Single vs. Dual Op Amps: Pros and Cons?

[ezkcdude]

After "hijacking" the AD8620 thread, I was prompted to start this new one. I think the subject may be of interest to some. I am posting a schematic of a differential to single-ended I/V stage for a DAC (PCM1794). This is part of a design I am working on now. There is nothing original. I have "stolen" the I/V from ADI's "Data Conversion Handbook" (Fig. 6.70). Anyway, the issue is whether single or dual op amps are better. In this case, AD8610 vs. AD8620. Obviously, a dual is more convenient, but I would imagine that many of you will tell me single op amps are more "audiophilic". I'd like to hear opinions on both sides.

 

[ezkcdude]

Apparently, you can't have attachments in two different threads? Well, if you want to see the schematic, just find my post in the AD8620 thread.

Edit: I deleted it from the other thread, so here it is.

 

[00940]

as your opamp will deal with high frequency crap coming from the dac, you should consider reading the application note slyt190 from TI.
http://focus.ti.com/lit/an/slyt190/slyt190.pdf

 

[JahJahBinks]

Your best bet is to ask someone who has listened to both

 

[ezkcdude]

Quote:

Originally Posted by JahJahBinks Your best bet is to ask someone who has listened to both

Or better yet, just try it for myself

 

[JahJahBinks]

Right, because you can't describe the sound signature from the data sheet.

 

[ATAT]

There was a DIYAudio post on the inadequacy of Opamps that talked about this,

Possible advantages

better channelmatching -> thermal coupling between duals may allow for thermal coupling

crosstalk -> this should be obvious, having two circuits at HF so close to each other can cause parasitic crosstalk.

There were a few others, but you could look up the thread yourself

 

[bhjazz]

Quote:

Originally Posted by ezkcdude Or better yet, just try it for myself

You betcha! Go for it...then give us the juicy details! Yay!

 

[ppl]

The primary reason that a Single Op-Amp is desirable over a similar Dual is isolation between op-Amps. When looking at the data sheet on any Dual you will see a graph showing the separation between the two internal op amps and it is almost an exponential decline with frequency. Consider the AD8620 you are referring to (see Pg.9 Fig2) http://www.analog.com/UploadedFiles/...D8610_20_d.pdf
Consider that the AD8620 is one of the Better performers in this spec for dual Op amps. Notice the decline from 138 dB of separation at 10 KHz to about 122dB at 300 KHz.. By using two AD8610’s with each properly decoupled with at least an R/C network on the supply rails with C consisting of the Bypass capacitors you put on the op amp power supply pins you completely eliminate this interaction.

Another consideration that was more of a concern with older op amps is most used a shared biasing network for both channels resulting in even worse isolation than is provided by the AD8620.

Another issue of using dual Op amps is that some designers will have the output of one half driving the input of the second half as opposed to using one half in one channel and the other half of the dual in the remaining channel and simply use another dual to form the following stage or section. Hope this helps Ya