OPA627 vs AD8610

[Dr.H]

Substituted the OPA627, biased into class A with JFETS (2n3819 with 2SK170) with an unbiased AD8610 in my pavel macura style pre-amp (OPA627 into BUF634) which feeds a NIGC). After 24 hour run-in, I could not really tell the difference in sound. Maybe biasing the AD8610 will shift it to another level?

Note that I am using a Theta dac with OA627 in both IV and filter positions. Also note that I am runningthe AD8610 at +-15V and it is well.

Time to try the biasing or AD825/AD8065....

[5th element]

Not to squander this class A bias thingy, but those opamps your talking about have such low distortion anyway biasing into class A aint really gonna do anything more then make an unhearable thing even more unhearable if you get me lol.

Plus doesnt the buffer have greater distorsion then the opamps anyway, so the low distortion of the opamps is going to be swamped by the higher dist of the buffer. So putting it into class A isnt really going to do anything. Or have I missed the point of class A biasing in opamps?

[analog_sa]

Time to try the biasing or AD825/AD8065

The difference between 627 and 8610 is very substantial. The difference with biasing- almost imperceptible, especially on the 8610. My ears, of course

[Dr.H]

Analog SA,

So you agree with my observation then: A biased OPA627 and unbiased AD8610 sound similar. And also, biasing the AD8610 leads to very little improvement over the unbiased AD8610. So to conclude, dont buy the OPA627 (which needs biasing to sound good-my observation as well, dark and compressed without biasing), rather buy the AD8610, unless of course you have serious allergies to SMT.

Any thoughts on AD8650? I'll be trying those tonight.

[analog_sa]

So you agree with my observation then: A biased OPA627 and unbiased AD8610 sound similar

I haven't really done this comparison. Biasing never really made a huge difference to me and some express the opinion that 8610 benefits more from positive rail biasing which i never tried.

[5th element]

Originally posted by Dr.H
dark and compressed without biasing

Thats most certainly not the conclusion I am at, with them performing filter duties in my active xover on the midrange, they also do I/V conversion in my DAC. Compressed it not a word I would use to describe the sound.

[Dr.H]

5th Element,

Have you tried a JFET bias yet, at about 2mA? I think you'll find that as good as teh OPA627 sounds unbiased, the sound will open up quite a bit with biasing. I guess it's all relative-OPA627 sounds good, better with biasing
Anyone else hear this?

[carlosfm]

Alone they are quite different.
P-A also said that with the BUF634 he didn't notice any difference between those two.
I only compared alone, in this case the AD8620 against the OPA2132, on a preamp (buffer input of my NI LM4780 GC).
I've put back the OPA2132, much better.
But that was at +/-15v and people say that the new ADs work better at +/-12v (+/-15 is over the top for them anyway).

I also tried AD8066, it worked well, but not to my taste, and I had to get it out, that thing got quite warm.
Don't try this one at +/-15v!


I only have initial impressions on these AD op-amps.
Testing conditions were not the ideal for them.
Someday I'll make a better test.

[fmak]

Time to try the biasing or AD825/AD8065.... [/B][/QUOTE]
------------------------------------------------------------------------------------
There are quite a few posts on biasing in the headphone forum. Biasing the AD825 to 3 mA with fets is certainly worthwhile - tames the slightly splashy treble.

I find the 8065 a little mechanical, too close to the 627 for my liking. The LM 6171 suitably bypassed bewteen V+ and V- also sound s very good without biasing!

[Dr.H]

Well, listened again last night to AD8610 (unbiased) vs OPA627 (biased into Class A, 2mA, with a JFET) and there is a difference.

The AD8610 seems to present a more open window, especially in the highs, on music, transients and realism of highs being enhanced.
I would say though that on some CD's it can come across as a bit bright. I will be bypassing using the Goudreau triplet and will post results here.