Review: ZERO 24 BIT/192KHz DAC/Headphone Amp/Pre-Amp

[Ash]

Originally Posted by Penchum

Hi, I guess I answered another of your posts in a different thread just a little while ago! Changing Opamps to customize the output is very easy. They are just a chip in a socket. You pull out the old and push in the new. There is enough information in this thread to get anyone started, if they want too.

See my other post also. See ya!!

Penchum,
Thanks for all the very useful advice and tips. I was under the impression that I would be soldering little capacitor like things when swapping Op-amps. But I guess - what u describe is very easy and doable.
Regards...

[Penchum]

Originally Posted by Ash

Penchum,
Thanks for all the very useful advice and tips. I was under the impression that I would be soldering little capacitor like things when swapping Op-amps. But I guess - what u describe is very easy and doable.
Regards...

Some of the guys here are doing serious mods to their Zero's. The rest of us, are only swapping Opamps. Not to worry, it is very simple and disgustingly fun.

[Fremen]

Previously on "Zero mod" :
- C9, C10 (560pf, 1nF) replaced by 2x470pF, R20 unchanged(10ohms) -> same reponse than the AD1852 datasheet (2x220pF, 200ohms)
- All the other caps of the LPF replaced by polypropylene caps, same value

Is the asymmetric filter the secret of the zero sound? After 2 hours listening to the zero with this mod, I think so.
The highs are lounder and very clean but the sound is not warm anymore. The zero signature is gone.
Another consequence: as the medium and highs are louder, I have to reduce the volume which reduces the basses too ...

So I resoldered the 2x560pf caps I removed on the 2x470pf (C10,C12) having then C9=470pF and C10=1,03nF (well, not good practice but it's temporary)and than...the zero sound was back, but not only: I think it's coming from the new polypropylene caps: the sound seems now more precise than before, in the all range.

If you want to build a filter, follow the rules, if you want a good sound, don't

[Fremen]

Originally Posted by RFL

Congrats! Seems like a fine job.


Me, however, I want to stay "safe" and follow the datasheet indications. So I'll keep the same 1 nF values and just put in the 220 pF ones and 200 ohm resistors. I like the "analog" sound in digital stuff, which the stock Zero has.

Also with the 200 ohm resistors the better opamps shouldn't oscillate anymore, thus they should reveal all their potential.

Do you mean C9=220pF, c10=1nF and R20=200ohms? If yes, you will completely change the sound of the zero.
From what I just saw with my mod, it seems to be better to follow the behaviour of the existing filter.
C9=C10=220pF and R20=200ohms will give the same result I had with C9=C10=470pF and R20=10ohms.

If you want to increase R20 around 200ohms but keep the zero signature, I suggest you try C9=220pF, C10=680pF which leads you to have R20=240ohms. It will give you the same response in frequency and the same trick in the group delay than the original zero.

EDIT: a more tricky one but closer to 200ohms for R20: C9=330pF, C10=780pF(560pf+ 220pF downside) and R20=220ohms

[Penchum]

Originally Posted by RFL

Well, after your experience I guess I don't want to touch the LPF anymore. The only thing I could still do might be changing all the polyester caps with WIMAs of the same value - but 560pF is a difficult value for me to find in polypropylene.

So probably I'll just leave the LPF as it is. What I must do in my Zero is replace the big filter caps, two of which I have damaged out of...a hysterical attack. I'll use some Panasonic FM and FC. With the occasion I'll probably replace the 100uF 25V electrolytics near the DAC chip & digital receiver with some 220uF 10V OSCONs that I have.


Btw, thanks!

You have those attacks too!!

Man, don't you hate it when it happens!

[Fremen]

Originally Posted by bada bing

Did you do anything with the 47nF electrolytics on the DAC output ? They are outside of the area pictured. I wonder about those 47nF caps. The DAC application recommended circuit doesn't include any DC blocking at all. It appears like offset might not be a problem after the balanced to SE conversion ? Did you consider temporarily shorting the two 47nF caps and measuring the offset at the filter output ?

47nF are pretty big honky dirty electrolytics on a high impedance, low current line. I'm curious what the rational was ? I'm considering as RFL said - ordering hiQ film caps and implimenting the data sheet circuit exactly - sans 47nF electrolytic blocking caps.

I didn't shorten the 47µF caps for the moment because if I make too many mods, I won't be able to follow and understand (as if I could )what would happen at the end. But I intend to try it later on.

[Fremen]

Originally Posted by RFL

Well, after your experience I guess I don't want to touch the LPF anymore. The only thing I could still do might be changing all the polyester caps with WIMAs of the same value - but 560pF is a difficult value for me to find in polypropylene.

Yes it's worth doing it. As for the 560pF caps they are are already in polypropylene

Originally Posted by RFL

So probably I'll just leave the LPF as it is.

Well... 220pF, 680pf and a 240 Ohms resistors are not that difficult to find.
May be I'll try it this week-end ...

[Fremen]

Originally Posted by RFL

Maybe not when it happens...but after.

I know exactly what you mean

[Fremen]

Originally Posted by RFL

It's that that 240 ohm in series followed by 10nF in parallel seems a bit...violent, to me. It may just be a psychological thing (since that's what we were at).

I agree. The intensity through it would increase from less than 1mA to about 7 mA at 10kHz and the tension on it from nearly nothing to 600mV at the same frequency...

[Fremen]

Originally Posted by RFL

I think that changing just the 10uF parallel caps (after the 10 ohm R) to polyprolylene should yield much of the benefit you got from changing all of the filter caps to polyprop.

If you mean the 10nF caps, I already changed them for polypropylene ones. they are the biggest ones on top.

[Fremen]

I have a pair of OPA627 on the way home (could not resist anymore ) and some THS4032 too. Does anyone have feedback on those THS?

Edit: Are the OPA627s hot in the DAC?

[Penchum]

Originally Posted by RFL

The full bodied sound seems to do good to Tom Waits's Blood Money. I like the full and textured midrange.

I suspect it gets that from the LT1364? Are you setup, where you can do a quicky comparison between the two units? Ow, and how much was it?
Thanks!

[Fremen]

Originally Posted by RFL

I also have and am curious about the THS4032. However I would not dare to try it in the Zero, since it's only stable at gains of 2 or higher... And then there's that 10000pf capacitive load thingy

The THS4032 is supposed to be stable with gain=2 or -1. In our case, the LPF has a gain of -1.1, so it should be OK.
What is the 10000pf capacitive load thingy ?

[Penchum]

Originally Posted by Fremen

I have a pair of OPA627 on the way home (could not resist anymore ) and some THS4032 too. Does anyone have feedback on those THS?

Edit: Are the OPA627s hot in the DAC?

They are "Hot" sounding. Just slightly warm as I recall. I still consider them to be the best sounding for the DAC output. Whether I run them into my vintage systems, or to my Super-T amp or to one of my headphone amps, the sound is fantastic! I just need to get another pair for Zero #2.

[pincellone]

Originally Posted by Allure

Penchum, you should try LT1364 + LME49720. In a headphone amp it's making my headphones sing like rarely happened.

I've reached the 100 hrs with my Zero and tried several opamp combinations. With my setup I have to say LT1364 + LME49720 is in fact the best combination you can get.

I'm using a Sennheiser HD-555 and a SB Audigy 2 ZS.

I've found the LME49720 to be more detailed then LT1469 in the AMP. The LT1364 in the DAC is unbeatable. A close competitor are a couple of OPA627BP on adaptor, but still it wins hands down.

BTW I also put an heatsink on each opamp, they can't do bad on them...