Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › AD797 now sounds like it should
New Posts  All Forums:Forum Nav:

AD797 now sounds like it should - Page 2

post #16 of 80
So, what did you do with the AD797 after reading that document? Gotta know if I need to do something more but I think the AD797 worked as it should in my circuit. It's good but not among my favourite.

Here is something more comprehensive:
http://focus.ti.com/lit/an/slod006b/slod006b.pdf
post #17 of 80
Thread Starter 
I just added the recommended capacitor from pin 8 to pin 6 for "distortion compensation". Jung says it's even better to put this capacitor between the buffers output to pin 8 to get a "distortion reduction" in the whole circuit. Maybe I'll try this if I build a dedicated amp for AD797.

I've put two of these in my LISAIII clone, so they are matched with AD829 in the ground channel. I use the cap mentioned above and 10 pF in the local feedback loop, but I've removed the resistor for the local loop, so there's no Jung multiloop at the moment. Now this amp has a more neutral tonality than with AD744 and there's not the typical JFET graininess or "Analog snap" anymore, for good and for bad. I think it softens the sound a bit, loosing some bite and aggressiveness. Maybe adding 100 pF instead of 50 pF "overcompensates" it, or it's the lack of multiloop. After the hollidays I'll try to find some better caps. Maybe it's the AD829, I've always found it a bit soft.

It still mates well with LM6171 as ground channel in the other amp only fed with 9.6 V and used with Jung multiloop. For some reason this amp is a bit brighter.

No doubt about it, AD797 is one of the best opamps.

Edit: I added local feedback to get the Jung multiloop, and it could be placebo but I think it's more neutral now = less soft.
post #18 of 80
Thread Starter 
Shopper, I always thought you're after some kind of coloration to get a "meatier" sound, yet you want all the details. I'm more a neutrality freak. AD797 is more neutral, very transparent. If I add this amp (the LISAIII-clone with AD797/AD829 with multiloops) between another amp and the phones, it's very hard to hear any difference. It adds very little, all to my liking. I have to give it some time, but I think this is a keeper, better than the original LISA. AD744 is a little lean and has that JFET sound. I think there'll be only bipolars in my "final" amp.

I've never heard OPA627 or 637. From what I've read, I might like the 637. I've heard OPA2107 which is supposed to sound like OPA627, and I think it's too colored and lacks details. I guess you should like OPA637. Why not try AD797? 7+9+7=23 so there's no reason for you not to try it.
post #19 of 80
2227 isn't JFET input, it's bipolar. 797 has lower distortion than any of the three you listed.
post #20 of 80
Quote:
Originally Posted by Shopper View Post
Ops, true. In fact, the OPA2134 is more musical sounding than the OPA2227 (as if it had a more ear-friendly type of distortion); yet both have that typical (full, smooth, without any apparent artificialness) Burr-Brown sound. I guess that probably the house sound thing has greater relevance than the FET/bipolar thing.


I'm sure that the AD797 does have (even) lower distortion. The LM4562 also has super low distortion on paper but I don't like it better than the OPA627, as it lacks some naturalness and body (some serious deficiencies in my book). The LME49720 is somewhat better.

I'd still really like to try the AD797, and I probably will even with the concern if it's going to be stable.
The 2227 isn't a particularly good op-amp for audio, IMO, but not because it's bjt input. The LM4562 is OK, although not great; I'd rather use the 797. AFAIK the 49720 and 4562 are the same chip; the data on them is identical.

BB doesn't have a "house sound" that I'm aware of (and I've used a lot of op-amps); that appears to be another one of those unifying principles that are popular on head-fi that I guess has managed to propogate itself, probably because the same core group of op-amps have been used and so there is sampling error due to inadequate set. JFET-input chips involve particular challenges in implementation that can perhaps produce some common distortion problems, although I haven't noticed a more "musical" sound as some sort of general rule. Then again, it's rather unclear just what that means.
post #21 of 80
Quote:
Originally Posted by Filburt View Post
The 2227 isn't a particularly good op-amp for audio, IMO, but not because it's bjt input. The LM4562 is OK, although not great; I'd rather use the 797. AFAIK the 49720 and 4562 are the same chip; the data on them is identical.
Have you tried the AD8599? If so, how does it compare in sound to the AD797? I've found the AD8599 to be warmer, detailed, but somewhat less so and less immediate sounding than the LM4562. I'm wondering if these characteristics also transfer to the AD797? If I were to guess, I would say the LM4562 has less distortion and less coloration. What is it about the LM4562 that you don't like? How about the LME49710 - any experience with that one?
post #22 of 80
Quote:
Originally Posted by Shopper View Post
LME49720 is highly recommended by me

Intersting - the AD8599 sounding less immediate than the LM4562 somehow justifies my lack of interest for this chip so far. The much trashed AD823 is a chip I like because it sounds immediate.
I didn't like the AD823 much at all.... found it sort of bright and brash, but lacking in true high frequency detail.... not very hi-fi.
post #23 of 80
!Why can't I delete my double post!
post #24 of 80
Thread Starter 
AD797 doesn't have any of the "Analog house sound". It's liquid smooth and unlike other high quality bipolar amps the mids don't sound recessed. I prefer it to AD8599. I'll have to see how it stands up for long time listening. Maybe I'll find it to be lacking aggressiveness (it takes longer to find out there's too little than too much aggressiveness), but at the moment this is the new champ. I hope I'll get some more AD825 soon to find out if they're even better.

The noise level is very low, and there's no problem listening to IEMs. The AD744 I used before is the opposite. I guess it's the noisiest HiQ opamp out there. The amp looks like a rats nest now and still the noise level is just slightly above Mini3 at the same gain.

About AD8599, I've thought of it as a JFET input opamp, but it's probably not if it's a compensated AD797. They don't tell which it is in the data sheet, but it's a bit grainy in comparison to LM4562 and AD797. It's warmer than neutral and the mids are a bit recessed just like in LM4562 but the latter has a brighter and somewhat clearer treble. I don't find it sounding less "immediate" than LM4562. LM4562 is a bit uninvolving, it's very apperent if you use it in all three channels, then it's really boring. I find AD8599 to be better in this aspect, more involving and warmer.
post #25 of 80
Interesting........ maybe the LM4562 is boring because it doesn't add much coloration (or distortion)? Some people like a bit of grit to liven up the sound. I agree that the AD8599 is warmer - it sounds a bit rounder in the lows to me as well.

As for the AD825, so far it is my favorite FET op amp. It's decently quiet, but still quite clear on top and clean and even overall, yet not harsh like a TLE2072 or AD823. It sounds quite uncolored to me. The TLE2072 seems to add some unnatural, somewhat brittle highs which sound impressive at first, but become fatiguing after a while.

Tonight I spent some time comparing the AD823, LT1057, OPA2604, OPA2132, OPA2134, AD825, TLO72, OP249, and TLE2072. Of these FETs, the AD825 sounds the best overall, with the OPA2132 coming in second. The LT1057 is kind of interesting in the midrange, but lacks openness in the highs. As for the OPA2134, I'm puzzled about why people say it sounds just like the OPA2132, because I clearly hear more detail and less veiling with the OPA2132. In comparison, the OPA2134 sounds kind of "phased" and rolled-off on top.
post #26 of 80
Thread Starter 
LT13xx is on my wishlist.

I think LM4562 is colored. The coloration is recessed mid. I also find the treble colored but not grainy or lacking details, just a bit artificial. It's not because it's too neutral I find it uninvolving. I think AD797+AD829 is more neutral, and AD8599+AD825 is also more neutral, but these combinations are more involving and sound more alive.
post #27 of 80
Shopper, I guess I would choose AD845 over the AD797 but honestly I wouldn't use any of them anywhere. For my today's needs they are both too far from being neutral.
OPA2107 does not sound like the OPA637, the latter is more neutral but sometimes "slow" and boring. OPA2107 has got great soundstage but not so great timbre and impact.
Op-amps which still matter to me are:
ADA4841-x, OPA1632, OPA(2)211, AD825, AD811, AD8045, AD8599, LM4562, AD8022, OPA(2)604
And the LT's which I haven't tried.
post #28 of 80
I heard the best Linear op-amp ever was LT1028. Gotta try. This one being not so new should be better suited for stand-alone devices providing high supply voltages, so maybe this one?
AD845 is warm and dark (whoaaa, dark AD op-amp ), AD825 is warm and rather bright, and has got very interesting bass.
post #29 of 80
Quote:
Originally Posted by Shopper View Post
The core difference being that I have the AD845 and not the AD825.

Which is the more transparent of the two, btw?
AD825 IMHO, gives better sound differentation while AD845 gives a bit of one-note sound presentation IMHO.
post #30 of 80
Quote:
Originally Posted by 12Bass View Post
Have you tried the AD8599? If so, how does it compare in sound to the AD797? I've found the AD8599 to be warmer, detailed, but somewhat less so and less immediate sounding than the LM4562. I'm wondering if these characteristics also transfer to the AD797? If I were to guess, I would say the LM4562 has less distortion and less coloration. What is it about the LM4562 that you don't like? How about the LME49710 - any experience with that one?
Seems like folks associate 'neutral' with something other than low distortion. The 797 is more neutral and seems to have better settling characteristics than the 8599, and its distortion is lower, so it's overall the superior of the two. The LM49710 is just a single channel version of the 4562/49720. It's fine, but nothing extraordinary. I'm not a huge fan of that series. I think the NE5532 might sound a bit better.

Shopper - I don't know what you mean by "house sound"; I've not experienced such a thing amongst the 40+ op-amps I've used from BB, AD, NatSemi, LT, etc. Haven't found the 8058 to be harsh, either; you probably implemented it wrong. Your comments remind me of (resemble those of) former member Andrea.
New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › AD797 now sounds like it should