[germanium]

I may be making too much of this fact so sorry if I was pushing this a little hard. I realize also to do this bypass where you are bypassing  surface mount capacitors is not an easy task & you will definately need a smaller tip soldering iron to do this. I just did this last night with my girlfreinds Xonar D2/PM & it was challenging to say the least just getting the solder to stick & not have a cold solder joint but I finaly got it & it sounds great as well.
 
It took me 1 hour before I got satisfactory results. These caps on the D2/PM are surface mount as well & very close to other suface mount devices like your situation. results were identical to the STX that I modded sound wise as far as the areas of improvement.

 

[PurpleAngel]

Germanium
Have you put any thought into getting the LME49860NAs or the LME49990MAs?

 

[Apocalypsee]

Quote:

I may be making too much of this fact so sorry if I was pushing this a little hard. I realize also to do this bypass where you are bypassing  surface mount capacitors is not an easy task & you will definately need a smaller tip soldering iron to do this. I just did this last night with my girlfreinds Xonar D2/PM & it was challenging to say the least just getting the solder to stick & not have a cold solder joint but I finaly got it & it sounds great as well.
 
It took me 1 hour before I got satisfactory results. These caps on the D2/PM are surface mount as well & very close to other suface mount devices like your situation. results were identical to the STX that I modded sound wise as far as the areas of improvement.

Yea its quite tough to get them to stick or to remove them and the trace is fragile especially on positive terminal. Its great to hear that you managed to get it done. I presume all the PCM179x sounds almost identical except for some small little details such as noise. And its good to hear that you say it sounds identical to STX.
 
I noticed opamp of choice in PCM1792A datasheet are LT1028 for its low noise, so I requesting a couple of sample, gonna put them on I/V and probably move my LME49720 on buffer because I don't like LM4562 vocals which is brought up too much up front. LME49720 on I/V, despite sounds great, I don't feel it sounds right. Some songs sounds great but others it looks like its missing bass. I feel there is roll off somewhere. And the soundstage is too deep its feel too much distant after a while listening to it, and changing the feedback caps makes it super wide.
 

 

[germanium]

Quote:

Germanium
Have you put any thought into getting the LME49860NAs or the LME49990MAs?

Was considering it but am getting such great results with stock opamps & just sticking to simple very cheap mods that happen to work extremely well. I believe that these mods I'm doing are more beneficial than opamp swaps. These actually improve resolution substantially without increasing harshness & in fact reduces harshness. Even the sounds that were harsh before though still bright make more musical sense as the improve resolution reveals thier musical purpose & said tones are no longer harsh sounding. this includes synth notes that are right at the edge of human hearingas well as electronic percusion instruments such as found on some Jean Luc Ponty albums. High frequencies are just sweeter sounding.
 
One person who tried the 49990 said they sounded too dry but I suspect that they really needed more bypasses in the power supply as the data sheet call for multiple bypasses with one being a rather large 2.2uf. this may get rid of the dry sound & that is what I'm aleady doing with the stock opamps here & performance is phenominal.

 

[germanium]

Quote:

Yea its quite tough to get them to stick or to remove them and the trace is fragile especially on positive terminal. Its great to hear that you managed to get it done. I presume all the PCM179x sounds almost identical except for some small little details such as noise. And its good to hear that you say it sounds identical to STX.
 
I noticed opamp of choice in PCM1792A datasheet are LT1028 for its low noise, so I requesting a couple of sample, gonna put them on I/V and probably move my LME49720 on buffer because I don't like LM4562 vocals which is brought up too much up front. LME49720 on I/V, despite sounds great, I don't feel it sounds right. Some songs sounds great but others it looks like its missing bass. I feel there is roll off somewhere. And the soundstage is too deep its feel too much distant after a while listening to it, and changing the feedback caps makes it super wide.
 

Not saying that overall the D2 sounds identical to the STX only the areas of improvement are identical though they are much closer sounding than before, not that they were that far apart to begin with.

 

[germanium]

Quote:

 
 
I noticed opamp of choice in PCM1792A datasheet are LT1028 for its low noise, so I requesting a couple of sample, gonna put them on I/V and probably move my LME49720 on buffer because I don't like LM4562 vocals which is brought up too much up front. LME49720 on I/V, despite sounds great, I don't feel it sounds right. Some songs sounds great but others it looks like its missing bass.
 

You are not likely getting bass rolloff but rather the higher frequencies being more transperant gives everything more bite on the attack & added detail across the range. The bass level being more natural & detailed yet sounding smoother all at the same time. The added warmth in the mids is due to the same & also contributes to the impression of rolled off bass but let me assure you it is not actually rolled off. D.C. coupled amp are flat all the way to zero hertz or shall we say D.C.. If you listen to the attack on the drums you will find that they have a more natural attack that you would get if listening to the same drums live.
 
Basically what I'm saying the more transperant you make the signal path the more natural the sound will be. Some tiimes natural can be in some ways less impressive than artificial but then you realize what you been missing when you get rid of the artificial. Soundstage, imaging, warmth, detail & musicality not to mention impact on percusion all improve if done correctly.
 
 

 

[germanium]

Just did the same type mod to to my speakers amps power supply, they are biamps internally. Same type improvements. Used 2-12uf noninductive metalized film caps, one for each power supply rail. Even more of an improvement than doing the soundcard alone. Tighter quicker punchier sounding bass without loss of extension as well. Even more warmth without any loss of detail, in fact detail very significantly improved as well. high frquencies are vrey sweet sounding & seem to extend forever. Depth improved even more.
 
Most the time people associate warmth in the sound with an attendant loss of detail but this need not be the case. It sure isn't in my case. I would characterize the sound extremely wet with the sound of the room the music was recorded in. No dry sound here.

 

[germanium]

Quote:

Oh in that case I need to find the ground plane then, what I done is simply just solder the capacitor leg in parallel to stock caps. I need to check the board again.
 
I did some testing just now since my dad went out so its pretty quiet, my oh my it really sounds good. Unlike other soundcard I tested before (Creative X-Fi XtremeMusic, Elite Pro, Audiotrak Prodigy HD2 and AudioAlchemy DDEv1.1 DAC) this Xonar isn't bright at all. Even on high volume its not cold, shrill or piercing to my ear. The worst of the bunch is HD2, its way too bright and piercing to my ear but funnily enough I use the same metallized film caps from the card to this card and it sounds better.

Looks like there is a ground plane trace that runs right under the buffer amp. Since I don't have your card you will have to use your volt ohm meter to test for continuity with ground at for example your RCA outputs connectors. One of the DAC output frequency filter caps connects to it just above the buffer amp so you could use that as a test point as well as a solder point to attach to ground if it be so that it is gound. 
 

 

[Apocalypsee]

I done more modding to the card, I been PM'ing someone here about capacitor value in relation to Fc frequency and so I removed the 8.2nF output feedback caps and replaced them with 4.7nF ones because the stock value means Fc of 50kHz, with this I could probably raise it by double around 100kHz. Removing the stock caps is a pain, a couple of those legs are such a bugger, but the process went without a hitch but one of the caps melted! I have backup though so its a non issue
 
With this mod it opens up some minute details that were smeared or missing before. So much detail filled the wide soundstage and it doesn't feel hollow/empty as before

 

[germanium]

Have you done anything new since your last post? Curius as to how things are going with your mods.

 

[germanium]

Ive been doing a lot of listening since my last mods to both speakers & soundcard & am liking very much what I am hearing. I also tranfered all my 24 bit 96KHZ recordings to my hard drives from the original DVD's (they are totaly unprotected, Thanks Chesky & Classic records for making it easier to store on my computer). They also sound fantastic. Much of it worth the step up to 24 bit/96KHz resolution which now appears to actually improve sound even more due to the superb resolution of my modded playback system though the improvement is pretty subtle.

 

[Apocalypsee]

I removed the previous 0.22uF bypass on opamp supply because I believe the value is too small to make any difference. I have one 1uF MKP and 1uF electrolyctic but because the space is so tight and the caps were huge I don't dare to put it there, and the leads going to be long as well. I will dug out my inventory, I think I did have some caps lying somewhere, or on old project of mine.
 
Then the aforementioned replacing 8200pF to lower 4700pF caps on buffer feedback. This adds detail significantly, like I said this brings out the missing details in the huge soundstage. But still it doesn't sound fatiguing at high volume, it just adds the missing detail. It seems like this is the last piece of the puzzle, and makes it sounds complete.
 
Other than that I done simple power mods by putting 1000uF on molex connector for both 5V and 12V. The caps over there is small 100uF value so I thought adding this could make the ripple lower.

 

[germanium]

Sounds cool. I think I'm finished with my mods as the sound is phenominal here. Just enjoying the music & am able to hear things I couldn't remember hearing before. Been listening 4-5 hours a night since doing my last mods..
 
If you could find 2 of those 1uf metalized film caps that would be great. It is interesting that some one opamp manufacturer is recomending double bypasses similar to what Ive done with the middle bypass about the same value as what I used on the STX card. There is already one bypass on the card of .1uf The opamp I'm refering to is the LME49990. They recomend 2.2uf ceramic as one of the bypasses. I haven't tried ceramic bypassses but I do really like the sound of the metalized film bypasses. This actually  confirms what I'v been doing for years but just haven't done till recently on this card as I liked the sound basically with just direct coupling. Little did I know how much better it could be. I thought I was the only one recommending or using large value (over 1uf) power supply bypasses.

 

[Apocalypsee]

Can you show any pics of your mods germanium? I would like to see it

 

[germanium]

 
Here is the front side showing the output coupling caps removed. Notice I also had to do a repair job on one of the traces. It was a very delicate repair & had to use very very tiny wire.
 
 
 

 
 Here is the back side of the board showing the 2uf caps across each of the + or - 12 volt rail power supply filter caps closest to the I/V amps in my case. This was the sigle biggest improvement in the sound of the card itself.
 
I did likewise to the speakers but used 12uf noninductive mylar metalized films.
 
Sorry still haven't taken any pics yet of the D2/PM I did the same to. Will add those when I take the pics.
 
Set up flicker account but linked photos were tiny. Not full resolution or even really all that viewable so I decided to try adding it directly which seems to work better
 

 
Here is a pic of my computer. All amps are D.C. coupled except the tweeter for obvious reasons even the subs amps. The tweeter amps are capable but D.C. amplification but D.C. is blocked at the input by the crossover