Discrete analog stages for multichannel players

[Steve Goff]

Some of the new multichannel combo (DVD-Audio and SACD) players have pretty good basic power supplies and DACs, but are strewn with cheap Op-Amps for current-to-voltage conversion, filtering, and amplification, sometimes with crappy electrolytic as coupling capacitors. I'd like to come up with a simple, discrete analog circuit to replace these, one that I could relicate six times on a PCB, which would also provide local regulation (probably also discrete, or a feedback design using an IC and a pass transistor).

So far this is only a thought, but I'd appreciate what others have to say on the subject. I know that there are folks out there doing modifications of some of these players, but they mostly amount to parts swaps, which are OK but don't do what I think is needed.




Steve Goff

[Chris D]

Steve-

I've never heard of people doing this actually, so I'll be most interested to hear what you see other people doing, and what you end up doing yourself. If you do anything like this, let me know, and I'd love to swing by your place and check it out in person.




CHRIS
Luke: "Hey, I'm not such a bad pilot myself, you know"

[TacoD]

Here comes my first post , (reading these forums some time now).

Some time back I came across this site, (very informative). This fellow has built a discrete class A stage for his SACD-player.

http://www.ultranalog.com/cdenhancer/sacdenhancer/

[Victor]

I read the materiel in the link provided for proposed improvement for the SAC-D player analog output. I must say that I am less then impressed with the solution. In fact, the solution is much worst in my view then the original circuit. The author proposes a 6 transistor single-ended circuit with no feedback that will most definitely have significantly greater distortion then the original op-amp based circuit. Well, I am not all that impressed with the original circuit either, but given the choice I would leave the original circuit in place. In the article there is a great deal of discussion about the ‘sins’ of the op-amp sound, - this is off course nonsense. The following quote says it all.

The opamp impresses with low distortion and low power dissipation, which to a certain extent compromises audio quality.

This is really rich!

The losses lie in crossover distortion and intermodulation due to its class-B design. The reason why you don't see this in the measured distortion (on 1kHz sine waves) is because of the high amount of global feedback.

So what is wrong with high amount of global feedback? You may prefer, and some do, a no feedback design; but, in principle, when the feedback is used properly, and it is in the case of this Sony SAC-D player, there is nothing wrong with it.

From where I stand it appears to me that the proposed ‘cure’ is much worst then the ‘disease’.

I built a few 2-channel outboard D/A converters in the past. I could achieve better then 18-bits resolution, which is still as good as it gets these days. In principle I can scale the design to 6-channels.

Here is what I would do, the I/V stage is not required since those D/A most likely output voltage in the balanced fashion. I would design a passive 6 pole balanced Bessel filter with -3 dB point of about 35 kHz and at most 1 dB loss at 20 kHz, followed by a balanced 2-op-amp driver with a bit of gain and that is it. It is not all that difficult.

I would then replicate the circuit 6 times on its own PCB with its own discrete feedback-type voltage regulator. Finally, I would either endeavor to put the PCB inside or leave it in its own enclosure and ‘borrow’ the power from the SAC-D player.

[Chris D]

Sssssssssssshhhhhhhhhooooooooooooooooooo......

Right over my head. Yikes. Definitely not something for me to mess with, only having had one EE class in college. I'll leave that sort of stuff to the EE experts, it looks like.




CHRIS
Luke: "Hey, I'm not such a bad pilot myself, you know"

[JonMarsh]

I have some thoughts about this, but I'm in the middle of the IEEE APEC 2004 conference, and really short on time- I'll get back to this around the end of the week. I do favor non loop feedback circuits myself, but there are lot of ways to do those, and there's no reason for noise or distortion to be high... just look at Ayre's digital products, for example, or Theta Digital.

The final test is listening to the circuit, and in my own experience not using loop feedback results in more natural, subtlely detailed sound, with more nuance.

But I wouldn't discourage anyone from trying lots of other approaches- I already have, too.

There's a bit of a raging debate going on at DIYAUDIO.COM on this same topic. JC, CH, NP, HD, a few others and myself on one side, and a variety of guys on the other. Might be interesting reading.

I have been working on my own DAC designs and circuits for fun; since one of them has been judged by some of my friends on a fairly nice system as significantly better than $3000 Sony SACD players on CD, and the CD layer rivaling the SACD performance (better in many areas, not quite as good in one area), I'm inclined to say that in my experience there's some potential validity to this as a workable design philosophy.

YMMV.

~Jon

Best regards,




Earth First!
_______________________________
We'll screw up the other planets later....