Asus STX OP AMPS

Soldato
Joined
7 Jan 2009
Posts
6,369
Anyone have any experience with experimenting changing these?

I dont dislike my STX but i do feel it lacks sometimes in the bass department.

Any recommendations to what i should get to try,And even more important wheres the best place to buy them as i dont want to buy fakes. :)
 

Kei

Kei

Soldato
Joined
24 Oct 2008
Posts
2,750
Location
South Wales
I tried a big variety and I found that there was almost no difference unless the opamp was not well suited to the STX design or was so old that it had quite high noise. I tried RMAA measurements and I couldn't even measure any appreciable difference either. Excessively fast opamps are likely to oscillate which will cause changes to the way the output sounds. (not in a good way either)

I tried:
AD826AN
OPA2604AP
OPA2134PA
LM4562NA
LM833N
TL072
TL2072
NJM2114D
OP275
NE5532P

I'm not 100% sure how I left it anymore as the only opamp that really made any appreciable difference was TL072 which had more hiss. I think I left it as LM4562 in both the buffer and I/V stages. If you really want to play with changing them, RS, mouser, farnel, digikey and possibly CPC are reputable electronics suppliers that sell a big range of opamps. You need dual DIP-8 package items.

As to actual frequency playback level, this is how my stock ST behaved @ 44.1/16 bit. Less than 0.5dB down at the extremes of human hearing. 0.5dB difference is inaudible. If you have a lack of bass, that's either the source media your speakers/headphones or you have a problem with the card. You can test the card by plugging the output straight back into the input and using RMAA to test it. (don't do this with the headphone amp output though)
fr.png
 
Man of Honour
Joined
13 Oct 2006
Posts
91,125
Everything below is my subjective opinion:

One issue you have with swapping out op amps on a card like that is you don't necessarily know what the circuitry is like - where gain is involved rather than just a buffer there might be compensation capacitors in the feedback loop that do bad things if you use an op amp that isn't unity gain stable, for op amps that aren't internally compensated they might not be stable if buffer stages and/or require a higher gain than the card is setup to use and then you have some complications with the power supply - whether there is enough voltage and whether it is setup for single or dual supply support, etc.

The op amps that tend to be fairly forgiving and work in most cases would be stuff like the OPA2132/2134, JRC4556, LM4562, the NE5000 series, etc.

The NE5532 is a "workhorse" amp largely transparent but a bit dated now and can sound a bit "raw".

A lot of people like the OPA2134 as its largely transparent with a fairly mellow sound but personally I think it is like someone playing the same tune for the 1000th time - accurate enough but with no passion - I prefer the OPA2227 as a replacement as it sounds a bit more "involved" with the music. (Or even better OPA2228 but you'd need carnal knowledge of the circuitry to know that would work in there).

LM4562 is very detailed, crisp and punchy and in buffer positions often allows the good bits of any previous gain stage through as well i.e. if the op amp in the gain stage has good bass the LM4562 tends to leave it relatively unmolested. However it can sound a little robotic and power supply noise can result in it being a bit metallic and harsh. One of my favourites however. I'm not sure what is going on with regards to manufacturing though as TI did say they were ending production of it in DIP format but most places still had good supply last I checked.

NJM2114 generally has very good stats but IIRC the bass is a bit lean and the treble a bit fatiguing.

One thing that might be worth trying is the LM4562 or one the LME equivalents with a 10uf "boutique" capacitor over its power supply pins i.e. Nichicon FG, KZ or Elna Silmic II in my experience that helps to give the low end more satisfying body to it and a cleaner background.

One of my favourites for bass is the AD8066 paired with LM4562 buffer(s) but you'd need to know the circuit supported it as it isn't just a drop in replacement (neither is it available in DIP format).

If you feel like playing around with adapters then the AD8610 in the IV stage might be worth a consideration.

I suspect the circuitry is very similar to fig 36 on page 35 http://www.ti.com/lit/ds/symlink/pcm1792a.pdf so the transimpedance op amps would probably have to be unity gain stable to avoid complications and destroying sound quality.

EDIT: I usually buy my op amps from Farnell or Mouser (pretty much guaranteed you won't get fakes) but Mouser ships from the US so is a bit expensive for small value orders.
 
Last edited:
Soldato
Joined
25 Jul 2012
Posts
7,377
Location
Ankh Morpork
Without the ability to tune the circuit around the op-amps I'd just leave it as is. Besides, if you want better bass then you should be looking at your speakers and headphones as that is where the vast majority of sound tuning is done.
 
Man of Honour
Joined
13 Oct 2006
Posts
91,125
As to actual frequency playback level, this is how my stock ST behaved @ 44.1/16 bit. Less than 0.5dB down at the extremes of human hearing. 0.5dB difference is inaudible.

One thing I found is that despite the numbers things can sound one way or another the defies what any amount of scientific analysis would tend to indicate - the 2114 for instance has very impressive stats on paper yet I've always found that in application the bass sounds lean and treble isn't perfect (I've spent quite a lot of time reading up on and experimenting with the O2 headphone amp that also uses it - IIRC its the default IV op amps in the STX).

Over the weekend I was playing about with a build that used the O2 type core circuitry for the audio part but with some of my own changes to the power supply section and OPA1612 gain stage and LMH6643 buffers like the Fiio X5 and that setup has very impressive numbers on paper but at the end of the day the 1612 has some tonal colouration - but it is noticeably different sonically to the stock O2 with far more spacious and lively sound.
 
Man of Honour
Joined
13 Oct 2006
Posts
91,125
Rather than start another thread - quite impressed with the combination of OPA1612 which has loads of width to the sound combined with the low noise of the O2 type design - once I worked out how to not overload the input stage :S (low voltage used with the LMH buffers makes it a bit easy to hit clipping and distortion if the source volume isn't controlled).
 
Back
Top Bottom