[diditmyself]

Help me make the 2009 Zero a better DAC! I've posted some possible problems in the analog/filter section and the headphone section before. I've done some preliminary tests, and I find the DAC to sound better with passive output than with either modded or standard filter/DAC. I've tried OPA627, AD797, AD8599, AD8066, AD825, LM4562 amongst others in that position, all with their own flaws and virtues.

I've changed the electrolytics around the AD1852 to the very good Sanyo OS-CON SVP (and some SA/SC types).

One problem is this - when run with passive filter (or straight without filter) it's humming and buzzing like if the power section is unclean and/or there are ground loops or I don't know what I'm doing and built myself a ground issue I'm not aware of.

Another "problem" is the unused neg output from the DAC. I don't have any balanced amps so balanced output is not an option at the moment. Is there a better way to hook it up other than just letting the neg output stay unused?

I'll be posting some schematics and 'scope pictures later.

Links to my earlier posts:
The original schematic except that the "33k" resistors really is 100k and the relay isn't showed.

http://www.head-fi.org/forums/f6/hel...elated-430000/
http://www.head-fi.org/forums/f5/tia...ml#post5785062
http://www.head-fi.org/forums/f6/zer...3/#post5769531

 

[diditmyself]

It seems like Sigma-Delta DACs need a filter on the output. That means the Lampizator mod can't be trusted.

This is the passive filter I'm using:

This is how a 1 kHz square wave looks before the filter

... and this is how it looks after the filter

This is the simulated frequency and phase response

... and pulse

More to come.

 

[supertramp]

Don't go too far or you will remain without a DAC...

I would not change the filter/buffer topology at all...

If you're full of good will, I'd change those yellow filter caps around the opamp with Wima FKS2 (same size but better) or FKP or something good for the purpose. Then I'd bypass those caps in the signal path with small polyesters (like 470nF).

Finally I'd change the opamp with OPA2211 or LT1355 (in no particular order).

 

[Bonthouse]

Quote:

Originally Posted by supertramp /img/forum/go_quote.gif Don't go too far or you will remain without a DAC...

No guts, no glory!

A lot of pics and all, but did you notice improvements?
It looks like the sounds should be a lot cleaner and clearer with the filter in place.

 

[diditmyself]

Quote:

Originally Posted by supertramp /img/forum/go_quote.gif Don't go too far or you will remain without a DAC... I would not change the filter/buffer topology at all... If you're full of good will, I'd change those yellow filter caps around the opamp with Wima FKS2 (same size but better) or FKP or something good for the purpose. Then I'd bypass those caps in the signal path with small polyesters (like 470nF). Finally I'd change the opamp with OPA2211 or LT1355 (in no particular order).

IMO the filter is erroneous. The 1nF C1 doesn't make sense. As you can see in my other threads a simulated square wave looks really messed up. I've used a modified active filter with WIMA and Evox/Rifa pp caps. The electrolytic coupling caps are removed and I use Solen SCR pp caps at the output instead. Still there was the sound of opamps. It sounds better with the passive filter, so I was hoping to get it to work without hum and buzz. There's a spark of magic that's lost when using an opamp.

I'm not so clever but I think the use of differential input as when the active filter is used cancels common mode noise. If I don't find a way to make the DAC "cleaner", I guess I have to go back to using an active filter, but a proper one.

 

[diditmyself]

Quote:

Originally Posted by Bonthouse /img/forum/go_quote.gif No guts, no glory! A lot of pics and all, but did you notice improvements? It looks like the sounds should be a lot cleaner and clearer with the filter in place.

I think the sound is smoother with the filter, but it could be imagination.
It seems that bypassing everything like in the Lampizator mod is to operate the AD1852 faulty. As you can see there's a lot of high frequency garbage that goes straight through the amp to the phones or speakers. It can't be good.

 

[supertramp]

Quote:

Originally Posted by diditmyself /img/forum/go_quote.gif IMO the filter is erroneous. The 1nF C1 doesn't make sense. As you can see in my other threads a simulated square wave looks really messed up. I've used a modified active filter with WIMA and Evox/Rifa pp caps. The electrolytic coupling caps are removed and I use Solen SCR pp caps at the output instead. Still there was the sound of opamps. It sounds better with the passive filter, so I was hoping to get it to work without hum and buzz. There's a spark of magic that's lost when using an opamp. I'm not so clever but I think the use of differential input as when the active filter is used cancels common mode noise. If I don't find a way to make the DAC "cleaner", I guess I have to go back to using an active filter, but a proper one.

There's a mistake... C4 is 560 pF, not 56 pF (I had the Zero). I know that in AD's schematic it should be symmetrical to the 1 nF C1, but I remember well that a guy with the 1st generation Zero changed C4 with a 1 nF cap and the sound (he said) became cold and unmusical...

So maybe it's not that erroneous.

Remember that the Zero is a precise copy of the Zhaolu D1.3, so that's where this peculiar filter comes from.

Music is not made of square waves... not really.

Why not just try and change the opamp with one of those I suggested...



To me the Zero sounded good and natural (even with the OPA2604), why want to mess it up that way... I think it should suffice to improve parts quality.

 

[diditmyself]

Quote:

Originally Posted by supertramp /img/forum/go_quote.gif There's a mistake... C4 is 560 pF, not 56 pF (I had the Zero). I know that in AD's schematic it should be symmetrical to the 1 nF C1, but I remember well that a guy with the 1st generation Zero changed C4 with a 1 nF cap and the sound (he said) became cold and unmusical... So maybe it's not that erroneous. Remember that the Zero is a precise copy of the Zhaolu D1.3, so that's where this peculiar filter comes from. Music is not made of square waves... not really. Why not just try and change the opamp with one of those I suggested... To me the Zero sounded good and natural (even with the OPA2604), why want to mess it up that way... I think it should suffice to improve parts quality.

You're right regarding the caps. My mistake. 560 pF makes sense. I'm sorry to have besmirched it.

Here's the true schematic

Why mod it? The sound did not impress me. Stock it was just a teeny bit better than iRiver IHP120. I expected more from a full size DAC. With the passive output stage it's enjoyable. How does it fare against your other DACs.

 

[exhausted]

Quote:

Originally Posted by diditmyself /img/forum/go_quote.gif Why mod it? The sound did not impress me. Stock it was just a teeny bit better than iRiver IHP120. I expected more from a full size DAC. With the passive output stage it's enjoyable. How does it fare against your other DACs.

Well the stock Zero improved a little on my Cambridge Audio D300SE (with CS4338, 24/96), so I was happy enough for the money spent.


I remember it was a smooth and easy listen, but it had good detail too. Tonality was a bit less saturated than that of the Cambridge Audio perhaps, but there was an improvement in transient accuracy. Bass was a bit cloudy but deep.

My current DACs benefit by the CS4398's more monitor-like sound, more dynamic and crisp, and with very good tonal richness. The AD1852 seemed to sound more "dreamy", if you will, an interesting sound anyway (it was very good with Leonard Cohen for instance).