Zhaolu DAC - a good value DAC?

Status
Not open for further replies.
Jump to Latest
#83
Interesting comments on the AD1852 vs CS4398. I choose the AD version partly because so many people said this was a good combo with the Zapfilter, because it appealed to the cheapskate in me, and because as I indicated in a prior post I am fairly partial to AD op amps and dacs due to recent career experience.. 😉

I'm pretty pleased with the choice, and I will reiterate that it makes my PS Audio Ultralink II sound broken by comparison.. :xeye:
 
#84
I am a bit off topic here. I have been toying with the idea of converting the Zhaolu into a upsampling DAC using a asynchronous sample rate convertor (ASRC) like the SRC4190 or SRC4192.

One would need to break 3 copper traces between the CS8416 and AD1852 chips and wire in the ASRC. It should be a simple operation.

Unfortunately, I have sold away my Zhaolu 2.0 sometime back and I have no candidates to operate on.

You guys might want to try it 🙂
 
#86
The OPA637 is a decompensated version of the 627 and is not stable at unity gain. Minimum gain for stability is 5. In this case since the circuits are likely unity or near unity gain LPF's you can't use the OPA637.

The LM4562 has very good performance and would be a viable alternative - I'm waiting for mine to arrive.

Anyone interested in a discrete solution, I've got a couple of ideas..
(Plug and play specifically for the 2.5)
 
#87
OPA637

Hello Kevin,

Thank you for the answer. I guess one needs to use
unity stable opamps.

I hope to play around with the different opamps when
I have the courage to open up the D2.5. For now,
I'll run it stock.

BTW, what do you have in mind for a discrete solution?
That would be quite interesting.

Cheers,
Dennis
 
#88
Hi Dennis,
I'm working on an all discrete bjt filter circuit with balanced voltage inputs and a single unbalanced output. The design consists of several long tail pairs, and a current mirror based output. It's very simple and direct. For good cmrr I will be using supermatched transistors in the LTP's. (These are also extremely low noise types.) The LPF function will be implemented as two single pole networks between the stages. I am not sure yet whether I will go with a dc servo and avoid the use of a single coupling cap in the path. I think I can tolerate one coupling cap, but I'll think more about that.

I am designing it to use the Zhaolu 15V supplies and hoping to put it in a form factor that will bolt in more or less. It's intended specifically for this application.. Anyone who is interested let me know.

All of this is very tentative, but the concept looks pretty good so far in simulations.
 
#90
Hi Dennis,
I guess you'd need to be able to read the Chinese language owner's manual to know that with total certainty .. LOL

That said I believe these features allow the dac to used as a bare bones pre-amp in addition to its dac function.

I suspect some sort of passive mixing between the dac analog out and the phone input is provided. You can verify this hypothesis by plugging a tuner or similar into the phone input, and plugging a pair of headphones into the headphone jack.

Probably the phone out is actually just that, (parallel output to the front panel jack) allowing you to use the dac as a simple pre-amp with a power amp.

This is all hearsay/suspicion on my part.. I have verified.... Nothing YMMV

😀
 
#92
Epicurean which version of the Zhaolu dac are you listening to?
Click to expand...

Sorry for the late reply. I guess I had enjoyed some beverages before writing as I did.

I'm mostly listening with my headphones(Sennheiser HD650) connected to a PPAv2 headphone amplifier that is using OPA637s and a OPA627 for the ground channel.

My Zhaolu is a D2.0 from Eddie at www.diykits.com.hk with his blackgate mod and the CS4398 chip. I've bypassed the output capacitors, disconnected the headphone amplifier and current op-amps are 3 LM4562s which I got today and I'm liking them a lot so far. Before I had 3 OPA2107s which IMO were clearly better than the OPA2604s and OPA2134s I tried before. My Zhaolu originally came equipped with a DY2000 in the center and 2 LT1057s in the sides. I listened to it for a while but it didn't match my headphones very well, which are very sensitive to bass quality. Switching to the factory standard OPA2604s improved just about everything and putting in the OPA2107s was quite a dramatic improvement once again. The OPA2107 sounded like the OPA2604 but with better soundstage, clarity and cleaner highs(the OPA2604 could make the highs quite shrill at times, even with the Sennehiser HD650s which are known to be very smooth and non-fatiguing).

I have to admit I haven't even touched the LT1057 since I removed them. Perhaps I should give them another try to see if I still don't like them.
 
#93
Probably won't change your impression too much, first impressions are usually not too far off the mark.

I am still waiting for the LM4562 samples I requested about 2 weeks ago now. Several of my friends have since purchased the Zhaolu 2.5A based on my recommendation, and one of them has installed the LM4562 - reports a large improvement overall, although he also thought the LT1057 was a pretty reasonable choice to start with.

I'll report back when I get the chips.

My discrete solution is looking fairly good on paper, needs a little more work on large signal handling.. The AD1852 puts out 4Vrms FS (IIRC) which is rather hotter than I would like from the linearity perspective with discretes on a +/- 15V supply with no global feedback. So far in simulations 2nd harmonic is about 50dB down and 3rd about 70dB down at FS - the distortion appears to scale proportionately with level over the limited range I've looked at. Note that simulations are not entirely accurate, and a lot depends on the accuracy of the transistor models, so it could in reality be a few dB worse.
 
#98
Starting to get a little frustrated with the delays in getting my LM4562 samples. Thought I would get smart and order some from Digikey instead - no stock on dips.. Went to Arrow Electronics and it's the same deal. :xeye:

It appears to be a long lead time item for the moment..

SOIC are available in limited quantities, and could be used on Brown Dog adaptors..
 
Status
Not open for further replies.
Top Bottom