AD8620 vs. AD8066

[Crazy_D]

Please give the pro's and cons to these two great op-amps.
The first con of the AD8620 is the price, but if the performance is better, no problem.

Would it be wise to decouple the V+ and V- with about .16uf oil cap?? Any other values seem to work better?

The application will be for a/d inputs.


Thanks!

 

[peranders]

Would it be wise to decouple the V+ and V- with about .16uf oil cap?? Any other values seem to work better?

Not very, I'll suspect. You should use caps with very good HF properties. I recommend SMD 0805 100 nF/ 50V XR7 or other ceramic type. 100 nF/63 polyester will do also probably. Use physical small caps.

 

[krishu]

Hello,

as I read in the analog letter I got yesterday, the ad8066 hat a much higher slew rate than ad8620, the latter one has less DC offset. there are also differences in distortion and maximal output swing. Biut soundwise ...?

I soldered both onto IC sockets and will try/compare both on weekend. I will report on monday probably.

Cheers
Christian.

 

[krishu]

Hello,

I burned in the AD8066, too and compared it (quickly) to the AD8620, which has been burned in moire intensively. I liked the sound of the 8620 more, it seems to be closer to what I feel is music. AD8066 was more like my double OPA627

I will listen to both again and more extensively when I have the time to do so. I will report here only if my judgement changes.

Cheers
Christian.

 

[pinnocchio]

So, after 8 years of waiting, what is the verdict?

Ciao!
Do

 

[gk7]

I have three AD8620 (on adpters for DIP8) and (purely subjective of course) they are absolutely unsuited for audio. They measure excellent but sound like a worn out cassette tape.

 

[pinnocchio]

I have three AD8620 (on adpters for DIP8) and (purely subjective of course) they are absolutely unsuited for audio. They measure excellent but sound like a worn out cassette tape.

Did you try the AD797BRZ or the OPA627 in can

 

[stratus46]

I have three AD8620 (on adpters for DIP8) and (purely subjective of course) they are absolutely unsuited for audio. They measure excellent but sound like a worn out cassette tape.

I'd be interested to see test signals into your usage of the chips on a 100MHz scope. I believe you're missing some parasitics.

If Analog Devices says it's for high performance audio, you can take that to the bank but these amps are quite fast and will not be forgiving of DIP adapters, single layer boards and poor bypassing. These would be excellent on a 4 layer board with a ground plane on layer 2 with 'holes' in the plane below the chip(s) to keep capacitance down on the '-' inputs which tends to cause peaking in the response. I've used many ADI chip in commercial video equipment and they always do what they claim. ADI is always my first choice because you CAN trust them but ALWAYS read the data sheet carefully so you use it properly. If you're looking for 'jelly bean' types to swap out, the much slower AD711/712/713 will give you no problems but are not much different from TL071/72/74.

I did HD analog video processing before the digitizing and 4 layer boards are the ONLY way to go for multi MHz systems. Liberal power supply bypasses are a must as well as the 'holes' in the ground plane and guard rings on high impedance nodes. The ground plane will 'randomize' the return currents and behave better than any ground 'tree' or 'snake' on a 1 layer board. You can sort of do a plane on a 2 layer board but there's always a temptation to break up the plane with a power or signal trace. You will not have that problem with 4 layers as you have 3 signal layers. All the solder flux residue must be cleaned off as well.

G²

 

[marce]

Use middle layers for power ground planes, or better still 8 layers and run all digital signals as stripline between ground planes.
But with todays high rise time devices etc using sockets etc is not good. Have a look at the ever decreasing size of SMD components, this is to minimise package parasitics andhelp remove heat from the actual chip (QFN's BGA's etc).

 

[pinnocchio]

Guys, this is all extremely interesting but what about actual listening tests of op-amps in your favorite DAC?

I'd like to hear some comments about the following:

OPA1612
AD797BRZ
Dual OPA627 (can version)

and any other worthy opamps out there.


I know that implementation is important but for people that already have a DAC and are not designing it would be great to know.

Thanks
Do

 

[analog_sa]

I have three AD8620 (on adpters for DIP8) and (purely subjective of course) they are absolutely unsuited for audio. They measure excellent but sound like a worn out cassette tape.

Goodness. No experience with the double but the 8610 is by far my favourite audio opamp. Not to mention the incredibly low offset bonus.