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

[Henmyr]

Originally Posted by RFL

Just FIY, if the LT1028 becomes hot, it might be oscillating and this may compromise its audio performance somewhat. With hot I mean burning your finger if you keep it upon the opamp.

I will most definitely burn my fingers if I touch LT1028, it is very hot. It does not make the LT1469 in the amp very hot though.

[ilgello]

Originally Posted by RFL

Sounds like the Beresford has been a great buy for you. How's its headphone amp?

Beresford Amp is something that you don't want to listen really poor to my ears (hd600, never tried it with my 325i), but I guess is the DAC you have to look at.....

[Henmyr]

Originally Posted by RFL

No good. I would bet it's oscillating; but perhaps of a different kind of oscillation that doesn't case the amp's opamps to oscillate too.

I suspect that those filtering capacitors around the opamp in the DAC board have nonlinearities at high frequency that are causing the faster opamps to oscillate. In fact they're just normal polyester capacitors - not the best choice but I guess it works with the stock opamp and similar ones.

What can happen if it is oscillating?

[OvidiuDanut]

All righy than. I finished all 182 pages and my conclusion is this:
Till i'll save for the Opera, i'll just have to try this cheap or should i say "inexpensive" sollution. Hopefully i'll get a flawless functioning unit
Thanks for the thread!

P.S - Thanks BadaBing and Penchum.

[Fremen]

Originally Posted by RFL

No good. I would bet it's oscillating; but perhaps of a different kind of oscillation that doesn't case the amp's opamps to oscillate too.

I suspect that those filtering capacitors around the opamp in the DAC board have nonlinearities at high frequency that are causing the faster opamps to oscillate. In fact they're just normal polyester capacitors - not the best choice but I guess it works with the stock opamp and similar ones.

It's also my guess. I'm working on it. I have ordered WIMA caps and will try direct decoupling of the DAC opamp.

I have done some "reverse engineering" on this DAC LPF circuit. In fact it's almost based on the typical application circuit described in the AD1852 datasheet. Two modifications have been made to this standard circuit.(plus the fact that the resistors have been rounded to the nearest standard value)
The first is the use of four electrolytic 47µF caps to isolate the AD1852 from the LPF op amp(2 per channel). It's good from a electrical point of view (no DC offset transmitted by the DAC) but not that good musically speaking (I will try later on to bypass them).
The second one is a different choice of 2 caps and a resistor in each channel of the filter (C9, C10 and R20 for the left channel on datasheet). This leads to a non linear group delay response in the audio frequency range (C9<>C10) and a reduction of the frequencies above 6khz.
With R20=10ohms, C9 and C10 should be 470pF to have the same response than the LP filter in the AD1852 datasheet. I will try it too when I receive my caps order.(of course on both channels)

The funny thing is that I found the schematic of the zhaolu here (the one with the AD1852) and even if the circuits are different, they have exactly the same behaviour in frequency and in group delay, plus the same 47µF caps mod... cousins or brothers?

[Fremen]

Originally Posted by RFL

Well... the design similarity with the AD1852 Zhaolu was pretty apparent throughout; so I'm not surprised at the discovery..

Yes that's why I simulated the zhaolu LPF too

Originally Posted by RFL

That asymmetrical filter is somewhat strange to me. I don't have the knowledge to judge why this choice.

I was wondering about some trick to remind some tube sound but I don't know much about tubes..
From what I remember from Bessel design filters, they are meant to keep constant phase and group delay...not to tweak them

Originally Posted by RFL

What I clearly see is that in the Zero the 560pf cap is of a different brand and perhaps dielectric from the 1nf one

The 560pF caps are made of polypropylen which is better than polyester like the 1nF yellow one, at least in the audio path.
I will replace all the caps in the LPF with WIMA FKP2 - 2.5% ones. They are in polypropylen and cost 0.2€ each, so not a big deal (total cost 2€)

Originally Posted by RFL

I had already realized that there were those electrolytics to block the DC offset between the AD1852 and the opamp. You can bypass them, but from what I've read somewhere you're going to find yourself with about 50 mV of DC offset on both channels. This is not a problem for the internal headphone amp since it has its own decoupling caps, again electrolytic (probably Elna Silmic II ?).

You are right. In fact I just listen to the Zero with my HD-650.

If I have some time I will do the same with the headphone amp board.(unless some courageous folks already looked up and down and up and down and ... to draw the circuitry )

[bada bing]

Originally Posted by Fremen

Yes that's why I simulated the zhaolu LPF too




I was wondering about some trick to remind some tube sound but I don't know much about tubes..
From what I remember from Bessel design filters, they are meant to keep constant phase and group delay...not to tweak them



The 560pF caps are made of polypropylen which is better than polyester like the 1nF yellow one, at least in the audio path.
I will replace all the caps in the LPF with WIMA FKP2 - 2.5% ones. They are in polypropylen and cost 0.2€ each, so not a big deal (total cost 2€)



You are right. In fact I just listen to the Zero with my HD-650.

If I have some time I will do the same with the headphone amp board.(unless some courageous folks already looked up and down and up and down and ... to draw the circuitry )

Awesome info, thanks for the work.

Besides the potential for opamp oscillation, one thing I have been wondering about is the bias on the final stage discrete transistors in the headphone amp. I haven't pulled the board out yet to trace it, but it appears like it has a pretty hefty bias current going through the final stage. The transistors stay pretty hot. Evidently set by the choice of fixed resistors. It could be a potential mod to put some trimmers in the amp section and fine tune the bias. Mine are definately pulling more current than needed for stable class A just going by the heat generated.

Also, there is no reason for more than one set of DC blocking caps. One stage is required by the offset inherent in the 1852, but two blocking stages is redundant, and electrolytics at that.

The sound of the Zero is pretty impressive, especially considering a couple of the design limitations. I bet for a few bucks in parts it can be made to really sing .

[Clok]

Anyone who has an order in heard anything from Lawrence?

[fl00r]

Originally Posted by Clok

Anyone who has an order in heard anything from Lawrence?

It's Chinese New Years holiday. As far as I know until February 22.

Bad timing to place an order

[Clok]

Oh, ok.

[fault151]

has anyone heard the zhalou dac? Im thinking of getting one but not sure which version is best? How do these compare the the zero? I have heard you need to do some upgrades to get the best out the zhalou?

[pincellone]

Originally Posted by fl00r

It's Chinese New Years holiday. As far as I know until February 22.

Bad timing to place an order

Actually, the New Year Holidays are already over. China is up and running since 16th February...

[fl00r]

Originally Posted by Clok

Anyone who has an order in heard anything from Lawrence?

I just sent Laurence an email and he replied:

'The transportation is improved in Mainland China after the natural disaster of snowing . The next stock will arrive next week and I'll run in for one day and ship them out.'

[dsmudger]

OK been away a couple of days so behind on reading the last few pages.. but just to say, I finally pulled the lever and ordered six OPA627s in metal TCO 99 package

I've just stuck two of them in the DAC leaving my LME49720s in the headamp section - and I do have to say I've had my socks blown clean off!

Bass is bigger than anything else I've tried so far (finally where I want it, nicely making up for the brightness of my Grados), nice integrated sound and not at the expense of detail.

Will try an all-627 setup later on, but wow - these 627s are a winner

[OvidiuDanut]

What's the ebay nick of the producer?