[carlmart]

Some opamps have very specific circuitry requirements otherwise they become unstable so you can't just blindly use them if you don't understand how the circuit they are in works..

OK, I agree with that. Very fast opamps can be tricky, for isntance

.

For instance some circuits use capacitors in the feedback loop to reduce or remove any gain on certain frequency ranges but if the opamp isn't stable i.e. isn't unity gain stable with that setup all kinds of undesirable things can happen even potentially the opamp killing itself..

Voltage feedback opamps can't have caps in feedback loop. Poorly designed pcbs can be tricky.

.

797 IIRC is pretty demanding on the placement and values of capacitors close to its power supply pins otherwise you tend to get a lot of problems with it especially if you are using it in a DIY fashion on prototype/breadboard, socket adapters or non-soldered.

Soldering bypass caps, at least 1uF if possible, directly on the chip supply pins and also bypassing between supply pins with another caps, should tame most (if not all) tricky opamps. Temp rise on the chip may be caused by oscillation. Walt Jung's superregulator was marginally stable when using the AD797, and there were experienced people working with it.

 

[Rroff]

Technically I guess the caps aren't in the feedback loop as their purpose is to act as a filter bypassing to ground rather than providing feedback. But yeah some opamps can be extremely picky about compensation capacitors and the like.

 

[leeperry]

Long story short you need some serious technical background to hear what 797 can truly offer, rolling it like a piece of lego is idiotic at best.

 

[Ivan TT]

I had misbehaving 797 in my DSD (although it worked just fine in Z3), both A and B grades, SOIC on adaptor, gain of 1: sweeping noises, RFI, overheating.

After trying everything suggested in datasheet, eventually solved the problem by adding RC filter tuned to about 150kHz on the opamps output (and buffer's input).

I disagree that 797's rolling is "idiotic" (or maybe just disagree with strong language used), in most cases it will work just fine, but sometimes needs minor tweaks to the circuit.

Strangely enough is 797 effectively biased in class A due to oscillation and I kind of liked the sound signature, it was a bit more vivid and involving.

 

[leeperry]

Scott Wurcer, inventor of AD797 made very clear that blind rolling it was not a smart idea, like not at all. We had a long thread about rolling opamps on diyaudio and the man is not kidding.

 

[carlmart]

Scott Wurcer, inventor of AD797 made very clear that blind rolling it was not a smart idea, like not at all. We had a long thread about rolling opamps on diyaudio and the man is not kidding.

Any final recommendation list on how to implement it? That would be useful.

 

[raoultrifan]

Perhaps the bypassing recommendations from https://www.analog.com/media/en/technical-documentation/data-sheets/AD797.pdf might help.

@Ivan TT, the low-pass filter was created at AD797 output? However, given the highly oscillations, I would probably remove it if I were you and replace it with something else; it just sounds safer to swap it for something else. BTW, I wonder how the ground was done there...or if there's any ground plane on the PCB.

 

[Ivan TT]

the low-pass filter was created at AD797 output?

Yes.

However, given the highly oscillations, I would probably remove it if I were you and replace it with something else; it just sounds safer to swap it for something else. BTW, I wonder how the ground was done there...or if there's any ground plane on the PCB.

Well, since I don't have oscillations anymore I don't need to remove it, it works and sounds just fine (although I have other opamps in my DSD atm, namely AD8065 and AD825 purely for synergy with my DIY IEMs).

As I mentioned before, I used every trick from AD797 datasheet, at some stage I had 4 bypass 100nf caps soldered directly onto AD797 power pins + 2 on DIP8 power terminals AND 50pf from pin 8 to output, made no difference whatsoever, sweeping noise and RFI and heat were exactly the same.

But Zishan DSD has it's processor literally touching opamp, so I'm not to surprised that AD797 was misbehaving as I don't think it's really RFI hardened and is sensitive to UHF noise.

RC filter on the output solved all of it and filter's cutoff is way outside of the audible range to be of concern and I used MELF resistors and PPS caps, so I'm quite happy, even if this approach is not in the datasheet.

 

[bogginhead]

Would I have to have two Muses03 if I wanted to use it in a Fiio E12 DIY since it's a single? Or could I just use one with some kind of adapter?

 

[Rroff]

Any final recommendation list on how to implement it? That would be useful.

Problem is it isn't always something that can be sorted by just adding or removing a few components - some of these opamps ideally require soldering with very short lead lengths to the components around them and attention paid to the implementation of the traces around them to avoid parasitics, etc.

The 797 has a fair amount of detail somewhere towards the end of the datasheet IIRC on best practises and so on but it isn't exhaustive. Bypassing the power supply directly on the pins with some small capacitors helps a lot.

 

[raoultrifan]

I'm successfully using 2 x AD797 in one of my Burson PLAY combo DAC/headmp units, in the pre-amplification stage (gain or voltage amplification), just before the output stage. I have no oscillation, no overheating or anything suspicious; had this tested with oscilloscope.
I have a "feeling" that PCB board layout, design and quality is with it counts more, of course...with enough decoupling around the op-amps, combined with a low-noise and very well regulated power supply.

 

[FritzS]

Yes.

Well, since I don't have oscillations anymore I don't need to remove it, it works and sounds just fine (although I have other opamps in my DSD atm, namely AD8065 and AD825 purely for synergy with my DIY IEMs).

As I mentioned before, I used every trick from AD797 datasheet, at some stage I had 4 bypass 100nf caps soldered directly onto AD797 power pins + 2 on DIP8 power terminals AND 50pf from pin 8 to output, made no difference whatsoever, sweeping noise and RFI and heat were exactly the same.

But Zishan DSD has it's processor literally touching opamp, so I'm not to surprised that AD797 was misbehaving as I don't think it's really RFI hardened and is sensitive to UHF noise.

RC filter on the output solved all of it and filter's cutoff is way outside of the audible range to be of concern and I used MELF resistors and PPS caps, so I'm quite happy, even if this approach is not in the datasheet.

I use Evox Rifa PHE 450 100nF near the socket and bypassed them with WIMA FKP2 10nF/63V
I don't use the recommend 50pf cap between pin 6 (out) and pin 8; I don't use the 20pf in NFB too.

I have no measure equipment at this time - but the sound result is very well! I hear with AKG K812 - this headphones don't forgive even the smallest of shortcomings.

 

[Nintendo Maniac]

I'm going to be ordering one of those fancy dual-mono to single-stereo DIP8 adapter things that were linked previously in this thread, but currently I actually don't have any mono op amps (I plan on getting a couple AD979 sometime in the future, but they're costly enough to not be an impulse-buy).

Are there any suggestions on an inexpensive (maybe just a couple bucks? The adapter is only $3.47) decently-good NE5532-compatable mono opamp just so I can make sure that this adapter is functioning alright? Outside of the stock NE5532, All I have right now is an LME49990 and an AD825 (both of which are dual-mono soldered onto a single stereo DIP8).


EDIT: And it should preferably be a solderless DIP8 form-factor opamp.

 

[Ivan TT]

I'm going to be ordering one of those fancy dual-mono to single-stereo DIP8 adapter things that were linked previously in this thread, but currently I actually don't have any mono op amps (I plan on getting a couple AD979 sometime in the future, but they're costly enough to not be an impulse-buy).

Are there any suggestions on an inexpensive (maybe just a couple bucks?) decently-good NE5532-compatable mono opamp just so I can make sure that this adapter is functioning alright? Outside of the stock NE5532, All I have right now is an LME49990 and an AD825 (both of which are dual-mono soldered onto a single stereo DIP8).

AD8065, if it works you get yourself a very decent sounding dual!

 

[Nintendo Maniac]

AD8065, if it works you get yourself a very decent sounding dual!

The adapter is only $3.47, and two of such opamps in a DIP8 form factor costs over 10 bucks even from Chinese ebay sellers...heck those same Chinese ebay sellers have the dual mono on a single DIP8 variant for only like a single dollar more!