Head-Fi.org › Forums › Equipment Forums › Computer Audio › ASUS Xonar HDAV1.3 mods
New Posts  All Forums:Forum Nav:

ASUS Xonar HDAV1.3 mods

post #1 of 53
Thread Starter 

Haven't been a while since I posted an article here, since its my semester break and I have some money I thought of buying new cards replacing dead X-Fi XtremeMusic, its Xonar HDAV1.3. I got it 2nd hand for less than the price of new one and since its out of warranty I looking to improve it straight away. Stock card sounds quite nice, it have wide soundstage but lacking in depth (like wide stage but with only a couple of meters of depth so everything sound packs together) and the percussion and details is muddy. Bass is quite nice though, and vocal seems to be in front a bit, I bet that is the work of LM4562. First thing I did is change the buffer NJM 2114D with LME49720NA. Immediately there is wider and deeper soundstage but its not as wide as I remember how LME49720 should sounds like so there is something needs to be improved more.

 

So next thing on the line is reading PCM1796 datasheet. After few minutes reading the datasheet and comparing it to the card layout I found two problem; one is there shouldn't be a 220uF caps on the output for each channel and the value of I/V feedback capacitor is wrong as well. Bypassing the final capacitor makes it sound less muddy and there is no DC offset to report but I made the mess of it by ripping off the solder pad on positive side of the caps confused_face.gif the PCB trace is very fragile but the trace that goes through the PCB is more robust. Luckily for me there is small caps nearby that goes before the 220uF cap pdas so I could tap from there and connect it to the negative solder pads. Problem solved.

 

The biggest sonic improvement comes from changing the stock 2200pF feedback caps in I/V stage with the right 2700pF from the datasheet (page 34, Figure 36). Soundstage opens like I never imagined or heard before, and the separation between each instruments are easy to differentiate. I did one more mods, opamp power supply bypass by using two 0.1uF caps in series and connect both rails together on each three opamps.

 

But one thing that impresses me is how PCM1796 sounds. This is the first time I heard Burr Brown DACs and already I liking it, its one of the best sounding DAC I heard, the other one that I really like is Cirrus Logic CS4398 which is more analytical than PCM1796, but the latter is a lot more musical and soothing to my ear. Its like drinking cup of warm tea in raining evening, yep its that good biggrin.gif

 

For the sound driver I'm using ASUS XONAR Unified Drivers 1.50

 

Some pics of my mods:

 

analogc.jpg   opamp.jpg

 

Any comment/discussion is welcomed.

post #2 of 53

Hi there Apocalypsee. I see you came to the same conclusion I did as far as direct coupling the output on your card. The sound being more focused & clear. I also found the highs to be smoother sounding as well as being more detailed.

 

The DAC on your card is the same as on my girlfriends Xonar D2/PM card which by the way is direct coupled from the factory. The DAC on the STX is one step up from that but no one will likely notice if you were to substitute one for the other in the same circuit. Both are fantastic sounding. Only difference is likely noise which is slightly higher on the PCM1796 compared to the PCM1792a.

 

All 3 of these cards likely have a slightly different sound due to power supply & layout differences the STX being more focused sounding than the D2/PM. Not sure where the HDAV1.3 fits in but I'm guessing between the STX & the D2/PM though probably closer to the STX in sound signature that is once it is direct coupled as I did with my STX & you did with your HDAV1.3.

 

Looks good as far as your mods are concerned. Too bad about the lands where the coupling cap was. You probably needed 2 soldering irons to lift them out without damaginng the trace. In spit of that you did a great job repairing it & I'm glad you like the sound of your new card.

post #3 of 53
Thread Starter 

Thanks for the reply germanium biggrin.gif This card is probably closer to STX like you said because it uses molex connector to power its analog out just like STX did but the capacitor used is not the one that being used on the STX, its just the regular caps. The short leg distance between the capacitor leads prevent me from changing that because its too close to each other as I have lots of big high quality caps lying around (Sanyo OSCON and Nichicon MUSE) and I'm worried I could damage the card.

 

Thanks for suggesting on using two solder iron when removing surface mounted caps, why haven't I thought of that? I saw that you do undervolt as well, since AMD Phenom II can be controlled by software I'm undervolting and underclocked all the time unless when I'm gaming.

 

I may wanted to do more, maybe by changing feedback capacitor by looking at opamp datasheet perhaps? But I don't know how to do that yet since opamps are generally to drive phono preamp or something, so transferring them into this soundcard is unknown to me. Or what I'm thinking is wrong?

post #4 of 53

I believe those are solid electrolytics as opposed to normal electrolytic which have a liquid electrolyte. I'm not sure of the quality of the solid electrolytics used as coupling caps but from your experience while they may be better than standard electrolytics they are still not the cat's meow in terms of accurate reproduction.

 

Looks like D.C. coupling is still the way to go like we did with these sound cards.To me the only other option would have been mettalized film caps but those are very expensive & way to large for the confines of most computers, plus they may be prone to pickup interferance from the computer due to thier large surface area.

post #5 of 53
Thread Starter 

Stock capacitor value is 220uF. 220uF for metallized film caps were huge, since there is no DC offset I don't want to use caps over there. More stuff on audio path isn't a good thing.

 

I'm thinking of modifying some caps on power section next, but I don't know whether it would make any significant difference. Is there any more mods that you recommend?

post #6 of 53
Quote:
Originally Posted by Apocalypsee View Post

Stock capacitor value is 220uF. 220uF for metallized film caps were huge, since there is no DC offset I don't want to use caps over there. More stuff on audio path isn't a good thing.

 

I'm thinking of modifying some caps on power section next, but I don't know whether it would make any significant difference. Is there any more mods that you recommend?



This may be somewhat difficult but you could try to bypass the analog powersupply caps with about 1-2 uf mettalized films. Do not bypass the powersupply caps to the DAC though as it will probably tilt the sound signature upward very significantly at the high end. This would improve high frequency detail without tilting the sound signature significantly. Make sure you attach the lead at the tail end of the writing to ground as that is the outer foil & you want to attach that to ground. While mettalized films are non directional it is important to minimize interferance. Yoou can get mettalized films in these sizes that are reasonably small if you get like 50 volt D.C. models

 

Just did this myself on my STX card, More upward tilt of sound signature than I was thinking there would be but still not bad. Highs are definately more open & airy. Bells stand out more but in good way. Less grainy sounding.  Highs even though compared to stock are tilted up a bit sound smooth & very detailed. Less harsh on music that sounded somwhat harsh before. Depth still there. Depth actually improved. This is all with the stock opamps by the way

 

 

 

I used one 2uf 100volt non inductive wound mylar cap per 12 volt rail nearest the I/V convertors. Yours are nearest the buffer amp.

 


Edited by germanium - 1/1/12 at 6:04pm
post #7 of 53

Too bad I can't really underclock. My computer won't boot if I take the base clock any lower. That's ok I get huge saving in power at the stock speed by undervolting, Much more that underclocking would provide on my system.

post #8 of 53

The effects of this last mod benefits well down into the midrange. Voices are clearer & smoother at the same time, this all takes place without emphasizing sylibants in an offensive way, no harshness. You can hear the singers breaths much more clearly. Percusion instruments such as shakers, sticks, bells & drums all sound better. Been doing a lot of listening today & not feeling tired at all. Electric guitars have all the bite that you hear from them in live situations yet there is lot more detail in the overtones than you hear when reproduced on average systems.

 

The depth of the soundstage is absolutely amazing but not artificial sounding at all. Imaging also improved as more sounds move beyond the confines of the speaker spacing. No artificial imaging enhancements were activated yet the depth & width of the image are vastly improved.


Edited by germanium - 1/2/12 at 12:23am
post #9 of 53

The depth of the soundstage is absolutely amazing but not artificial sounding at all. Imaging also improved as more sounds move beyond the confines of the speaker spacing. No artificial imaging enhancements were activated yet the depth & width of the image are vastly improved.

 

 

This post was meant to be an edit but edit wouldn't take.

post #10 of 53
Thread Starter 

Thanks again germanium. I guess what you mean the two 47uF solid caps on the right of buffer is it?

 

I'm out of capacitor really, I found some 0.22uF red mylar caps from dead CRT circuits. There is plenty of mylar caps on the PCB but I need to check for pairs but so far that is the largest I found, most of them are 104, which is 0.1uF. Soldering them is tough job because the solder iron I use is not a small tip and the area of solder is small and there is a couple of SMD caps/resistor there. But I managed to solder them anyway. I'm not going to test it yet since its evening and things are pretty warm and quite noisy with my dad playing his system downstairs (a couple of transmission line floorstander, you know the bass of these babies are)

post #11 of 53
Quote:
Originally Posted by Apocalypsee View Post

Thanks again germanium. I guess what you mean the two 47uF solid caps on the right of buffer is it?

 

I'm out of capacitor really, I found some 0.22uF red mylar caps from dead CRT circuits. There is plenty of mylar caps on the PCB but I need to check for pairs but so far that is the largest I found, most of them are 104, which is 0.1uF. Soldering them is tough job because the solder iron I use is not a small tip and the area of solder is small and there is a couple of SMD caps/resistor there. But I managed to solder them anyway. I'm not going to test it yet since its evening and things are pretty warm and quite noisy with my dad playing his system downstairs (a couple of transmission line floorstander, you know the bass of these babies are)


Yes the 47uf caps These bypass caps must attach to ground & the +&-12 volt rail in order to provide a return path for the signal, I noticed the caps on the underside of the board do not attach to ground but only across the power terminals of the opamps. By puting 2 caps in series not attached to ground you also half the available capacitance but double the voltage capability. The power supply is in the signal path so one must provide a return path for the signal to get the most benefit. I would definately go with something bigger than .22uf, .5 uf should do it though you could go as large as 2uf.you parralell 2-.22 uf caps if you have 4 of them these circuits are feeding a high impedance circuit so they don't need to be absolutly huge to make quite an impression. I like to make the bypasses at least 1% of the size of the filter caps they bypass which is why I chose 2uf as my filter caps are 220uf. Cap type should be noninductively wound metalized films. If both terminals come out one end they are likely not noninductively wound. Noninductively wound caps typically have terminals coming out both ends. while it is possable to make a radial connected noninductive cap it would be extremely rare if at all.
 

Yes I know what transmission line speakers are. They can have quite a lot of bass that is clean & reasonably extended if properly done. I've had many types of speakers in the past myself. Planar ribbons with 3-12 inch woofers  aumenting the ribbon for bass with each having 2' travel operating in free air being only attached to a flat panel that could deliver flat response to 20Hz & even pressurize a room at 10Hz sufficiently to rattle windows. Couldn't hear the 10Hz sound but could definately hear & the windows rattle from it as well as feel it. 


Edited by germanium - 1/2/12 at 1:27am
post #12 of 53
Thread Starter 

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.

post #13 of 53
Quote:
Originally Posted by Apocalypsee View Post

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.



As long as you went accross the legs of the capacitors in question they are grounded thus providing the signal return path. When I was talking of the caps that weren't grounded I was talking of your added caps on the back side of the board (blue ones). The solid electrolytics are grounded so if you go across each one individually your set to go. Thats what it sounds like you did so your ok. 

 

I agree with you that even though on the surface it seems that that the signature is tipped up slightly it retains it's warm sound character & in some ways even gained more warmth but now I find the sound very tantilizing & am drawn into the music much more than before. I can listen for hours & hours where before I felt tired after 1 or 2 hours & needed to sleep. Yesterday I listened for over 12 hours & still didn't really feel tired but had to give my house mates a rest as I was playing it pretty loud for at least 10 of those hours.

 


Edited by germanium - 1/2/12 at 6:20pm
post #14 of 53

Deleted as edit which said the same thing finally took a day later.


Edited by germanium - 1/2/12 at 6:23pm
post #15 of 53
Thread Starter 
Quote:
Originally Posted by germanium View Post



As long as you went accross the legs of the capacitors in question they are grounded thus providing the signal return path. When I was talking of the caps that weren't grounded I was talking of your added caps on the back side of the board (blue ones). The solid electrolytics are grounded so if you go across each one individually your set to go. Thats what it sounds like you did so your ok. 

 

I agree with you that even though on the surface it seems that that the signature is tipped up slightly it retains it's warm sound character & in some ways even gained more warmth but now I find the sound very tantilizing & am drawn into the music much more than before. I can listen for hours & hours where before I felt tired after 1 or 2 hours & needed to sleep. Yesterday I listened for over 12 hours & still didn't really feel tired but had to give my house mates a rest as I was playing it pretty loud for at least 10 of those hours.

 


Hmm weird because I see lots of peeps doing the things I did on Xonar threads, user like csy and shimm doing what I did. I will try to ground them properly and see whether if that made any noticeable difference.

 

Its kinda bad because its not fatiguing at high volume because now I liked to ramp up the volume. I afraid it might be bad for my hearing in long term, I need to limit the volume or the time I listen to music tongue_smile.gif

 

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Computer Audio
Head-Fi.org › Forums › Equipment Forums › Computer Audio › ASUS Xonar HDAV1.3 mods