-2009, 19:35
I'm tempted :eyebrows: problem is it'll be only a few days before temptation gets the better of me to start fiddling transfering the Paul Hynes regs over etc ;) http://www.twistedpearaudio.com/digital/buffalo.aspx

jkeny
21-04-2009, 01:49
I'm new here & just read through this thread - Great review of mods, Leo - the most comprehensive mods I've seen so far on this DAC.

Looks like you've nearly optimised the PSUs. Only area for improvement I can see is a separate very low noise 3.3V supply for the clock. At this level of refinement I believe it will make a noticeable difference.

You now have to try the Counterpoint discrete IV stage or a different IV stage!

rolls
21-04-2009, 01:59
There is a totally different approach:
no regulators
no mains
no IVY, passive I/V stage
seperate supply for digital, clock, analog left, analog right.
I have done it and will never return to transformers/regulators
regards
rolls

jkeny
21-04-2009, 02:05
Rolls, what do you mean no regulators? - you need regulated supplies for all the different Ps feeds.

I can understand your preference for transformer IV rather than active IV (especially op-amp ones). What trafo are you using?

NRG
21-04-2009, 06:40
Batteries.....?

rolls
21-04-2009, 10:11
Rolls, what do you mean no regulators? - you need regulated supplies for all the different Ps feeds.

I can understand your preference for transformer IV rather than active IV (especially op-amp ones). What trafo are you using?

Hello
I have already posted in the buffalo forum. probably you read there, but I think Leo has been trying hard to get more out of this chip. I agree that there is a lot more in it.
Yes I work with Batteries, Transformers from S&B, inferior to those from Audio Consulting and ...no regs! I have removed reg1 and the opamp. Just the 1.2 volt is in place at the moment, but will be removed later.

jkeny
21-04-2009, 10:51
Hello
I have already posted in the buffalo forum. probably you read there, but I think Leo has been trying hard to get more out of this chip. I agree that there is a lot more in it.
Yes I work with Batteries, Transformers from S&B, inferior to those from Audio Consulting and ...no regs! I have removed reg1 and the opamp. Just the 1.2 volt is in place at the moment, but will be removed later.

Hi Rolls., yes I read some of the Buffalo thread on the other forum & noticed your posts. Apologies if you've already answered these Qs over there!

I applaud you for going an IV route other than the IVY or Counterpoint. Are the Audio Consulting trafos expensive? I'm looking at a different IV stage also.

I don't understand your PS - you must need some sort of regulation (not 3 legged chip variety) on the batteries to get the accurate voltages required by the chip - no?

rolls
21-04-2009, 12:29
Hi Rolls., yes I read some of the Buffalo thread on the other forum & noticed your posts. Apologies if you've already answered these Qs over there!

I applaud you for going an IV route other than the IVY or Counterpoint. Are the Audio Consulting trafos expensive? I'm looking at a different IV stage also.

I don't understand your PS - you must need some sort of regulation (not 3 legged chip variety) on the batteries to get the accurate voltages required by the chip - no?

Hello Ikeni
you are the first who actually think about my battery supply. So you must know, that the "father" behing this is Mr. Serge Schmidlin, Audio Consulting, who designed the prototype of this I/V converter that beats my Silver winded S&B easily, even still in copper. Final version will be cryo silver winding, will be quite expensive. We were there with the (mains driven) buffalo for a visit when he asked to replace the power supply boards with batteries. Even with the regs still in place you listened through a window that had just been cleaned. I promised him not to tell anything about the batteries, so you have to make up your mind for youself, but if you think about...., think we have actually no regs between the batteries and the chip.
Look at the datasheet.
At the moment I am designing a new dac with the 32bit chip, mainly the control board with voltage detectors, relays and so on, because a battery dac should not be more inconvenient than a normal dac and security is important, because the batteries could deliver up to 100Amp ....
regards
Andre

jkeny
21-04-2009, 13:14
Hmm, intriguing! I can't find the ESS9008 datasheet other than the promotional pdf on the ESS website. Any pointers?

I always liked the idea of batteries and have 8*12V 7ah SLAs. I've used them singly, with some success, in the past but I believe the secret is to parallel them for very low internal impedance!

Keep us posted, please, on your progress with the 32 bit chip (ES9018 or ES9012?)

rolls
21-04-2009, 13:36
Batteries and ESS/Sabre, a happy marriage, think what they need..., definitely not 12V!
I have just bought a new buffalo for evaluation, but actually I am just interested in the clock, the chip and the control chip. Unfortunately I am not able to program it.
The most expensive part will be the transformer, but I am sure it is worth the money.
Look at Mr. Schmidlins page:
http://www.audio-consulting.ch/
Don't look at the pricelist!
At the moment i have an Accuphase DP-400 for evaluation, because I am interested to know how my buffalo performs against the leaders. This player costs a fortune outside Japan and nobody bothers, but if they should pay just less than 500$ for a new buffalo....

jkeny
21-04-2009, 14:36
I'm not against regulators & believe, like in everything else, there are good & bad ones & some might even be good in one situation but not good in another!

Shunt regulators are generally regarded as good sounding & I've no doubt Paul Hynes's ones are excellent!

I would consider using battery(s) with suitable regulators but I still think that nobody has yet addressed the major limiting factor in good sound ie reducing jitter. Forget about low jitter clocks - none of these will live up to spec unless the clock PS is a separate, very low noise supply. The clock is very sensitive to any modulation or noise on this PS

jkeny
21-04-2009, 15:40
Hi Rolls,
An interesting idea was just posted on DiyHiFi by wildmonkeysects in which he says that maybe the sweet spot is to use V out & ensure your AVCC PS is
a really, really clean, preferrably open loop Avcc. No overshoots, no close phase margins if using feedback, no cheezy ceramic caps, no deriving the reference from Vdd

And this
internally, there is probably a sweet spot for the external load where the Sabre's thermal modulation is minimal. Not convinced it would be all I or all V. Dunno yet. Dunno the internal layout.

Now I'm wondering what is the load that the sabre sees using your Audio Consulting transformer as opposed to the S&B? Might this partially explain the difference in sound between them?

Just some things to consider!

leo
21-04-2009, 16:30
Interesting chaps, I'm pleased to hear others are starting to post their experiences with this dac.
How about some pics of your work;)

AVCC voltages are so damn critical to the performance, The separate L&R Paul hynes regs really brought out the potentional in mine tbh.

Since then I have tried quite a few other regs now with varying performance, none yet have got close to Pauls imo

BTW I've been playing about with Counterpoint , one of the prototypes, its quite a nice circuit, it had a few stability problems but I think TP has sorted that out in the production units

Also I'll have the new Buffalo32s to compare soon:eyebrows:

jkeny
21-04-2009, 16:55
Leo,
Have you tried a separate 3.3V low noise supply for the XO? Would love to hear about this? What regs did you compare to, can you say?

Did you ever try as a Vout DAC as wildmokeysects said?

leo
21-04-2009, 18:05
I did try separating the XO and using a separate reg, regs tried here was the one posted on Guido Tents site for XO's, a Pinkfish Flea modified to output 3.3v, One of the older Audiocom super regs, various devices from Linear Technology.
I actually thought all these sounded worse than having it running from the main PH shunt reg.
Maybe the PH reg was far better than the other regs I tried even though it was providing two shared supplies, I'm not sure yet, I may have to try another PH reg separate feeding the XO and see how that compares


Regarding Buffalo as V out, I've only tried it using + and ground of the differential output with a coupling capacitor, its not ideal tbh this way but was the easiest thing to try at the time., I liked it with IVY better
.
The other way to convert the differential out to single ended would have been transformers, I don't like signal transformers though (various reasons) , before anybody chucks bricks at me its just a personal thing:)

jkeny
21-04-2009, 18:15
Hi Leo,
Yea, I'm surprised the Tent or Audiocomm PS, connected as close to the XO as possible, didn't show any advantage over the shared 3.3V PH shunt reg! Maybe PH can provide an explanation?

rolls
21-04-2009, 19:00
The trannies are the same, 1:10, the dac sees 10 Ohms, is more in current mode. I have listened in voltage mod for more than a month, but found current better.
There is no dc involved. The quality of the trannies is different. I have an Audio Consulting silver for my 300b, I know his quality, I should buy a Lundahl to see if there is a cheaper way, but I doubt.
Interestingly we made an experiment to verify thermal modulation/distorsion, quite a new topic: We cooled the ESS chip with a peltier for a short while and could hear the effect, but then I had to add another and very big battery and condensate could destroy the chip...
I agree with Wildmonkeysects and have a battery for left and right Avcc, buffered with Caps. I should even try without caps.

leo
21-04-2009, 19:06
How about some more details on the batteries Rolls? type? how you power this the stuff up as in switches?

rolls
21-04-2009, 19:20
How about some more details on the batteries Rolls? type? how you power this the stuff up as in switches?

I have said everything possible about the batteries in post 53. There is a control board with relays, 12 at the moment. 6 relays are for the Buffalo and the Wolfson receiver, the other six for charging. I can listen about 20hrs before charging. Things will improve with the final board. There must be an undervoltage shut off and more. I have yet to find out the current of the 1.2V supply, then I wonder if it is possible to run the receiver with just one battery, because the PLL needs just 1.7mA, if I can not hear a difference I can save a battery.
nice to meet you
regards
Andre

leo
21-04-2009, 19:23
Sorry old Rolls, I must have been asleep:o

Nice to meet you too;)

Regards,
Leo

jkeny
21-04-2009, 20:31
Final questions, Andre (for now anyway),
Did you notice much of a change in the sound while cooling the ESS chip?
How would you rate the various changes you have made to your DAC in terms of score out of 10 for
- change from standard PS to battery PS
- Transformer IV

You never tried other Vregs like Leo, did you?

rolls
21-04-2009, 21:40
Final questions, Andre (for now anyway),
Did you notice much of a change in the sound while cooling the ESS chip?
How would you rate the various changes you have made to your DAC in terms of score out of 10 for
- change from standard PS to battery PS
- Transformer IV

You never tried other Vregs like Leo, did you?

It is difficult to answer the way you like it, but I am going to tell you how the batdac evolved.
First was and still is my "mains" buffalo with separate outputs for IVYbal/unbal and direct. Next was
mains Buffalo with S&B transformer, with 10cents I/V resistors.
Mr. Schmidlin was impressed, but tried the batteries.
After the visit we were at work, he winded his transformer and I worked on my second buffalo. Was already a big improvement to the old dac, even if the regs were still there, I was not prepared to remove the chips yet...
Then I got the new transformers, and after a considerable burn in time, more than 20 hours, they were and are much better, even still in copper.
I removed the reg and the opamp and was very nervous..
This was another big step, and finally i replaced the i/v resistors with bulk metals, ooooohh (10 hours later!), I should have done this much earlier.
It is a pity, if you are used to listen to the "mains" with whatever regs you can not imagine that there is still a lot of sh... on it. I can even hear it now on my Accuphase DP-400, which is not a "grainy" player. Now I listen to music even in the afternoon, not just in the night.
I must confess that Mr. Schmidlins equipment was an eye/ear opener, looks very strange all these unused sockets in the wall...
And, by the way, the battery dac modul is just about 22cm X 12cm x 9cm, batteries included. The most important thing is the control board.
I thank you for your interest.
regards
Andre

jkeny
22-04-2009, 02:28
Thanks Andre,
Can you try to put numbers that signify marks out of 10 on:
- "mains" buffalo
- "mains" buffalo with Audio Consulting Transformer
- "battery" buffalo with Audio Consulting transformer
- with bulk metal foil IV resistor

It would help me to get some idea of the relative improvements in each step!

What change in sound did you notice when cooling the ESS DAC?

rolls
22-04-2009, 10:05
As I have told you, I can't tell it in numbers, please,
but is was always a clear step, cooling the chip was again like
a clearer window. With all the audiences, six up to now, I had always the mains buffalo
with me. Of course I think that better regs will give an improvement as well,
a battery not suited can destroy the sound completely, at first I tried a cheap lead acid, with the regs, but a guy in Germany told me recently, that he uses lead acids parallel, 3 at least.
When I compare with my "reference", the Accuphase, I am quite satisfied.
regards
Andre

jkeny
22-04-2009, 10:07
Thanks Andre,
Didn't mean to pressure you!

leo
24-04-2009, 12:46
I've just tried this new input filter recommended by Russ for the IVY http://www.twistedpearaudio.com/forum/default.aspx?g=posts&t=587
So far I'm liking this better than the original values, its certainly worth trying!

leo
16-05-2009, 12:59
;)

http://i273.photobucket.com/albums/jj220/kingbusoms/018.jpg

leo
17-05-2009, 19:13
Anybody had experience with Oscon SVP caps? these seem to be polymer based instead of electrolyte fluid
Theres quite a few on this new 32bit ES9018 based dac

Reason I ask is because they are on the anolgue decoupling too, the older Oscons wasn't really suited for analogue but they wasn't polymer
I've not a lot of experience with Polymer caps

leo
23-05-2009, 10:39
I've been running this newer ES9018 Sabre based dac for over a week now.
There was a few things which needed sorting on the grounds but apart from that its been running in very nice

Again psu and regulation are extremely important in a dac like this, the standard supplied regs in this particular design by Twisted Pear are based around separate LT1763 and LM49710/20 used as voltage buffers for the dac chips AVCC L & R
Theres no onboard regulation for the outputstage so I first tried well filtered LM317/337 set to +/-15v, I then borrowed the Paul Hynes +/-15v Z1A/Z1N shunt regulation , these gave an immediate improvement


Its true this newer Sabre improves on the older version in some area's, I've yet to try a few more tweaks to make comparing more fair but so far its pretty impressive:eyebrows:

Audiocom AV
23-05-2009, 20:57
Anybody had experience with Oscon SVP caps? these seem to be polymer based instead of electrolyte fluid
Theres quite a few on this new 32bit ES9018 based dac


Hi Leo

I have used the OSCON SVP capacitor for upgrade mods in several unit including the Benchmark DAC1, Squeezebox Transporter; results have been very good.

Best Wishes
Mark

PS: I sent you a PM on the TX2575.

leo
24-05-2009, 19:52
Thanks and thanks Mark:)

jkeny
25-05-2009, 18:18
Leo or anyone,
Any idea why the LT1763 V Regs are used here at IC3,4,5 below the clock - designed for battery use & so atrocious PSRR by the look of the datasheet - what are they supplying PS to, the chip, clock? I'm not sure Russ wants to answer this Q as posted on the Buffalo tweakers thread on DIYA

Any further tests with PH shunt regs?

rolls
25-05-2009, 18:49
Leo or anyone,
Any idea why the LT1763 V Regs are used here at IC3,4,5 below the clock - designed for battery use & so atrocious PSRR by the look of the datasheet - what are they supplying PS to, the chip, clock? I'm not sure Russ wants to answer this Q as posted on the Buffalo tweakers thread on DIYA

Any further tests with PH shunt regs?

I think they are chosen because of their "low" 20uV noise. Anagram chooses the same on their DAC EV board, you have a choice of shunts, e.g. Paul's, or of no regs....

regards Andre

jkeny
25-05-2009, 19:04
Hi Andre,
I can't fathom it - How does low generated noise represent an advantage compared to low PSRR (PSRR <50dB (1KHz - 20KHz) <40dB over 100KHz) which allows noise coming in from the PS, a greater source of noise, if you ask me? If this DAC was intended to run off batteries then I'd understand!

BTW, I asked over on DIYA about using protoboards with Sabre DAC but Russ put it down as sub-optimal with traces & gnd. I made the point that this would probably be more than made up for by using better & multiple PS & clock. (I'm not even sure how sub-optimal it would be)

leo
26-05-2009, 01:54
Leo or anyone,
Any idea why the LT1763 V Regs are used here at IC3,4,5 below the clock - designed for battery use & so atrocious PSRR by the look of the datasheet - what are they supplying PS to, the chip, clock? I'm not sure Russ wants to answer this Q as posted on the Buffalo tweakers thread on DIYA

Any further tests with PH shunt regs?

As a guess (I've not closely traced the circuit yet so could be wrong) LT1763 regs provide the 1.2v core, XO and dacs 3.3v , I'd imagine the 3.3v is then split off through CRC filtering + op-amp voltage buffer for the AVCC L&R.
My next plan is to feed the LT1763's from a PH shunt reg set to 5v, this should clean a lot of s**t up before it gets to those regs.
removing the LT1763 altogether and installing separate PH regs would probably be the best

This Buffalo32s pcb allows track cuts so external separate regs for the AVCC L & R can be used , a pair of PH regs will be used to do this duty as I did with the original Buffalo dac

The PH regs will certainly be staying for the IVY section, they are vastly superior to any other regulator combo I've tried
Also none of the LM317/337 based regs seemed happy with the SVP caps seen on the output , PH regs have no such problem with very low ESR caps

leo
26-05-2009, 02:01
I think they are chosen because of their "low" 20uV noise. Anagram chooses the same on their DAC EV board, you have a choice of shunts, e.g. Paul's, or of no regs....

regards Andre

I'd love to try bats at some stage , for now I'm very happy with the results from the shunt regs

BTW please keep us updated Andre, its nice to see the alternatives to poplular regulation types

leo
26-05-2009, 20:22
Bloody hell! this thing is starting to sound real damn good:eek:

Dan Giovanni
27-05-2009, 23:05
Hello Leo,

I'd be interested in hearing about your experiences with other regulators. Also, do you have any idea how much of the improvement of the new Buffalo is down to the 32 bit chip?

Dan

leo
28-05-2009, 02:33
Hi Dan,

A range of the usual three pin types LM317/337, LT317/337 etc
ALW super regs, PFM flea, Audiocom super reg, Teddy reg , Burson discrete, Paul Hynes series and shunt

I tried most different options with the first Buffalo

TBH My hotrodded Buffalo gave the Buffalo32s using external lm317/337 regulated based supplies a very hard time in a lot of area's although to be fair the Buffalo32s is better than what the standard Buffalo was (going by memory)

The design of the Buffalo32s is certainly better than the old version, XO, regulation , passive components etc
Its hard to say how much is down to the ES9018 regarding improvement, ideally they would have to compared using the same board

Audiocom AV
28-05-2009, 08:21
Hi Dan,

A range of the usual three pin types LM317/337, LT317/337 etc
ALW super regs, PFM flea, Audiocom super reg, Teddy reg , Burson discrete, Paul Hynes series and shunt


Hi Leo

Which version of the Audiocom Super-regulator did you try, 1, 2, or 3?

The last version 3 is totally different to version 2, which in turn was different to version 1.

leo
28-05-2009, 10:34
Hi Leo

Which version of the Audiocom Super-regulator did you try, 1, 2, or 3?

The last version 3 is totally different to version 2, which in turn was different to version 1.

Hi Mark,

It was version 1(I just noticed your now on version 3)

Leo

Audiocom AV
28-05-2009, 11:21
Hi Mark,

It was version 1(I just noticed your now on version 3)

Leo

Hi Leo

OK.

Well version 1 was a stab in the right direction but version 3 is superior in every respect. As well as using a very low noise reference and gain amplifier, the reference is post filtered and the output bootstrapped. The clever start up means that a 5V reg will operate from 5.5V.

leo
28-05-2009, 21:36
Hi Leo

OK.

Well version 1 was a stab in the right direction but version 3 is superior in every respect. As well as using a very low noise reference and gain amplifier, the reference is post filtered and the output bootstrapped. The clever start up means that a 5V reg will operate from 5.5V.


Thanks for the info Mark, another one to try out sometime:)
Being able to operate from such a small voltage drop is quite handy indeed

Regards,
Leo

jkeny
28-05-2009, 22:48
A range of the usual three pin types LM317/337, LT317/337 etc ALW super regs, PFM flea, Audiocom super reg, Teddy reg , Burson discrete, Paul Hynes series and shunt
Wow, that's a fair selection of regs with a lot of well regarded ones in there - just missing the battery option as per Rolls


TBH My hotrodded Buffalo gave the Buffalo32s using external lm317/337 regulated based supplies a very hard time in a lot of area's although to be fair the Buffalo32s is better than what the standard Buffalo was (going by memory) Just shows how sensitive these digital circuits are to PS quality!


The design of the Buffalo32s is certainly better than the old version, XO, regulation , passive components etc
Its hard to say how much is down to the ES9018 regarding improvement, ideally they would have to compared using the same board
Dan (Spartacus) is doing just that - replacing a ES9008 with ES9018 on a mark 1 Buffalo - his impressions should be interesting!

Dan Giovanni
29-05-2009, 13:30
Hi Dan,

A range of the usual three pin types LM317/337, LT317/337 etc
ALW super regs, PFM flea, Audiocom super reg, Teddy reg , Burson discrete, Paul Hynes series and shunt

I tried most different options with the first Buffalo

TBH My hotrodded Buffalo gave the Buffalo32s using external lm317/337 regulated based supplies a very hard time in a lot of area's although to be fair the Buffalo32s is better than what the standard Buffalo was (going by memory)

The design of the Buffalo32s is certainly better than the old version, XO, regulation , passive components etc
Its hard to say how much is down to the ES9018 regarding improvement, ideally they would have to compared using the same board

Thanks for the reply Leo. Like you I've found the Buffalo to be very responsive to power supply improvements. For the analogue side, I'm currently using a discrete regulator by Greg Ball, pre-regulated by something of my own design. I'm interested in the PH shunts, but am pretty satisfied with what I have at the moment. Perhaps later when I'm bored. Haven't done much to the digital supplies (got rid of those LM317s) ....not sure how much there is to be gained there.

As JKeny says, I've got a ESS9018, and just need to send my DAC off to get the swap done by someone with the tools. The trouble is having to put up with on-board sound via my PC while the Buffalo is away......:( will report in when I finally get off my rear.

leo
29-05-2009, 14:13
Good idea guys, the original board for comparing Sabre chips would be ideal.
The ES9018 is certainly going to sound at least a little different, even the output test signals measure differently with these chips, the ES9018 looks better higher up in the frequencies

leo
30-05-2009, 15:46
Thanks for the reply Leo. Like you I've found the Buffalo to be very responsive to power supply improvements. For the analogue side, I'm currently using a discrete regulator by Greg Ball, pre-regulated by something of my own design. I'm interested in the PH shunts, but am pretty satisfied with what I have at the moment. Perhaps later when I'm bored. Haven't done much to the digital supplies (got rid of those LM317s) ....not sure how much there is to be gained there.

As JKeny says, I've got a ESS9018, and just need to send my DAC off to get the swap done by someone with the tools. The trouble is having to put up with on-board sound via my PC while the Buffalo is away......:( will report in when I finally get off my rear.

No probs, its nice to share experiences and idea's with this stuff

Gregs regulators look quite compact, any idea what they are as in are they series types? I don't think they are shunts
Is the output impedance low? you got any nice pics

leo
30-05-2009, 15:54
BTW Some albums really bring this dac to life, I was listening to Mary Black collection the other night and it sounded so real it freaked me out :P

I tried comparing the analogue output from Squeezebox , wouldn't take a golden ear to hear the difference , the silent spaces around instruments etc turns into a layer of noise with the squeezebox, cymbals go splashy , it just sounds messy

Dan Giovanni
31-05-2009, 12:42
No probs, its nice to share experiences and idea's with this stuff

Gregs regulators look quite compact, any idea what they are as in are they series types? I don't think they are shunts
Is the output impedance low? you got any nice pics

They are closed loop series regulators. Line rejection for the latest version is around 100db wideband. I don't have any figures for output Z, but as they have plenty of loop gain it will be pretty low. I'm not sure at what point low Z stops becoming meaningful .... at some point the resistance of the tracks will dominate.

Pics here ....
http://www.ska-audio.com/Forum/YaBB.pl?num=1221704470

The more I think about it, the more I'm tempted to give the PH regs a go sometime. IT's only money after all ....

jkeny
31-05-2009, 13:54
Dan, Leo,
I'm sure you've seen this low voltage shunt reg which is just nearing final design & has been built & tested by a number of people who report it sounding good: http://www.diyaudio.com/forums/showthread.php?s=&postid=1842017#post1842017

It would make an interesting comparison with the PH shunt & SKA series regs

Dan Giovanni
01-06-2009, 05:02
Dan, Leo,
I'm sure you've seen this low voltage shunt reg which is just nearing final design & has been built & tested by a number of people who report it sounding good: http://www.diyaudio.com/forums/showthread.php?s=&postid=1842017#post1842017

It would make an interesting comparison with the PH shunt & SKA series regs

Looks a good reg. Would have to swap the zener for a low noise reference. Vout = Vref + 0.6V, so we'd need a reference of 2.7V. 2.5V would probably be OK, alternatively it might be worth looking at different LEDs. I'd also want to use an RC filter with the reference to get the noise right down.

James G
06-06-2009, 10:03
Hi Leo. I am wondering how you're getting along with those Oscon SVP caps. I'm thinking to mod one of my DACs with a few SEPC types. Any special considerations I need to take when using them?

leo
06-06-2009, 10:22
Hi Leo. I am wondering how you're getting along with those Oscon SVP caps. I'm thinking to mod one of my DACs with a few SEPC types. Any special considerations I need to take when using them?


Their currently sounding very good but I'm still using Paul Hynes shunt regs

As long as you don't get instability with the SEP's you should be ok, the ear is a good tool , if you hear things like singing in a barrel effect, harsh etc thats a good clue its not stable

One thing you can try if you experience problems, stick a low value resistor in series with the cap, this can help instability of low ESR caps

James G
07-06-2009, 03:53
Thanks Leo. I'll listen out for that.
The designer was nice enough to email me a photo marking suggestions for replacements.

leo
13-06-2009, 17:14
I'm getting PM's from people and can't seem to reply to them, they look to be new users, if I don't PM anybody back it means I can't and not ignoring you:)
I think all new users have to pop into the welcome section just to say hello etc, is this right Marco?

Cheers,
Leo

jkeny
06-07-2009, 01:47
Just thought you guys might be interested in my Tweakers Sabre DAC - basically an ES9018 soldered (by hand) to a prototype board - total cost $70 http://www.diyaudio.com/forums/showthread.php?postid=1870036#post1870036

leo
06-07-2009, 20:37
Interesting, keep us updated

Those proto boards are handy for experimenting, the copper pads should take more soldering and desoldering than standard pcb tracks

jkeny
06-07-2009, 21:05
Will do, Leo, as you can see there are those who say it's doomed, "doomed I tell you" :)

Dan Giovanni
21-07-2009, 22:01
Hi John,

how are you getting on with the DAC? Still fired up to prove the doubters wrong i hope :)

I've gotten my ESS9018 working by the way - the Sabre chip needs heating up in order for it to lock on, so still a dodgy joint in there somewhere. Sounds great of course, but impossible to make a meaningful comparison with the older 24 bit Sabre.

Dan

jkeny
21-07-2009, 23:52
Hi Dan,
Good to hear that you have the problem figured out - only issue is how to resolve it - did you solder this yourself, by hand or send it out? Pity a meaningful comparison isn't possible between the 32 & 24 bit DAC on the same board - is this just because you now don't have an original buffalo board to compare it to? This would be an ideal opportunity to evaluate the sonic character of just the DACs.

Actually, I wa thinking of you the other day in regard to my experimental Sabre board and wondered if it tuned out to be a disaster would Dan be able to desolder it from the board with his friends hot air station?

This is not a problem on my protoboard as I did a continuity check on all pins when I finished soldering the DAC.

Yes, still working on it but I'm doing a TAS1020B (48LQFP) USB receiver first on protoboard to try out some ideas and avoid mistakes on an expensive Sabre DAC. So far it has proven easy to put 0.1uF C0G PS decoupling caps as close to the PS/Gnd pins as I want (within a couple of milimeters - the limit being the soldering iron tip width) - I'll post some pics if there's interest.

Dan Giovanni
22-07-2009, 22:19
Glad you are making progress John. If you run into trouble I can give you the contact details of the guy who did the work for me. Looking forward to seeing the pics and any comments from your doubters :)

As for mine, in order to do a fair comparison I'd have to undo all the mods I've done, and I really can't be arsed. My feeling is that there is little difference between them - certainly nothing large.

leo
30-07-2009, 22:25
Wanted to see how this dac sounded driving headphones so I wired up a jack socket and plugged in the cans, hmm pretty good.
So lets lets try cutting the headphone cable and adding a male and female 4pin XLR to allow the option of balanced seeing as though the Buffalo's offer differential output based around OPA1632's

I don't have a Volumite yet for controlling output or a balanced attenuator so I had to do it the compromised way by adjusting the SPDIF to suit

Anyway initial impressions was :eek: really took me by surprise!
Think I'd like the option of balanced inputs with the next diy speaker amp I try

jkeny
31-07-2009, 00:45
Leo, do you reckon the balanced O/P sounds better than the single-ended IVY stage?
A number of users have maintained that a transformer I/V is the best sounding O/P stage for the Sabre!

jkeny
31-07-2009, 01:28
Guys,
Just nearly finished the TAS1020B USB streaming controller on protoboard & thought I'd post some pics. I'm using it as a test bed for ideas/techniques before I tackle the Sabre DAC protoboard.

I have two ground planes on this - the bottom of the protoboard is a solid analogue ground plane & the bottom of the matrix board is a copper foil filled digital ground plane. There is also a small copper clock ground plane under the crystal & PLL components. All gnd planes are directly connected under the chip although the TAS1020B datasheet recommends to use a ferrite bead to join the clock gnd plane to the analogue ground plane.

Small X5R 0.1uF caps are placed close to the chip but could go closer. SEPC cap with bypass on analog crystal supply.

Any comments/improvements/mistakes?

leo
31-07-2009, 15:18
Leo, do you reckon the balanced O/P sounds better than the single-ended IVY stage?
A number of users have maintained that a transformer I/V is the best sounding O/P stage for the Sabre!

With headphones,DEFINITELY!
With the speaker amps, well only amp I currently have giving the option of balanced in is the UCD's , I'd say balanced is best with that amp, it just sounds a little more open and effortless

As noted before I've never been a great fan of signal transformers, I don't discuss much why because it usually causes arguments:lol: if I do get chance to try one with the Sabre and I'm proven wrong I'd be sure to post it on here;)

leo
31-07-2009, 15:19
BTW, anybody know anywhere we can download the ES9018 datasheet? theres a lot of things I'd like to look at , I know the ES9008 is readily available

leo
31-07-2009, 15:22
Guys,
Just nearly finished the TAS1020B USB streaming controller on protoboard & thought I'd post some pics. I'm using it as a test bed for ideas/techniques before I tackle the Sabre DAC protoboard.

I have two ground planes on this - the bottom of the protoboard is a solid analogue ground plane & the bottom of the matrix board is a copper foil filled digital ground plane. There is also a small copper clock ground plane under the crystal & PLL components. All gnd planes are directly connected under the chip although the TAS1020B datasheet recommends to use a ferrite bead to join the clock gnd plane to the analogue ground plane.

Small X5R 0.1uF caps are placed close to the chip but could go closer. SEPC cap with bypass on analog crystal supply.

Any comments/improvements/mistakes?

Your doing a good job with what you have to work with.
Look forward to seeing updates on this one;)

leo
20-09-2009, 12:23
I've been a little busy lately so the diy and fiddling has been slow, I'll update as soon as I get chance

The first Sabre dac really improved with the Paul Hynes regs replacing the standard onboard 3pin types so it seemed obvious to try some with the newer 32bit Sabre based dac

I've only just fitted these so they'll need time to run in, tidy up etc
The dac board is sat on the usual chopping board for evaluation purpose:)

First with PR3G2 supplying VD
This is a handy little adjustable supply, the one I have was set to have a range of up to 12v, I'll be trying this as a supply for the modded 7520 too:eyebrows:

http://i273.photobucket.com/albums/jj220/kingbusoms/011.jpg

Discrete regs before fitting

http://i273.photobucket.com/albums/jj220/kingbusoms/012-1.jpg

Next the onboard LT1763 output pin lifted, 1 x Z1703v3 shunt supplying the dac chip VD, 1 x Z1703v3 shunt supplying the XO and the S17LN-1v2 series supplying the 1v2 dacs core voltage

http://i273.photobucket.com/albums/jj220/kingbusoms/PHregs.jpg

15v +/- shunts for the output stage (can also use a pair for both channels
http://i273.photobucket.com/albums/jj220/kingbusoms/008-4.jpg

leo
21-09-2009, 14:37
After some run in this thing sounds so damn good! :eek:
This really proves how important good regulation is even for these newer dacs

A single pair of Z1A/Z1N +/-15v shunts can be used to supply left and right or can be run using two pairs in mono
I'll compare at some stage to see how much difference there is

Only downside is I need a new case now, it needs to be a little higher :lol:

http://i273.photobucket.com/albums/jj220/kingbusoms/PHregs1.jpg

Tony Moore
21-09-2009, 15:43
Awwww, c'mon, stop teasing us! What does it _really_ sound like? ;) How's it compare with the AYA? :eyebrows:

Good job with the psu!

Cheers,
Tony

leo
21-09-2009, 17:22
Thanks Tony:)

I like and appreciate the things this dac now does, it sounds a lot more real than anything else I've heard so far.
The variation in recordings can be pretty major now, I need to sort a better source capable of hi-res I think (any idea's from anybody would be appreciated)

I'll need to leave it running in longer before commenting too much yet, theres a few other things I want to try too

If you want the best from this dac though it'll never be a cheap project, theres a few things in the design of this actual chip makes it much more sensitive to supply quality than a lot of other dacs

Tony Moore
21-09-2009, 18:02
Hi Leo,

How about the Squeezebox Touch when it comes out? It seems to suggest support for 24/96khz encoding...

It's a pity they don't have a high res capable receiver though. Something with the bare bones for a digital source, I2S output with multiple modes, upgradable power supplies, etc. I guess it's not high volume market enough for them now that they're Logitech?

Cheers,
Tony

leo
21-09-2009, 18:22
Hmm, I'll have a look at the Touch, ta Tone! 24/176 capable would be nice to have though

I tyhink your spot on with what you say about Logitech , its about time some place did something suitable for us fiddlers:eyebrows:

leo
26-09-2009, 09:27
The TDA1541 dac will now need some more work :doh: not sure how much more can be squeezed out of it

Audiocom AV
26-09-2009, 10:13
The TDA1541 dac will now need some more work :doh: not sure how much more can be squeezed out of it

Leo

Does this mean the mighty TDA1541 is going to be retired?

leo
26-09-2009, 10:22
Leo

Does this mean the mighty TDA1541 is going to be retired?

Not really retired but knocked off the top spot

Tony Moore
26-09-2009, 10:51
:violin:

leo
26-09-2009, 12:34
:violin:

:lolsign:
More flipping fiddling to try yet:mental:

leo
20-10-2009, 16:45
The voltage buffer IC's on the Buffalo32s usually run pretty warm so the latest mod was to add a resistor on the output to the positive rail, the resistor acts as a CCS so it gives the op-amp an easier time
It works well, the chip runs cooler , the zener is a little higher than the output of the voltage on the buffers out but lower than the chips max rating, its mainly just a safety thing

http://i273.photobucket.com/albums/jj220/kingbusoms/008-5.jpg

jkeny
25-10-2009, 01:20
Leo,
I would highly recommend a Musiland 01USD USB to SPDIF converter - will do 24/192 asynchronous. It sounds great stock but if you're in the mood will respond well to some modding - external PS (great mod) & I'll be trying a change of the 24MHz crystal soon.

I have modded a 01US which contains an internal PCM1793 DAC - details here: http://www.diyaudio.com/forums/digital-source/153191-m2tech-hiface-usb-spdif-24-192khz-asynch-4.html#post1955840

Oh, BTW, I haven't visited this forum since summer & only now picked up your PM to me - reply sent (if it's not too late) :)

leo
25-10-2009, 08:21
Leo,
I would highly recommend a Musiland 01USD USB to SPDIF converter - will do 24/192 asynchronous. It sounds great stock but if you're in the mood will respond well to some modding - external PS (great mod) & I'll be trying a change of the 24MHz crystal soon.

I have modded a 01US which contains an internal PCM1793 DAC - details here: http://www.diyaudio.com/forums/digital-source/153191-m2tech-hiface-usb-spdif-24-192khz-asynch-4.html#post1955840

Oh, BTW, I haven't visited this forum since summer & only now picked up your PM to me - reply sent (if it's not too late) :)

I've been looking at the Musilands ta, think I'll give one a try .
Its a shame the US version doesn't have co-axial out so I think I'd try the USD version and try some mods

I had an email this morning saying you tried to send me a PM but my inbox was full, anyway I cleared a load out again now:)

jkeny
25-10-2009, 13:20
There are some great options around now for high res USB to SPDIF conversion:

There is a comparison of 4 different USB to SPDIF units in which rates them in the following order of preference:
- M2Tech Hiface
- Teralink -x
- Musiland 01USD
- EMU 0404 USB

The M2Tech uses 2 different low jitter clocks and some jitter test have been performed here: http://www.head-fi.org/forums/f46/usb-24-192khz-m2tech-hiface-446375/index4.html#post6063226

Hope my PM got through?

leo
25-10-2009, 17:33
Hmm, so we now have a few interesting looking options, thanks for making us aware of those, time for some reading I guess;)

Afraid PM didn't get through although it should be in your sent folder so could always forward over:)

jkeny
25-10-2009, 18:30
I resent the PM. Leo!

Marco
26-10-2009, 18:50
Explain please.

{Edit: Sorry, I see that you meant 're-sent' rather than resent - I was a bit worried for a moment, there!} :)

Marco.

jkeny
26-10-2009, 19:03
Oh sorry, I see how it reads now :)

leo
03-01-2010, 21:05
Sounds nice driving a pair of headphones balanced;)

http://i273.photobucket.com/albums/jj220/kingbusoms/P1000965.jpg

leo
11-01-2010, 22:30
Tried some more tweaks (which I'll probably update later) also a few things can't be seen in the last pic

I'm hearing some brilliant sounds out of this thing, next dac I'm playing with is going to have to go a long way to better this:eek:

jkeny
11-01-2010, 22:42
Great Leo, I like your style - never one to rest on your laurels you're tweaking this thing to within an inch of it's life. Please post details when you get a chance.

What are you driving it with? I just (nearly) finished modding a HiFace. Great option for hi-res sound out of PC - I think you have one, don't you?

I recommend trying a separate external 3.3V supply (I use a battery) to the clocks to hear what this thing is capable of - it's held back significantly by it's switching supplies!

leo
12-01-2010, 11:39
Great Leo, I like your style - never one to rest on your laurels you're tweaking this thing to within an inch of it's life. Please post details when you get a chance.

What are you driving it with? I just (nearly) finished modding a HiFace. Great option for hi-res sound out of PC - I think you have one, don't you?

I recommend trying a separate external 3.3V supply (I use a battery) to the clocks to hear what this thing is capable of - it's held back significantly by it's switching supplies!



Sounds like you've been achieving great results with the Hiface John, I only had one on loan but do intend on getting my own sometime after I've finished tweaking the Musiland, theres a few idea's I'd like to try with the Hiface and look forward to having a probe around inside seeing what else can be done to improve things, just a shame the units are sealed and require breaking open.

I recently tried running an external supply to the Musiland which was based around a little linear job going to a Paul Hynes Z1A-5 shunt , a few minutes of use showed this to be extremely worthwhile! the filter mods on the USB line suggested by Thorsten was a worthwhile little cheap and easy tweak but isolating the USB source and running from the PH regged supply :eek: the sound really starts to improve noticeably after its had a little time to warm up.
Again this trashes the opinions all spdif sources sound the same;)

How long do you get out of the batts?

jkeny
12-01-2010, 12:04
Sounds like you've been achieving great results with the Hiface John, I only had one on loan but do intend on getting my own sometime after I've finished tweaking the Musiland, theres a few idea's I'd like to try with the Hiface and look forward to having a probe around inside seeing what else can be done to improve things, just a shame the units are sealed and require breaking open.

I recently tried running an external supply to the Musiland which was based around a little linear job going to a Paul Hynes Z1A-5 shunt , a few minutes of use showed this to be extremely worthwhile! the filter mods on the USB line suggested by Thorsten was a worthwhile little cheap and easy tweak but isolating the USB source and running from the PH regged supply :eek: the sound really starts to improve noticeably after its had a little time to warm up.
Again this trashes the opinions all spdif sources sound the same;)

How long do you get out of the batts?

I would say the results are astounding with the HiFace - I have to do some re-work on the Musiland & do a AB comparison but I don't remember the Musiland sounding as good as this.

In the Musiland, external supply on the USB is an improvement & running the PC off batteries or Linear PS is also an improvement. I think getting rid of the switching supplies helps also. As I said I have to get back to it & do some work & listening.

On the HiFace, I have 3 battery supplies - one feeding 5V via USB, one feeding 3.3V to clocks & One feeding 3.3V to the Xilinx CPLD. I also have a 3.3V battery powering my DAC & some salvaged LiPo from a laptop battery pack running my TA2020 amp so I'm completely off grid except for my laptop (I refuse to buy more batteries for it!)

I will be testing to see what supplies I can combine together & retain SQ but I was listening for about 6 hours yesterday on full batteries & when I measure them today they're all at about 3.29V. This really is no indication of what is left in them as they hold their voltage output fairly steady until exhausted (unlike other batteries). But so far I'm a happy bunny :) These are 2300mAH & I don't know the nominal current usage of the various elements in this system. I'll know the charge longevity better after a bit of usage but even if it was 12 hours I'd be happy - these LiFePo4 also have about 10 times the cycle life compared to Li ion batteries!

jkeny
12-01-2010, 17:26
I left it running from 11:00am this morning to see how long it ran for - the 5V Duracell pack just gave out at 17:00. All the 3.3V batteries are running fine at about 3.29V :). I've put a 5V transformer on the Duracells & am up and running again to see which of the 3.3V batteries goes first & when. Pic attached of the makeshift set-up. You see 3 LiFePO4 batteries 1 for clocks, 1 for CPLD I/O banks, 1 running ESS DAC.

Leo, I hope I can box it up as neat as your Sabre set-up?

PS Leo, how can you attach big images like yours above when the size limit is 97K?

jkeny
12-01-2010, 23:27
Leo, just to let you know this is still playing at 11:00 pm now so 12 hours non-stop on the 3.3V batteries so far (only the 5V Duracells ran out). I'm pretty happy so far - I'll resume tomorrow to see how long I get out of them! I expect the one supplying the ESS DAC will run out first as it's a Vout DAC with a 2V RMS output running off this one battery!

leo
12-01-2010, 23:40
Sorry I'm late getting back John:) thats good going with the batteries, certainly enough time for most folks, thank you for going to the trouble testing their run time, I'm getting the itch now with buying a Hiface:eyebrows:
I like the way you've separated things

With the pics I just upload them on photo bucket rather than uploading them here, then copy and past the image code link

leo
12-01-2010, 23:46
BTW I'm sure you'll be ok casing that lot up, if I'm honest the casing up for me is the worse bit, especially finding things suitable, I'd have preferred a bigger case with the dac, I'm tempted to fit something like a mesh type lid to allow plenty of ventilation for the shunt regs, I normally leave the lid off things anyway when in use so probably do the same with this

jkeny
12-01-2010, 23:55
Sorry I'm late getting back John:) thats good going with the batteries, certainly enough time for most folks, thank you for going to the trouble testing their run time, I'm getting the itch now with buying a Hiface:eyebrows:
I like the way you've separated thingsThanks. I reckon that giving the clock an individual clean supply is paramount to low jitter. Giving the I/O banks of the Xilinx CPLD it's own 3.3V feed also proved worthwhile. There are only 2 other improvements that I may get to - removing the final 1.8V dc-dc converter & isolating the I2S out (can't do anything to isolate the USB 2.0 at Hi-Speed 480mBits/s)


With the pics I just upload them on photo bucket rather than uploading them here, then copy and past the image code linkAh, yes :)

Ciu
24-01-2010, 12:36
Hello Leo

Is there any simple solution to "interrupt" the signal (cut the tracks, or pick-up a resistor ) from Dac Chip 9018 to the I/V, and to connect to another output stage ?
Just to obtain a direct output like the 9008 board had...

Regards
Richard

leo
24-01-2010, 15:03
Cutting the tracks on the output of the chip would be the easiest and running wires, I've not tried it yet with B32s

Dan Giovanni
24-01-2010, 17:24
Hi Jon and Leo,

Really enjoyed reading of your experiences modding the Hiface and Musiland USB converters. I gather the Hiface is the better sounding unit stock, but is there a reason why one should sound better than the other with both fully modded and using I2S output? All things being equal, I'd prefer to buy the Musiland as a) it's £50 cheaper and b) looks easier to work with. However if the Hiface is fundamentally a better unit ....

Cheers,
Dan

jkeny
05-02-2010, 22:14
Hi Jon and Leo,

Really enjoyed reading of your experiences modding the Hiface and Musiland USB converters. I gather the Hiface is the better sounding unit stock, but is there a reason why one should sound better than the other with both fully modded and using I2S output?The Hiface has 2 low-jitter clocks that it uses directly as the audio clock whereas the Musiland uses a single 25MHz clock from which the audio clocks are synthesised giving higher jitter. The Hiface clocks can be given a separate PS which is a huge leap in SQ & low jitter can be heard.
All things being equal, I'd prefer to buy the Musiland as a) it's £50 cheaper and b) looks easier to work with. However if the Hiface is fundamentally a better unit ....

Cheers,
Dan

Ali Tait
06-02-2010, 10:34
I read on diyaudio that the Buffalo II dac will be available for pre-order on the website shortly.

Themis
06-02-2010, 12:09
It's just the dac module, isn't it ? Not boxed ? :scratch:

Ali Tait
06-02-2010, 12:47
Yes,but they do other kits for output stages.I was round at a friends recently,and he has the 9008 board.Using a bench supply,he'd knocked up an ouput stage of 6n6p in push-pull with some little Sower output transformers.Sounded stunning!

leo
06-02-2010, 17:33
Been thinking of trying a valve output stage with my original ES9008 dac and see how it compares against the modded B32 with ES9018

Whats the valve guys opinion on this circuit ? http://www.lampizator.eu/lampizator/REFERENCES/Buffalo%20DAC/Loredana/3maja09%20048.jpg anything better? I can build a valve circuit no problem if theres a schematic , I know a lot more about ss than valve stuff;)

jkeny
06-02-2010, 21:51
Why don't you guys try a transformer output stage - anywhere I've tried them on voltage out DACs they have sounded wonderfully natural & I've heard similar opinions about a transformer output on the Sabre DAC. Forget about the THD/SNR numbers - we all know they don't tell what sounds good

Dan Giovanni
06-02-2010, 23:45
I'm going to be ordering one of the Jensens soon, probably their 4:1 line input transformer. Also got a Hiface ordered, and will of course mod the crap out of it. :)

Dan Giovanni
06-02-2010, 23:58
Been thinking of trying a valve output stage with my original ES9008 dac and see how it compares against the modded B32 with ES9018

Whats the valve guys opinion on this circuit ? http://www.lampizator.eu/lampizator/REFERENCES/Buffalo%20DAC/Loredana/3maja09%20048.jpg anything better? I can build a valve circuit no problem if theres a schematic , I know a lot more about ss than valve stuff;)

This guy seems to know what he's doing ....

http://www.tubecad.com/2006/08/blog0075.htm

jkeny
07-02-2010, 14:57
Just so you guys don't think the Buffalo is the only kid on the block & to give some competitive variety & consumer choice in your choosing of a Sabre DAC implementation - the Acko board is nearing completion & prices have been announced - $50 for board (for all us DIYers with NDA's - that's about all 3 of us), from $220 for stuffed board. http://www.diyaudio.com/forums/digital-line-level/147817-ackodac-based-es9018-11.html#post2076598

One thing that always struck me about the Buffalo was the use of screw connectors for high speed signals - did nobody cringe at this? I'm no RF E'ee but my brief 101 reading about high speed signalling and all the care & caution necessary with regard to avoiding sharp corners & impedance jumps, did make me think WHAT-THE- ****?

I think they have addressed it in version II but I don't know ?

Dan Giovanni
09-02-2010, 23:52
Yeah, there have been quite a details of the TP DACs that have been less than optimal - I guess they are learning as they go. The worst thing to my mind was supplying the clock, digital section of the DAC, the S/PDIF comparator and uProcessor all from the same supply (a 50p 3 pin reg) without even ferrite beads to decouple them. Also deriving the critical analog reference from this same supply with only a 150Hz low pass filter. Still, it would be kinda boring if there was nothing to tweak.

leo
08-06-2010, 16:03
Updated pic with the PR3 installed used as a pre-regulated PSU for onboard shunts , I also moved the AVCC L&R shunts closer the dac (across the decoupling ceramics)

http://i273.photobucket.com/albums/jj220/kingbusoms/P1060343.jpg

Next thing is probably BII to experiment with alternative output stages

leo
29-06-2010, 11:18
I've been wanting to try the ultra cheap ES9022 chip so I lashed one up at the weekend, currently running from a 9v battery (into a 3v3 LDO reg)
I tapped into the modified Duet's I2S lines

Sound isn't half bad at all for such a simple cheap thing :)

leo
03-08-2010, 21:53
Another one to try out (BII) , this one uses a separate I/V output stage so makes it much easier to try alternatives like Nicks valve output stage compared to the onboard one found on the Buffalo32 I last modded

http://i273.photobucket.com/albums/jj220/kingbusoms/P1070563.jpg

squeezethatbuffalo
04-08-2010, 11:34
Here's my modded Buffalo I (sorry about the poor image quality, blame Apple ;)):

http://img155.imageshack.us/img155/6434/img0046zk.jpg


BIG thanks to leo for his inspiration on Buffalo modding! Also, kudos to TPA, Paul Hynes, and K&K audio for their great products & support!

My "Original" Buffalo: stock Buffalo DAC, powered by two LT1086-based PS. IVY power by a super Jung type PS.

Modded Buffalo: Buffalo DAC (PH Z173v3 supplying AVCC), powered by Z1P5v45 (with LM317 supply as pre-reg), directly connected to Lundahl 1527 transformers.

Though falling short from being "hi-end", the Buffalo was already a great DAC in stock form (especially when one considers its price tag). After the mods, however, it is capable of putting quite a few big name DACs to shame. :eek:

After hearing what the 1st gen Buffalo can do, now I really can't wait to get my Buffalo II! Still one month to go... :doh:

Cheers,
Alex

leo
04-08-2010, 16:12
Hi Alex

Nice work there! The pics are fine and good enough to show us what you've done.
The Lundahl's are much smaller than I thought they would be, I'd like to try transformer output myself sometime

Oh BTW Please could you just nip into the welcome section of the forum and introduce yourself before the site owners jump on you ;)

Cheers,
Leo

squeezethatbuffalo
05-08-2010, 03:01
Hi Leo,

To be honest, I never expected the preregs to make such a difference. The improvement brought by Z17s was anticipated, but the Z1P definitely took me by surprise! I guess that even run-of-the-mill LDOs can be "saved" by high quality power supplies.

I think the Lundahls transformers are definitely worth trying. Even their "low-end" 1527s sound pretty good. I've been thinking about pairing my Buffalo II with their amorphous core products.

BTW, should I increase the size of the filter cap? The 470uf muse kz that came with the kit sounds fine, but I suspect it might be a tad small. What do you think?

Cheers,
Alex

leo
06-08-2010, 23:10
I think when you use the dac as voltage output the regulation for parts of the dac has even greater influence on the sound .

Is the 470uf kz cap before or after the regulator ? if its before the regulator on the raw DC it maybe worth making it a little bigger, if its after the regulator it should be fine.

squeezethatbuffalo
07-08-2010, 12:42
Hi Leo,

I think it proves once again that PH regs are just that much better than 3-pin regulators ;)

The 470uf KZ is before the regulator. Would a 2200uf cap be excessive (the current draw is about 530ma)?

Thanks,

Reid Malenfant
07-08-2010, 12:52
By increasing the capacitance on the DC rail prior to the regulator you'll reduce the ripple voltage that the regulator has to deal with. Reduced ripple on the input will result in a reduced ripple on the output (even though it might be vanishingly small anyway).

With just over a half Amp of current draw i'd definately be tempted to increase the storage capacitance. 2200uf would result in less than a quarter of the ripple voltage on the DC rail, sounds good to me if you have a similar specification capacitor :)

leo
07-08-2010, 12:59
Hi Alex,

2200uf should be fine used on the RAW DC if its nice and stiff, The shunts take more current anyway so the transformer is obviously meaty enough so not to sag too much on the output

If its ok now though with the 470uf you may want to leave it as is ;)

squeezethatbuffalo
07-08-2010, 13:04
Hi Mark,

In fact I do have a 2200uf/35v Black Gate FK lying around somewhere. I'm not sure about the specifications, but from past experience BG FKs rarely disappoint. Now I just have to find it...

Regards,

squeezethatbuffalo
07-08-2010, 13:10
Hi Leo,

The transformer is 72VA rated, but I'm only using "18V/1A" of it right now.

The sound is really good, but it could be a little bit more bassy and less straightforward. Hmm, I think I'll invite Miss Golden Ears over tomorrow and see what she thinks.

Regards,
Alex

Edit: Leo, can you please elaborate? I only got the part that "using a 470uf as main filter cap is okay", but not sure what you meant by "the shunts take more current anyway so the transformer is obviously meaty enough so not to sag too much on the output". Are the two sentences related, or am I just being a confused newbie?

Just a thought... since I'm using a LM317 reg module as a pre-reg to the Paul Hynes shunt, the ripple at the shunt's input should be quite low anyway (compared to "normal" rectified DC), right? If so, then perhaps I can get away with using a small filter cap on the 317 module. :confused:

leo
07-08-2010, 15:18
Hi Alex,

I just mean't a shunt reg consumes more current than say a LM317, your linear psu is running it fine so upping the capacitance of the filter cap won't be a problem. Sometimes using high value capacitance with a weedy transformer can give poor results, yours is fine so 2200uf may even give you a bit more bass.

BG FK's are supposed to make great filter caps, It maybe worth giving it a go anyway if you can find it.
The only real way to know is try it, your now getting to the icing on the cake stage :)

squeezethatbuffalo
07-08-2010, 16:03
Hi Leo,

I just found a 1000uf BG NX! Too bad that it's rated at 25v.

I'll keep looking for that 2200uf/35v FK. This is one of the rare occasions where I wish I was more organized... :doh:

Oh, and I disagree. I won't even have my new cake until September! ;)

Cheers,

leo
15-08-2010, 21:30
Just wired up quick for testing out

squeezethatbuffalo
15-08-2010, 21:35
Sweet :eek: Eager to hear your thoughts on the BufII/IVYIII!

Edit: Do you still plan on trying out Nick's tube output stage?

leo
16-08-2010, 21:57
Hi Alex,

I'll probably try the valve output stage last mainly because it will be the most expensive thing regarding parts, especially the transformers .

First will be IVYIII, next will be Legato probably followed by other various idea's etc.

I've been lisening to the dac tonight, sound wise so far it seems to offer a lot of potential , tbh I expect a definite gain in performance after fitting the PH regs on dacs board for the AVCC L&R, XO, VD and core

leo
16-08-2010, 21:58
BTW so far I feel its not as good as the modified B32, hopefully that will soon change

squeezethatbuffalo
16-08-2010, 22:59
Hi Leo,

Don't worry, the PH shunts will work their magic ;)

I hope it's not too much to ask, but can you please compare your PH against the "stock" mini shunt? I'm really curious about the TPA shunt's performance, as Russ claims that it has been fine tuned for this particular application.

To be honest, I've been having second thoughts about the valve output stage. Cost is one factor, but my main concern is due to my listening habits; I normally leaving my system on/mute when I'm away, and I don't think tube equipment likes that. :P

Cheers,

leo
17-08-2010, 00:25
Hi Alex,

I intend to compare the PH regs against TP's shunt, I'm more than aware of the claims but I know at least one person already who has compared them :eyebrows: of course I prefer to try these things myself before saying anything.
Its going to be fiddly because I've fitted the PH shunts very close to the pads on the B32, I have to be careful not to damage anything removing them from the B32, one of the advantages of BII is the sockets to allow simpler swapping about. Once removed I have to replace the legs on the PH shunts too because they are currently cut short (too short for the sockets )

squeezethatbuffalo
17-08-2010, 11:28
Hi Leo,

I know exactly what you mean by "fiddly". Did I mention that I used to own a 32s? Well, now it's just another broken piece of PCB :doh:

Don't be too harsh on the Buffalo II :lol: I remember that the 32s sounded funny at first :( but improved significantly after a little (okay, more like alot) run in.

leo
23-08-2010, 23:54
Legato is ordered, being impatient I've also knocked one up on matrix board , I've not tried it yet, also instead of using a dual op-amp for the differential to line for both channels I added a separate single op-amp on each channel.

I'll mainly be using differential output but the chips are added for convenience so I can have both balanced and unbalanced

squeezethatbuffalo
24-08-2010, 01:27
Hi Leo,

Do you plan on changing the output coupling caps? Have you considered about using transformers instead?

Cheers,

leo
01-09-2010, 20:13
Just running in Legato output stage, they use discrete for the differential and op-amp for the differential to line, of course you can easily run it discrete out for balanced and run it into say a transformer converting it to SE if you don't want to use the op-amp.
I've already briefly tried Legato I knocked up on vero but will run in the proper one before commenting on the sound :)

http://i273.photobucket.com/albums/jj220/kingbusoms/P1090066.jpg

squeezethatbuffalo
02-09-2010, 15:38
Very looking forward to the battle between Legato and IVYIII ;)

I can't help but notice that your Legato looks quite different from stock. Do you mind discussing the mods with us?

leo
03-09-2010, 01:49
Hi Alex,

After some run in I still wasn't liking BII + IVYIII as much as highly modified B32, I don't know the circuit for the IVY on B32, maybe its different to IVYIII :scratch:

Initial impressions of Legato are good, so far I like it a lot better than IVYIII , in comparison IVY sounds mechanical

I've added sockets for the coupling caps to allow comparisons and also sockets for the output filter, I've been trying lower values.
I've never been a huge fan of Wima caps so those 15nf's was the first to go :)

squeezethatbuffalo
03-09-2010, 03:54
Hi Leo,

That's very interesting, as I remember TPA folks saying that the IVYII and IVYIII are "almost identical". I guess "almost" just doesn't cut it ;)

It's good to know that the Legato sounds good. I plan on swapping out the decoupling caps for BC037s, as I never really liked Nichicon caps. The bipolars will have to go as well, I think I'll use my LL1527s instead of caps.

Not sure how it'll sound, but Russ seems to think it's a good idea.

This actually is a very good scenario because the DAC sees a near 0 ohm load, and the transformer sees a low impedance source. Both ends are happy.

leo
05-09-2010, 00:06
I lowered the output voltage of the Z1A/Z1N from 15v to 12v , Legato runs a little cooler. Just the XO, VD and core regs need upgrading to PH shunts (series reg for the 1v2 core)

Think I'll change those Nichicon KW decoupling caps too tbh, anybody have any recommendations for something decent on analog supplies which doesn't add lots of signature?

I also need to try alternatives for the coupling caps

http://i273.photobucket.com/albums/jj220/kingbusoms/BIIlegato1.jpg

http://i273.photobucket.com/albums/jj220/kingbusoms/BIIlegato.jpg

squeezethatbuffalo
05-09-2010, 04:21
Hi Leo,

I find the BC037s rather decent and less "colorful" than many other caps I've used. But it sounds like that you aren't really interested ;)

I've yet to see (hear) how they go with Legato though. Hurry up Mr. Donegan! :lol:

As for coupling caps, have you considered using film caps? May be worth trying.

Cheers,

leo
05-09-2010, 10:27
Hi Alex,

I'd like to try the BC037's but I don't think they are available from places like Farnell anymore, I'll have to check

I may try film caps for coupling although tbh lytics are not too bad if properly biased , a lot of film caps over 1uf I've tried sound what they are, plasticky :lolsign:

Edit according to Farnell, no longer manufactured http://uk.farnell.com/vishay-bc-components/2222-037-51101/capacitor-100uf-50v/dp/1165335

squeezethatbuffalo
06-09-2010, 12:15
Hi Leo,

That's a shame. The BC037s still available locally, so it never occurred to me that they went out of production.

I'd like to recommend some very nice but inexpensive film caps, but they're probably difficult for you to obtain as well. Maybe I should stay quiet this time to avoid looking like an ass :lol:

Just one more question, please. You said:

I lowered the output voltage of the Z1A/Z1N from 15v to 12v , Legato runs a little cooler.
Does lowering the voltage affect sonic performance?

Thanks,

leo
06-09-2010, 18:51
Most of the good components are harder to find over here thanks to damn ROHS :wanker:
Anyway let us know which film caps you recommend , maybe ebay will have some

I decided to change the 4 x bipolar 100uf Nichicon coupling caps to 4 x 10uf 50v Wima. I'm not a huge fan of Wima but must admit these are clearly better than the Nichicon bipolars :) the added bonus is their compact too being almost the same size as the Nichicons. Bass to high's are cleaner for sure so until I can find something better the Wima's will stay inplace .
For those with a Legato , try an alternative for the coupling caps, you may be pleasantly surprised :eyebrows:

I'm not sure lowering the voltage with Legato makes a huge difference sonically , tbh if there was it must have been small

squeezethatbuffalo
07-09-2010, 03:03
Most of the good components are harder to find over here thanks to damn ROHS :wanker:
Anyway let us know which film caps you recommend , maybe ebay will have some

I decided to change the 4 x bipolar 100uf Nichicon coupling caps to 4 x 10uf 50v Wima. I'm not a huge fan of Wima but must admit these are clearly better than the Nichicon bipolars :) the added bonus is their compact too being almost the same size as the Nichicons. Bass to high's are cleaner for sure so until I can find something better the Wima's will stay inplace .
For those with a Legato , try an alternative for the coupling caps, you may be pleasantly surprised :eyebrows:

I'm not sure lowering the voltage with Legato makes a huge difference sonically , tbh if there was it must have been small

ROHS? That's something I'm still blissfully unaware of :lol:

The caps I'm referring to are the MKP2000 (http://theartofsound.net/forum/showpost.php?p=147613&postcount=93). I'd describe them as "better Wimas", as they're relatively inexpensive and compact compared to more exotic film caps, and I find it OK to use them in multiple locations without adding too much color to the sound. Oh, did I mention that they sound noticeably better than the Wimas? ;)

The official page is here (http://ucccap.myweb.hinet.net/). They also have an incomplete English (http://ucccap.myweb.hinet.net/ucccapeng.htm) version of it, which has no mention of the MKP2000s. I guess it's some kind of jealously guarded secret... :ner:

If you're interested, just send the site owner an e-mail. There's no need for Google translate btw, English is a mandatory school subject in Taiwan :eyebrows:

And thanks for yours thoughts on Legato's voltage. Better set it at +/-12v then. Silly class A devices!

Cheers,

leo
07-09-2010, 08:28
Thanks Alex, I'll have to give the MKP2000 a try out :) Their bigger than the compact Wima's but at least their not huge compared to a lot of other plastic film types.

BTW I found this , I presume they are the same ? http://cgi.ebay.co.uk/Metalized-Polypropylene-Capacitor-10uf-100V-PIO-OIL-/270627396289?pt=Vintage_Electronics_R2&hash=item3f02a65ac1

Cheers,
Leo

squeezethatbuffalo
07-09-2010, 11:14
Hi Leo,

The picture is the same. But I've never dealt with the ebay seller, so I'm not 100% sure if they'll deliver the right ones. Though judging from the seller's feedback rating, I guess we can be 99.9% sure :lol:

I'd personally buy the caps direct for 20%+ cheaper price. Then again, if ebay is more convenient, then maybe you should use that instead.

Cheers,

leo
07-09-2010, 15:13
Hi Alex,

The ebay shop seemed easier to use, I'll send UCC Audio a mail first though asking payment details etc :)

Cheers,
Leo

squeezethatbuffalo
07-09-2010, 15:40
Hi Leo,

IMO, the MKP2000s wipe the floor with Wimas as output coupling capacitors. I hope you'll feel the same way :)

Cheers,

leo
07-09-2010, 22:22
I'll let you know how I find the MKP2000 if/when I can get hold of them

Just been listening to the dac tonight , got to admit some tracks really blow my bloody socks off :eek: I've got the XO, VD and core regs to upgrade yet

squeezethatbuffalo
08-09-2010, 04:08
Sweet! How does Buffalo II + Legato compare to your modified Buffalo 32s? Which one do you like better?

Edit: I just received confirmation from ucc audio. The ebay seller is an authorized dealer, so it's 100% okay ;)

leo
08-09-2010, 14:44
Hi Alex,

I've not heard back from UCC yet but thanks for finding out that the ebay shop is an authourized seller :)

BII + Legato is getting closer to modded B32, I still prefer the B32 in some ways but hopefully that will change after BII's XO, VD and core regs are replaced, oh and I've tried those caps ;) BII doesn't quite have that flow the modded B32s has YET

squeezethatbuffalo
08-09-2010, 14:50
Hi Leo,

What are "those caps"? The MKP2000s? How did they magically appear on your Legato? ;)

Cheers,

leo
08-09-2010, 16:45
Hi Alex,

I mean't the performance may be improved further after I have replaced the Wima's with the new caps (MKP2000)
Sometimes I type too quickly and it don't make much sense :lol:


Cheers,
Leo

squeezethatbuffalo
09-09-2010, 08:05
Hi Leo,

It would make alot more sense if "I've tried" were removed from the sentence ;)

By the way, have you heard from ucc? Perhaps you should try ebay if he's being lazy :lol:

Edit: Earlier in this thread, you mentioned that you don't like signal transformers. Do you mind sharing the reasons? Don't worry, I'm just curious, no heavy objects coming your way :P

Cheers,

Dan Giovanni
09-09-2010, 23:06
Hi Alex,

have you been able to compare the MKP2000s to other caps, e.g. Solens, Sonicaps, Ansars, etc? I've got a couple of 47uF MKP2000s knocking around that I've not found a use for yet. May use them in my crossovers.

leo
10-09-2010, 19:04
Hi Alex,

Still no email from UCC so I've just ordered 4 x 10uf 100v from the ebay link.

squeezethatbuffalo
13-09-2010, 19:08
Hi Alex,

have you been able to compare the MKP2000s to other caps, e.g. Solens, Sonicaps, Ansars, etc? I've got a couple of 47uF MKP2000s knocking around that I've not found a use for yet. May use them in my crossovers.

Hi Dan,

I never knew 47uF MKP2000s existed! Are you referring to the 47uF Fast Power caps from the same company, or are those custom made?

Never tried MKP2000/Fast Power in crossovers, so I can't really help you there. I do know that they work well in power supplies though. Maybe you can use them in a tube amp!

Cheers,

squeezethatbuffalo
13-09-2010, 19:18
Hi Alex,

Still no email from UCC so I've just ordered 4 x 10uf 100v from the ebay link.

Hi Leo,

Great! Let us know how it sounds in your Legato!

I have a feeling that you'll receive the caps before I get my Buffalo II (the batch which was supposed to ship on the week of Sep 5). I guess the good folks of TPA are overwhelmed by the orders. Delivery was punctual back when I purchased my Buffalo32 :lol:

Cheers,

Dan Giovanni
13-09-2010, 22:17
Hi Alex,

Maybe I'm mistaken - I also bought them from Valab and assumed they were the same brand as the 10uFs you were talking about. The 10uFs sound ideal to me, except I have a slight reluctance based on some cheap MKP caps sounding tizzy e.g. Solen.

Thanks,
Dan


Hi Dan,

I never knew 47uF MKP2000s existed! Are you referring to the 47uF Fast Power caps from the same company, or are those custom made?

Never tried MKP2000/Fast Power in crossovers, so I can't really help you there. I do know that they work well in power supplies though. Maybe you can use them in a tube amp!

Cheers,

leo
13-09-2010, 23:26
Hi Leo,

Great! Let us know how it sounds in your Legato!

I have a feeling that you'll receive the caps before I get my Buffalo II (the batch which was supposed to ship on the week of Sep 5). I guess the good folks of TPA are overwhelmed by the orders. Delivery was punctual back when I purchased my Buffalo32 :lol:

Cheers,

Hi Alex,

Yes, I'll be sure to update how I find the MKP2000, hopefully their better than the other polypropylenes I've tried.

The XO,VD and core regs from Paul Hynes have arrived, I'm running them in on a jig for a few days before fitting them in the BII

leo
13-09-2010, 23:30
Hi Alex,

Maybe I'm mistaken - I also bought them from Valab and assumed they were the same brand as the 10uFs you were talking about. The 10uFs sound ideal to me, except I have a slight reluctance based on some cheap MKP caps sounding tizzy e.g. Solen.

Thanks,
Dan

Hi Dan,

I've also found a lot of polypropylene caps to sound tizzy and often quite synthetic/ plasticky so I'm hoping the MKP2000's are better

squeezethatbuffalo
15-09-2010, 14:27
Hi Alex,

Maybe I'm mistaken - I also bought them from Valab and assumed they were the same brand as the 10uFs you were talking about. The 10uFs sound ideal to me, except I have a slight reluctance based on some cheap MKP caps sounding tizzy e.g. Solen.

Thanks,
Dan

Hi Dan,

They're from the same brand, but seem to belong to different product lines. Since I haven't found a feasible way to compare them due to different capacitance/voltage rating, I can't tell whether they're the "same" caps with different names (they do look alike).

If you can provide more details of the intended application, I may be able to contact UCC and ask for recommended types/values.

Btw, I'm learning Japanese right now, so don't be surprised if I start recommending gadgets from Japan in a year or two. :eyebrows:

Cheers,

squeezethatbuffalo
15-09-2010, 14:35
Hi Alex,

Yes, I'll be sure to update how I find the MKP2000, hopefully their better than the other polypropylenes I've tried.

The XO,VD and core regs from Paul Hynes have arrived, I'm running them in on a jig for a few days before fitting them in the BII

Hi Leo,

My BII is finally on its way! Unfortunately, the chance of finishing the DAC this weekend is quite slim, as it's still in the States as I'm typing this. But who knows, maybe EMS will outdo itself this time.

Still curious about your thoughts on transformers, but it seems like you don't want to talk about it :lol:

Cheers,

leo
15-09-2010, 17:53
Hi Leo,

My BII is finally on its way! Unfortunately, the chance of finishing the DAC this weekend is quite slim, as it's still in the States as I'm typing this. But who knows, maybe EMS will outdo itself this time.

Still curious about your thoughts on transformers, but it seems like you don't want to talk about it :lol:

Cheers,

Hi Alex,

It'll give you something to look forward to ;)

A lot transformers are reported to suffer from things like phase distortion etc , this puts me off using them tbh . I may try some out sometime though if I can find decent ones

squeezethatbuffalo
15-09-2010, 18:41
Hi Alex,

It'll give you something to look forward to ;)

A lot transformers are reported to suffer from things like phase distortion etc , this puts me off using them tbh . I may try some out sometime though if I can find decent ones

Hi Leo,

I understand what you mean. I've heard my fair share of horror stories about transformers.

To be honest, the main reason I went for transformer passive output was because IVYIII went out of stock and I didn't feel like waiting :lol: But I'm happy to report that it turned out more than fine. Going by memory, I enjoy my Buf24 more than my old Buf32s. Then again, the latter had no PH shunts on it, so the comparison isn't really fair.

I plan on comparing the transformers to the "stock" bipolars and other caps once I receive my BII + Legato. Unfortunately, my current amp is single-ended, so I'm not sure how helpful it will be to those with proper balanced equipment ;)

Cheers,

Dan Giovanni
15-09-2010, 23:14
I have some Jensen 4:1 transformers that I'll be using with my Legato in the not-to-distant future. There's around 6-7mV dc offset on the balanced output of my Legato, which also drifts around with temperature. I need to check with Jensen to make sure this is OK (without coupling caps). The good thing about the Legato is that the buffers give very low output impedance. The Jensen data-sheets clearly show transformer distortion at it's best with very low driving impedance.

squeezethatbuffalo
16-09-2010, 13:13
Hi Dan,

Do you already have your Legato, or are you still waiting for it?

I agree. Low input/output impedance along with lower distortion are the main reasons I'm putting a Legato between the BII and my transformers. I'm also curious on how Legato sounds stock. Hopefully I'll be able to find out in a few days.

Cheers,

Dan Giovanni
16-09-2010, 14:44
Yes, mine came a few weeks ago - I ordered it as soon as it became available. I've heard from Jensen, and it seems I don't need the coupling caps, so that's one less thing to worry about. Hopefully it sounds as good, or better, with the trannies as it does with the opamp. It's very good already.

I've also got in mind some PSRR mods for the Legato, but they're on hold 'till other stuff is sorted.

Look forward to hearing your impressions .... I think you'll like it! :)


Hi Dan,

Do you already have your Legato, or are you still waiting for it?

I agree. Low input/output impedance along with lower distortion are the main reasons I'm putting a Legato between the BII and my transformers. I'm also curious on how Legato sounds stock. Hopefully I'll be able to find out in a few days.

Cheers,

squeezethatbuffalo
16-09-2010, 15:17
Yes, mine came a few weeks ago - I ordered it as soon as it became available. I've heard from Jensen, and it seems I don't need the coupling caps, so that's one less thing to worry about. Hopefully it sounds as good, or better, with the trannies as it does with the opamp. It's very good already.

I've also got in mind some PSRR mods for the Legato, but they're on hold 'till other stuff is sorted.

Look forward to hearing your impressions .... I think you'll like it! :)

Sounds like you've been using the built-in BALSE stage. Let us know how they compare to the transformers :)

Do you mind sharing what you have in mind? Sounds very intriguing!

I'll find time to build the DAC after my order arrives, hopefully I'll be able to listen to it during the weekend.

Dan Giovanni
17-09-2010, 21:37
Sounds like you've been using the built-in BALSE stage. Let us know how they compare to the transformers :)

Do you mind sharing what you have in mind? Sounds very intriguing!

I'll find time to build the DAC after my order arrives, hopefully I'll be able to listen to it during the weekend.

The Legato uses resistors for current sources, and while a simple and elegant solution, they will give the circuit rather poor PSRR.

The mods I'm thinking of are either adding extra RC filtering on the rails, or using active current sources. I've got a few ideas, but need to work out the details. Got a bunch of other stuff to do first -hopefully will make some headway over the weekend.

leo
17-09-2010, 22:37
Just having a play about with the BII regs, once the SMD inductors on the underside of pcb are removed to disconnect the BII onboard regs (Core, VDD, XO) it seems the caps going to the input of the stock regs are disconnected too . It maybe worth trying adding input caps to the alternative regs to be fitted especially if the main supply source is not close to the regs, I'm going to try some caps with the inputs to the PH regs

http://i273.photobucket.com/albums/jj220/kingbusoms/P1090331.jpg

leo
17-09-2010, 22:40
Also thing to note, even though the input of the stock regs is disconnected their output is still connected so it may be worth doing a small track cut to isolate them from the upgraded regs output

squeezethatbuffalo
18-09-2010, 04:18
The Legato uses resistors for current sources, and while a simple and elegant solution, they will give the circuit rather poor PSRR.

The mods I'm thinking of are either adding extra RC filtering on the rails, or using active current sources. I've got a few ideas, but need to work out the details. Got a bunch of other stuff to do first -hopefully will make some headway over the weekend.

Great, I'm already thinking about modding my Legato though I don't even have it :lol:

Just one more question, please. What supply(s) are you using?

squeezethatbuffalo
18-09-2010, 04:20
Also thing to note, even though the input of the stock regs is disconnected their output is still connected so it may be worth doing a small track cut to isolate them from the upgraded regs output

Thank you for sharing! I'll keep that in mind :)

squeezethatbuffalo
19-09-2010, 10:06
Yes, mine came a few weeks ago - I ordered it as soon as it became available. I've heard from Jensen, and it seems I don't need the coupling caps, so that's one less thing to worry about. Hopefully it sounds as good, or better, with the trannies as it does with the opamp. It's very good already.

I've also got in mind some PSRR mods for the Legato, but they're on hold 'till other stuff is sorted.

Look forward to hearing your impressions .... I think you'll like it! :)

Hi Dan,

Which Jensen transformers will you be using with the Legato?

Somehow I missed the DC offset part of your previous post. According to the datasheet, the core of LL1527 "is a high permeability mu metal core", no mention of air gaps. Now I'm worried about saturation :doh:

Cheers,

Dan Giovanni
19-09-2010, 11:40
Hi Alex,

My supply is currently a "Dual Reg" by Greg Ball ...

http://www.ska-audio.com/Forum/YaBB.pl?custompage=products

Good, but not great. I've got a discrete shunt reg based on Salas' design almost finished too that should be better. What about you - what are you doing for a PSU?

My trannies are Jensen 4:1 step down types. The Lundahls my be OK without caps too .... perhaps you could email them to find out?

Dan

squeezethatbuffalo
19-09-2010, 13:12
Hi Alex,

My supply is currently a "Dual Reg" by Greg Ball ...

http://www.ska-audio.com/Forum/YaBB.pl?custompage=products

Good, but not great. I've got a discrete shunt reg based on Salas' design almost finished too that should be better. What about you - what are you doing for a PSU?

My trannies are Jensen 4:1 step down types. The Lundahls my be OK without caps too .... perhaps you could email them to find out?

Dan

Hi Dan,

I plan on using "Golden Reference" (http://ucccap.myweb.hinet.net/pics/NGRsch01.jpg) power supplies for my Legato, and Paul Hynes for the BII. They're no Paul Hynes, but I find them really good for analog applications. Not nearly as good for digital though.

I got a reply from Kevin (from K&K Audio), he said 6~7mV would "cause you to lose some low bass". I just wrote him another e-mail asking about the severity of the issue. If it seriously hinders performance, I may add caps or go for gapped transformers. Not sure which is the lesser evil ;) Did you hear something similar from Jensen?

Edit: Oh and btw, I finally got my BII+Legato today! "The post office doesn't deliver mail on Sundays", you say. No they don't, but I was determined enough to drive through a typhoon to pick it up. :lol:
http://library.thinkquest.org/03oct/01027/typhoon.jpg

Cheers,
Alex

Dan Giovanni
19-09-2010, 21:27
Those regs look like a text-book implementation of the Walt Jung reg, in other words very good! Better than the TP Placed supplies, as the opamp gets its juice from its own regulated rails, rather than the raw supply.

I was told by Jensen that the offset would give no problems, that it was well within spec. I think in your case just trying the transformers with and without caps will give you your answer.

In my case I wasn't massively impressed with the Jensens directly on the output of the Sabre - a bit soft and bland - so I'm hoping they do better fed with the much lower impedance of the Legato. If not it's back to finding the best op-amp I can.


Hi Dan,

I plan on using "Golden Reference" (http://ucccap.myweb.hinet.net/pics/NGRsch01.jpg) power supplies for my Legato, and Paul Hynes for the BII. They're no Paul Hynes, but I find them really good for analog applications. Not nearly as good for digital though.

I got a reply from Kevin (from K&K Audio), he said 6~7mV would "cause you to lose some low bass". I just wrote him another e-mail asking about the severity of the issue. If it seriously hinders performance, I may add caps or go for gapped transformers. Not sure which is the lesser evil ;) Did you hear something similar from Jensen?

Edit: Oh and btw, I finally got my BII+Legato today! "The post office doesn't deliver mail on Sundays", you say. No they don't, but I was determined enough to drive through a typhoon to pick it up. :lol:
http://library.thinkquest.org/03oct/01027/typhoon.jpg

Cheers,
Alex

squeezethatbuffalo
20-09-2010, 09:08
Those regs look like a text-book implementation of the Walt Jung reg, in other words very good! Better than the TP Placed supplies, as the opamp gets its juice from its own regulated rails, rather than the raw supply.

I was told by Jensen that the offset would give no problems, that it was well within spec. I think in your case just trying the transformers with and without caps will give you your answer.

In my case I wasn't massively impressed with the Jensens directly on the output of the Sabre - a bit soft and bland - so I'm hoping they do better fed with the much lower impedance of the Legato. If not it's back to finding the best op-amp I can.

The Golden Reference regulators are Jung-type regs with optimized layout and carefully selected components. That's what they say, anyway ;)

They're available as kits or built modules from ucc audio, the same fellas who sell fast power and MKP2000 caps. More details here (http://ucccap.myweb.hinet.net/howtodoGUMR.html).

I really like the regs. Very good performance, inexpensive, and ships right away (as opposed to like 30 days later :ner:). Unlike good ol' Paul Hynes regs, however, they don't seem to work well as digital supplies. A decent 3-pin regulator implementation betters it in that regard.

TP Placids aren't bad, but it appears that they work best with a pre-reg. Not my first or second choice :lol:

I think I'll use the LL1527s with caps for now, and upgrade to a bigger/better transformer later.

I hope the Jensens would work well with your BII/Legato!

Edit: I think the Golden Reference and the Mmini regulator (http://www.diyaudio.com/forums/power-supplies/63844-mmini-regulator-sample-offering.html) are designed by the same guy

andreime
23-09-2010, 20:42
try some known brands for output caps- Mundorf, Ampohm and such - MKP or PIO or paper etc.

I had a very nice experience with an MKP with polystyrene bypass. give it a try, the one I used was a 1uF Mundorf Mcap with a .001 bypass.

I'm using it in an arcam alpha 5+ and will experiment soon with a better MKP from them and even some paper ones. Will post results :)

leo
23-09-2010, 21:56
Hi Alex,

Those 10uf MKP2000 caps have arrived, I was surprised how light they are, I actually thought the box would be empty :lol:
I've fitted them in the Beresford Caiman dac first to see how they sound in that, I'm going to give them some running in for a few days

BTW, I have one of those mini regs too ;) well I think the one I have is an older version, I was testing them along with ALW's etc

squeezethatbuffalo
23-09-2010, 23:43
Hi Leo,

I think you won't be disappointed :eyebrows:

The mini regs were long out of production before I became aware of the company, but the latest version works for me.

Oh, and I'm almost done with my Buffalo II! So psyched! Expect my impressions soon®.

Cheers,
Alex

*soon® is a trademark of Twisted Pear Audio, registered in numerous countries :lol:

squeezethatbuffalo
24-09-2010, 13:46
The Buffalo II sings! Kind of...

Ironically, right after I sang praises about it, one of the GR regs refused to work properly (negative rail supplies -8v instead of -15) :( Gonna send it back to UCC for repair on Monday. No dual mono for me until it returns, oh well.

And for some reason, my Volumite refuses to cooperate with my Buffalo II. I'm supplying the onboard LDO with a separate 5v supply, the SLA/SCL wires are connected to BufII, and the chip on BufII has been removed as instructed. I checked all connections, and everything seems ok (except for the fact that Volumite doesn't work). :scratch: Have I missed something obvious, or is it more likely that the Volumite is defective?

Oh well, at least my BufII is playing music :lol:

Edit: Okay, Russ kindly reminded me that the Volumite GND should be connected to the BuffII GND. That should be it :P

leo
24-09-2010, 16:07
Nice one Alex,

Shame about the faulty reg, at least it shouldn't be too hard to repair. What error amp is fitted in your regs BTW?

Let us know how you get on with the Volumite , I've not actually fitted one yet in the BII, I might borrow the one from my B32

BTW the MKP 2000 caps are running in ok, they are still in the Caiman dac, I'll transfer them over to the BII in a few days.
First impression when fitted in the Caiman the sound was a little strained but its starting to open out. How do you find the bass with these caps ? I'm finding it a little soft, this may be because their new and need more running in :) I'll comment more on their progress ina few days

andreime
24-09-2010, 21:18
To what caps are you comparing them?

Dan Giovanni
24-09-2010, 21:26
Sorry you are having issues getting it all going - does anything ever work first time?!

I don't know your level of technical experience, but if you like I can give you some pointers for fault finding the regulator. First thing .... does the reg give -8V when NOT connected to the DAC, and is anything getting hot?

squeezethatbuffalo
24-09-2010, 23:36
Nice one Alex,

Shame about the faulty reg, at least it shouldn't be too hard to repair. What error amp is fitted in your regs BTW?

Let us know how you get on with the Volumite , I've not actually fitted one yet in the BII, I might borrow the one from my B32

BTW the MKP 2000 caps are running in ok, they are still in the Caiman dac, I'll transfer them over to the BII in a few days.
First impression when fitted in the Caiman the sound was a little strained but its starting to open out. How do you find the bass with these caps ? I'm finding it a little soft, this may be because their new and need more running in :) I'll comment more on their progress ina few days

Hi Leo,

Not sure what happened to the regs, but it's okay, repairs are usually free. I just don't feel like waiting. Oh, I think it's LME49710NA, if that's what you're referring to.

I'll be out of town for a few days, so no fiddling with audio until I get back. :(

The MKP2000s are IMO quite balanced, they sound clean but not too lean. You won't get exceptional bass with these caps, but if it sounds "soft", then you may want to give the caps a few more days to run in. How do they compare to the Wimas so far?

Edit: I think the ebay seller is exaggerating, IME, the MKP2000 takes a bit longer than three days to mature. How about the treble? I recall that it sounds a bit harsh freshly installed.
Just one more question. In your past experience, how big of a difference does a "dual mono" (separate transformers and supplies) output stage make?

Cheers,

squeezethatbuffalo
24-09-2010, 23:43
Sorry you are having issues getting it all going - does anything ever work first time?!

I don't know your level of technical experience, but if you like I can give you some pointers for fault finding the regulator. First thing .... does the reg give -8V when NOT connected to the DAC, and is anything getting hot?

Hi Dan,

I don't have much (if any) technical experience :lol: Thank you for offering help!

Yes, the regs gave -8v even when not connected to the DAC. IIRC, nothing got dangerously hot on the faulty reg, at least not noticeably hotter than the functioning one.

Cheers,

Spirit
25-09-2010, 05:15
Expect my impressions soon®.
<snip>
*soon® is a trademark of Twisted Pear Audio, registered in numerous countries :lol:

:lol:
Oh so true.

I'm (im)patiently waiting for their USB transport...
(After (im)patiently waiting for Buff2 ;p)

Marco
25-09-2010, 05:20
Hi 'Spirit',

Welcome to AOS :)

Could you please pop into the Welcome area and introduce yourself to our community by supplying your first name, basic geographic location, system details and music tastes, as this is the requirement for all new members joining AOS.

Cheers! :cool:

Marco.

Dan Giovanni
25-09-2010, 16:01
Hi Dan,

I don't have much (if any) technical experience :lol: Thank you for offering help!

Yes, the regs gave -8v even when not connected to the DAC. IIRC, nothing got dangerously hot on the faulty reg, at least not noticeably hotter than the functioning one.

Cheers,

OK, next is the raw supply the same on both sides? Also, the voltage across C16 should be the same as that across C6. Ditto R5 and R15, and ZD1 and ZD2.

squeezethatbuffalo
02-10-2010, 06:24
OK, next is the raw supply the same on both sides? Also, the voltage across C16 should be the same as that across C6. Ditto R5 and R15, and ZD1 and ZD2.

Hi Dan,

Sorry for the late reply, I just came back from a trip ;)

Since I didn't have any spare parts at hand, I decided to send them back to UCC for repairs. Turns out that the 2SK170GR and LM336 5V were broken. Repairs incl. shipping were free. I guess that is why they don't sell kits/modules overseas :lol:

Edit: Have you had the opportunity to listen to the Legato+Jensens? I hope that the transformers sound better with a proper I/V stage. They do in my case!

Cheers,

squeezethatbuffalo
02-10-2010, 06:25
:lol:
Oh so true.

I'm (im)patiently waiting for their USB transport...
(After (im)patiently waiting for Buff2 ;p)

Heh, I know how you feel. By the way, do you happen to have the same ID on another UK audio forum?

squeezethatbuffalo
02-10-2010, 06:51
Finally got everything to work properly, whew! :eyebrows:

BufII is powered by Paul Hynes Z1P5v45, Volumite by a separate LT1086 5v supply, and each channel of the Legato gets its own Golden Reference bipolar supply+transformer.

Legato is mostly stock, I switched the FW caps to BC037s, and used MKP2000 3.3uf over Muse ES. The Lundahl transformers are hidden under the Legato board. I tried the compact Wima 6.8uf caps, and they actually sounded fine. However, the bass was a bit bloated and muddy in my system, so they had to go. The MKP2000s are still running in, but they already sound better to my ears. More transparent sound, better separation - maybe it's the "tidiness" Leo talked about - and "normal" bass (probably due to lower capacitance though).

Although I like my old Buffalo alot, the BufII is obviously in a different class. I'd like to compare the BufII+Legato to much pricier Sabre 9018 DACs, such as the DAC202 and Moon 750D. Now I just need find someone who owns one of those DACs in Taiwan :lol:

Oh, almost forgot about the pic! Here ya go, and sorry again about the iPhone quality picture. I'll update a better one soon® (lol, can't help it).

http://img833.imageshack.us/img833/9376/img0090se.jpg

Cheers,

Ali Tait
02-10-2010, 07:23
Looks good Alex. I've been looking at these,and also the Rakk dac. That gets pricy if you opt for the valve output stage though.

squeezethatbuffalo
02-10-2010, 07:46
Thanks Ali! As you know, my friend is going to make me a VC-less valve pre, and I'm not sure I want that many tubes in my system. Okay, the main reason is that I probably need a new, taller case to fit the valve output stage, and I'm too lazy. :lol:

Edit: I just realized that you're talking about Rakk's active output stage, not the BufII valve output stage. Silly me!

leo
02-10-2010, 08:39
Coming on nicely Alex, well done :smoking: Glad to see you got the faulty reg sorted too

squeezethatbuffalo
02-10-2010, 10:28
Coming on nicely Alex, well done :smoking: Glad to see you got the faulty reg sorted too

Hi Leo,

Can't think of praises to sing about the Buffalo II, other than "it didn't disappoint". I did have very, very high expectations though ;)

For lack of a better way to put it, I'm just happy with my new Buffalo. It's the first piece of equipment that has allowed me to simply enjoy music, and not worry about further modifications/upgrades at all. Very unaudiophile-like! :lol:

Btw, how do the MKP2000s sound on your BufII after more run in? Do you like them?

Cheers,

leo
02-10-2010, 17:30
Hi Alex,

Mission accomplished if your happy ;)

The MKP2000's seem to be improving slightly after continuous running in , they are still a little light in the bass but quite sweet in the high's. For such a low price they are very good, certainly better than the Wima's

squeezethatbuffalo
02-10-2010, 17:55
Hi Leo,

I was doing my best to not sound like a complete fanboy. But I guess it was way too obvious. Mission accomplished indeed :lol:

I chose the MKP2000s because they're the "best" compromise I can find. There are many better sounding caps out there, but those that I'm aware of are either pricey or sized and shaped like a coke can (often both). :scratch:

Glad that they sounded okay to you!

Cheers,

needsp
08-10-2010, 07:13
Hi

These might be a good alternative, and they're a similar size to the MKP2000's-

Obbligato Premium 10.0uF (250V)

http://www.diyhifisupply.com/node/733

I've had excellent results with Obbligato Gold Premium caps in my valve amps, speaker crossovers and lower voltage gear too. (For example, I much preferred them to Hovland Musicaps, despite the latter's film and foil construction). So have many other people! All of these caps are polypropylene types, but the 10u/250V is physically much smaller, as it's rated at 250V, not 630V.

Hope this helps

Paul

Marco
08-10-2010, 08:01
Hi Paul,

Welcome to AOS :)

Could you please pop into the Welcome area and introduce yourself to our community, tell us what system you've got and about your music tastes, etc, as this is the requirement for all new members joining AOS.

Cheers! :cool:

Marco.

domo
10-10-2010, 00:16
Just running in Legato output stage, they use discrete for the differential and op-amp for the differential to line, of course you can easily run it discrete out for balanced and run it into say a transformer converting it to SE if you don't want to use the op-amp.
I've already briefly tried Legato I knocked up on vero but will run in the proper one before commenting on the sound :)

http://i273.photobucket.com/albums/jj220/kingbusoms/P1090066.jpg

Hi,

I just joined here, mainly because of this thread. I ordered a Legato a month or so back but it looks like the package has been lost, hopefully I should get a replacement this week, or next, and am looking forward to trying a few things - Jensen JT-11-P1 transformers for Bal/SE conversion, upgraded coupling caps, paralleled transistors, etc.

I like the use of the SIL sockets for capacitor swapping/comparisons, great idea I think I will do the same. I see you are also using the Hynes sregs, how do you think they compare to the standard avcc reg. And have you also tried/compared them to any of the other 3pin regulator replacements that are available such the Tentlabs shunt regs, DEXA series regs, etc?

Marco
10-10-2010, 18:56
Hi 'domo',

Welcome to AOS :)

Could you please pop into the Welcome area and introduce yourself to our community by supplying your first name, system details and music tastes, as this is the requirement for all new members joining AOS.

Cheers! :cool:

Marco.

Spirit
22-10-2010, 06:33
Alex - I presume you mean RG, in which case, yes ;)

Marco - Will do, shortly.