OPAMP Comparison in DAC Analog Ouput Stage

Tweakers' Asylum

Tweaks for systems, rooms and Do It Yourself (DIY) help. FAQ.

dtm1962

I have tried the following OPAMPS in the Analog Output of my AA DDE1.2 DAC: OP275, OP249, OPA2132, AD826, OPA627AU, AD8620, AD825 & AD8610.
And the winner is (actually are) AD825 as the I/V Converter & AD8610
as the Buffer/Output Stage.
Case History:
The OP275GP (original) sounded OK but hazy & uninteresting with an average soundstage.
The OP249GP actually sounded strange especially the midrange ( maybe some oscillation going on )
The OPA2132P was a definite improvement over the OP275. It has a smoother refined sound and deeper soundstage.
The AD826 was added an the results were totally opposite with respect to the OP2132P. The sound was in-your-face aggressive with
a very wide soundstage but much less depth.It was more exciting but
listener fatigue would set in quickly. I added Class A Biasing with resistors and it seemed to tame it a little bit.
Next up was the SOIC version of OPA627AU on BrownDog Adapter and the results were an immense and deep soundstage relative to the AD826.
It was smoother , but had a siblilance problem. My ProAC 2.5 Clones have a little bit of sibilance as well , so situation made it worse.
I am assuming since I have a OPA627 for the I/V Converter as well as the Output buffer, that having 2 OPA627's compounded the sibilance problem even further.
I added Negative Biasing via resistors , then CRDs, then CRDs with resistors seemed to make it somewhat better.
Next up was the AD8620 and the sound was more forward with wider soundstage, a sound very similar to the AD826. I added negative biasing with crd's and resistors, but did not notice any improvements. I then added positive biasing again with CRDs & resistors and noticed somewhat of improvement, but I still noticed I was suffering from listeners fatigue. I still wasn't happy.
In goes the OPA627's again and it brought much less listeners fatigue, but that sibilance still bugged me and reminded me that
it was a stereo equipment I was listening instead of real music.
In my last bid I tried the AD825 as the I/V and the AD8610 as the output buffer/stage (WITHOUT Class A biasing). After the first song
with cold OPAMP chips & on-board unbroken-in OSCON decoupling caps,
the sound was both smooth ( BB charteristic)& had a certain clarity (AD characteristic)and was exciting to boot. It offered pretty good depth and width with a good sense of ambience and low level details.
I added Cascoded FETs with 560 ohm resistor and the sound improved slightly in all categories.
I assume it will improve even more in the coming weeks as the OPAMPs/Fets & OSCON caps break in.
I believe that this combination of AD825 & AD8610 work so well because the AD825 has very quick settling time (2V to 0.01 % at less
than 100nS) which is required in the I/V Converter (Ignoring the feedback loop). The AD1862 DAC Ouput has a settling time of 350nS. Therefore I believe you would want a I/V Converter faster than the DAC feeding it.
The second reason why it works so well may be the fact that if an OPAMP has a weakness (i.e. forward sound)and you use 2 of them in series (I/V Converter & Output Buffer) then the problem is compounded/multiplied and is made worse. When you use 2 different Opamps devices in this case in series, then the sound may be balanced ( i.e. forward sound from one opamp balanced by a smoother sound from the second opamp etc.)
Any comments welcomed!
Dan

This post is made possible by the generous support of people like you and our sponsors:

WEET Music Caps

Topic - OPAMP Comparison in DAC Analog Ouput Stage - dtm1962 18:00:57 04/26/03 (0)

You can not post to an archived thread.