Hotrodding the X-Fi: A Layman's Guide (No 56k)

[ViciousXUSMC]

Thanks Herbal,

@cotdt... what you say makes sense. I have noticed that phenomina before. I figured tho that its because you have no signal "power" or strenght and thus just need to turn it up some.

I notice nothing diffrent from 10% to 50% but at 0 volume I just figured the signal is too weak, not that its a direct influence on signal quality.

I actually thought on my own from trial and error that max volume was too much and causing distortions because the signal was over powered, and that some of the quality was now being subsituted for raw amplitude.

Of course this is all on the other side of the tree for me, I am more familiar with say how a speaker works and the relationship of soundwaves and there amplitude and how high amplitude can cause sound wave distortions due to the inperfections of speakers. So I dont doubt what you say, but I have to play around with it some and see what happens.

Im still doing the mod when the parts are here, I dont listen to my headphones nearly as often as I used too so I would benifit from the better SQ on my speakers. I will also try to power my headphones from the reciever and see if there is any noticable quality loss.

When I use my headphones I go for maximum sound clairty and sound stage, usually 3dcmss mode, I figured it would be disturbed some running thru my reciver and equalizer even tho I havnt tried.

[cotdt]

if you're powering your headphones straight out of the soundcard, then yeah if you turn it up too much it will start clipping. it shouldn't be a problem for soundcard -> amp since the input impedance is very high (ie. the amp won't "suck" as much power as headphones).

yes, you'll need DSPs to get some soundstage out of your headphones. or consider getting a K701.

[Seba]

I could hear noticeable difference in dynamics between 90% and 100% volumes from the soundcard. Music kinda opens up more when it is set to 100%. I use to keep the 90% because old soundcards and cheap computer speakers seem to have distortion if set to full volume, but I have no problems with the latest set.

[ViciousXUSMC]

I was right about the problems at 100%. Not driving my headphones directly (100% would blow my ears up) but 100% out on the computer going into my sterio reciever gives me some craking on certian songs in the high treble. Im doing 90% now vs the 50% I was at and so far no issues.

Also having 100% makes my volume so touchy on my reciever (its super powerfull 135w per channel and just pushing small spekers right now)

Not sure what the deal is with the poping/cracking, I coudlnt hear a diffrence between 90 and 100% myself anyways.

[Awake77]

I just wanted to pop in and say thanks to cotd and everyone else for this awesome thread!! I just modded my X-Fi XtremeGamer with the BlackGate 2200uf and LM4562 on the main out.

The improvement in sound quality is dramatic - my X-Fi is a pleasure to listen to know and sounds nearly as nice as my Echo Layla 24 (a pro audio interface).

I'd like to try replacing the op-amps and caps on the breakout box too...has anyone done this yet?

Thanks,
A77

[cotdt]

Originally Posted by Awake77

I just wanted to pop in and say thanks to cotd and everyone else for this awesome thread!! I just modded my X-Fi XtremeGamer with the BlackGate 2200uf and LM4562 on the main out.

The improvement in sound quality is dramatic - my X-Fi is a pleasure to listen to know and sounds nearly as nice as my Echo Layla 24 (a pro audio interface).

I'd like to try replacing the op-amps and caps on the breakout box too...has anyone done this yet?

Thanks,
A77

glad you like it. i've modded the breakout box, but it sounds best straight out of the soundcard. it's worth modding if you have no choice but to use it, though.

[cotdt]

Originally Posted by ViciousXUSMC

I was right about the problems at 100%. Not driving my headphones directly (100% would blow my ears up) but 100% out on the computer going into my sterio reciever gives me some craking on certian songs in the high treble. Im doing 90% now vs the 50% I was at and so far no issues.

strange, but don't worry you're not the only person who reported this problem. i was never able to figure out what the problem is since i've never experienced it myself.

[Awake77]

glad you like it. i've modded the breakout box, but it sounds best straight out of the soundcard. it's worth modding if you have no choice but to use it, though.

What did you do to mod the breakout box? Just swap out the op-amps?

I dont need to use it I guess but I like plugging into it for gaming, etc. If it makes a difference Id like to give it a shot.

Thanks!
A77

[cotdt]

Originally Posted by Awake77

What did you do to mod the breakout box? Just swap out the op-amps?

I dont need to use it I guess but I like plugging into it for gaming, etc. If it makes a difference Id like to give it a shot.

Thanks!
A77

i also used LM4562, but there are just too many things that can be modded on the breakout box, you might as well make your own breakout box. personally i would stick a nicer headphone amp in there.

[trodas]

I would like to do this mod as well, but since I already exchanged some caps on my X-Fi Fatal1ty ( WTB: K3000 or HP3000 ) so I would rather measure the RMAA results first now, as they are, and then compare what each change do to the X-Fi sound quality.

The thing is, I can't seems to get the RMAA work with my X-Fi
I definitively will contribute my RMAA result - as soon, as I figure out where to plug the loopback thing I made... Getting weird results (eg. the RMAA complaining that signal is too high or low, no matter what I do - or that there are too many noise in the signal and definitively the signal is unbalanced...) when using the jack to jack cable between the mic input and "green" output, supposedly used for the 2/2.1 channels output... Any ideas?

Also I have question about the big Audio rated caps used there. I measured and on the cap 1.25V on it and like 50mV ripple = it must, simply MUST be a just power line for the circuit.
It is connected to the chip Vcore pin and it it - on the other side - connected to that voltage regulator thing on the card.

So in a voltage filtering job perform simply MUCH better Low or Super Low ESR caps (like Panny FM, Samxon GC, Os-cons...) that the Audio caps, witch does suxx at the job badly. Hence we are back at my original question WHY use BlackGate cap there.
If the sound get better - by some measurments like using RMAA - then I start looking where to gather one of these BlackGate cap. But I bet that when I exchange the Panny FM 470uF 16V I use there now for say like Sanyo Os-con ESVP 820uF 4V then nothing happen to the audio quality output.
(even the difference between the caps specs are like night and day...)

But the other mods sounds very reasonable to me - time to order the samples...

[Awake77]

i also used LM4562, but there are just too many things that can be modded on the breakout box, you might as well make your own breakout box. personally i would stick a nicer headphone amp in there.

That's actually a great idea - any recommendations? Ideally I'd like to build my own headphone pre...

[cotdt]

Originally Posted by Awake77

That's actually a great idea - any recommendations? Ideally I'd like to build my own headphone pre...

any amp that can be powered by 12V would be nice. there are many choices.

[cotdt]

RMAA measurements have been discussed previously. According to it, the mods make no difference. According to it, the CMoy is one of the best amps ever...

Bottom line is that it cannot do what are considered the more important measurements. If you want low ESR, the Blackgates are unbeatable in the realm of electrolytics.

[LeonvB]

Just picked up the opamps that finally arrived. Replacing the opamp was very easy, cutting the legs on one side and using a big iron to solder off the rest did the trick. A little of the masking tape did wonders for positioning the new opamp. Thanks for those tips.
The difference is indeed bigger than swapping out the cap. IMO The cap improved mainly the low end, while the opamp dusts off the entire spectrum.

[phalanger]

I put a AD8599 in place of the 4562 today with good results (mostly for diagnostic reasons but also curiosity). The resulting sound is very good. Hard to say which is better but since a few have mentioned wondering about it i'd like to report it does work fine and sounds very nice. The reason I wasn't too sure is that on the AD site its rated for a higher voltage range than 5v (9-36v), but googling some more found me a info sheet that did mention 5v as being adequate.. either way yeah it works just fine

It scored a bit better in RMAA i think then 4562 but ive not been able to do optimal tests on either opamp due to the same RMAA complaining about the signal being either too soft or clipped, which is why i tried the different opamp, but unfortunately that problem still exists even though when i do run RMAA the test results are still fairly good so it doesnt seem to be a serious issue. How is this for others, any problems getting RMAA to give the 'green light' without having to resort to increasing the recording level to >0dB?
Since I started modding this card right away I don't know if its any different with stock parts. I thought it was just me having this problem but since someone else mentioned the exact same, Im curious how RMAA goes for others.

Also I cleaned the mods up a bit and reshorted the coupling caps (the nice way this time ). tomorrow ill do the line-in and surround opamps Does anyone know if any (and if so, which) capacitors are in the line-in signal path?

And yes the masking tape made it a breeze to put/keep the SOIC in place while soldering the first legs, thanks again for that tip