Hotrodding the X-Fi: A Layman's Guide (No 56k)
Mar 6, 2008 at 6:11 PM Post #1,561 of 2,194
Quote:

Originally Posted by bichi /img/forum/go_quote.gif
eboy2k3: more toys to play with, cheap! And its local to us too!
- eBay Store - MeasureExplorer: Prototype Boards: ME-PB-101RW wireless circuit proto prototype PCB board
- be prepared to spend more cash, next time we meet. LOL! (read: LM3886)
- model number of desolder tool I was yakking about: Hakko 808 (Frys)
- pictured, lower left, in the 633 stand: X-FI-MOD-001b - eSnips, share anything



bichi
The hakko 808 is around $180
mad.gif
I will have to think about it.

Chipamp? Hmm, next project I want to build a DAC w/ dual AD1955 or WM8741. Then I will probably dump my modded SB0550 and get a SB0460 other cheaper sound card for SPDIF out only. No space for speakers in my computer room, so I probably have to stick with head-fi'er route.
redface.gif


The 680uf tant for the SB0550 ADC is interesting. Do you see meaningful performance improvement on that part?

It seems no one responding for my stereo crosstalk pattern issue. Possible meet this Saturday here to swap the card to try?
 
Mar 6, 2008 at 6:57 PM Post #1,562 of 2,194
Quote:

Originally Posted by Random Murderer /img/forum/go_quote.gif
Has nobody done this mod to an audigy 2 zs? i need to know which opamp is the stereo opamp before i go soldering...


well, i said "screw it" and went ahead soldering. found the right opamp. had the stereo and line in opamps reversed in my head, but that's cool, i replaced both
wink.gif

lm4562ma's sound incredible... listening to rush right now, it's incredible. things i've never heard before have appeared.
oh, and the 4562's SIGNIFICANTLY louder than the stockers.
 
Mar 6, 2008 at 10:09 PM Post #1,563 of 2,194
Quote:

Originally Posted by eboy2003 /img/forum/go_quote.gif
bichi
The 680uf tant for the SB0550 ADC is interesting. Do you see meaningful performance improvement on that part?



Only if one is recording and would like improved low-freg (20~200hz) response.
- better THD and improved low-end crosstalk by 2~3db.
- no issue increasing Vref Offset Cal period from 2.4sec to 3.4sec, relative to ucode. (see AK5394 ref below)
- sonically, it reproduces better LF tracking/phase, but only if listening critically or doing delta-phase measurement.
- for most, probably not worth $16+ USD new unit price, low-quant, for Kemet T510 at that value. (DigiKey: 399-3996-1-ND)
- planning to use modded SB0550 and analysis software to measure LM3886 performance, is main reason why I made the change.

ASAHI KASEI AK5394 - see page 15 and 21
http://www.asahi-kasei.co.jp/akm/en/...5394a_f03e.pdf

Crosstalk improvement results:
X-FI-MOD-110 - eSnips, share anything

Quote:

Originally Posted by eboy2003 /img/forum/go_quote.gif
bichi
The hakko 808 is around $180
mad.gif
I will have to think about it.



Yikes!
- price doubled in 2 years??
- still worth it, in time savings, preventing lifted/dinged pads, less heat and redoing joints. (think cost of replacing ruined stuff...)
- oh wait! - you're a soft-handed software GURU, never mind... - LOL!
FRYS.com | Hakko

Quote:

Originally Posted by eboy2003 /img/forum/go_quote.gif
bichi
Possible meet this Saturday here to swap the card to try?



Right!
- usually there from 9:30am~2pm.
- but if you are going to sell your SB0550 and go digital, is it worth the trouble?
- pssssst... I have a SB0460, expertly modded (no weak pads) to "beat" X-FI Elite, complete with sockets for sale... (joking!)
 
Mar 6, 2008 at 11:57 PM Post #1,564 of 2,194
Quote:

Originally Posted by Random Murderer /img/forum/go_quote.gif
well, i finished hotrodding my buddy's x-fi xtreme music, but he only gets audio out of the left channel... at first i thought it was because maybe the op-amp wasn't soldered on completely, so i resoldered it, then i thought it was because i removed the decoupling caps, so i replaced them to no avail. still only audio out of the left channel. What? help?


found the problem. either i had a bad opamp or i fried part of it somehow. we removed it and now his audio is a bit quiter, but even across the left and right channels. he'll hafta wait until i get my 49860's in before he gets an opamp, but he's cool with that.
 
Mar 7, 2008 at 12:21 AM Post #1,565 of 2,194
Quote:

Originally Posted by bichi /img/forum/go_quote.gif
Only if one is recording and would like improved low-freg (20~200hz) response.
- better THD and improved low-end crosstalk by 2~3db.
- no issue increasing Vref Offset Cal period from 2.4sec to 3.4sec, relative to ucode. (see AK5394 ref below)
- sonically, it reproduces better LF tracking/phase, but only if listening critically or doing delta-phase measurement.
- for most, probably not worth $16+ USD new unit price, low-quant, for Kemet T510 at that value. (DigiKey: 399-3996-1-ND)
- planning to use modded SB0550 and analysis software to measure LM3886 performance, is main reason why I made the change.



OK, I will try replacing those 4 470u caps on ADC. They are the only stock caps left (well other than rest of the channel coupling caps) anyway. Will see if that improves my stereo crosstalk issue. I use the SB0550 RMAA loopback with my LD MK3 headphone amp in preamp mode to do the tube matching when rolling tubes. I think this is the best computerized tube matching mechanism. Should patent this. LOL


Quote:

- pssssst... I have a SB0460, expertly modded (no weak pads) to "beat" X-FI Elite, complete with sockets for sale... (joking!)


biggrin.gif
Stop laughing at my ebay marketing tactics. But I do find SB0460's soldering pad is stronger than SB0550.
 
Mar 11, 2008 at 7:01 AM Post #1,567 of 2,194
Quote:

Originally Posted by Random Murderer /img/forum/go_quote.gif
is there a cheaper alternative to ers paper? it's really expensive!


Quote:

Originally Posted by Bad_dude /img/forum/go_quote.gif
Would Fish paper works as well as the ESR paper?

Thanks.



no, fish paper will not work, although i did take it upon myself to find a cheap alternative to ers. all you need is a cell phone that has service and a couple of anti-static bags(the clear gray bags that certain esd sensitive devices come in). not all anti-static bags are suitable, hence why you need the cell phone. grab a bag, put the cell phone in it, seal the bag, and if you completely lose signal, it's a suitable bag. i used an anti-static bag from a hard drive i bought a while back, dropped the card inside, and cut a slit for the pci connector to fit through. it looks ghetto, but i immediately noticed a difference in the sound, so it is working.
 
Mar 11, 2008 at 7:10 PM Post #1,568 of 2,194
eboy2k3,
- have you tried TDK MLCC 22uf @ 16vdc and/or OSCON 4.7uf @ 25vdc on U7 carrier yet?
- interesting how it changes speed/attack, imaging and mids...

U7 Carrier Version 2
- added RC from (V+) and (V-) to ground.
X-FI-MOD-111 - eSnips, share anything

Other Stuff
- missed you at Deanza meet
- more stuff to play with:
LME49600 - High Performance, High Fidelity, High Current Audio Buffer
(who needs, hot, fragile, high-voltage, vacuum-based, electron-shooters driving muffled, headache inducing head-clamps anymore? LOL!)
- give you any ideas about modding Line-Out stage with this buffer?
- might do a design this weekend, if parts get here in time....
 
Mar 16, 2008 at 7:01 AM Post #1,569 of 2,194
bichi, sorry I have been busy at work and did not come here so often. I have been fairly happy with the current opamp carrier with 0.1u ceramic and 10u tant decoupling, but will give oscon a try when I get time. Plan to get a o'scope as a serious hobbyist.
smily_headphones1.gif


I start feeling 4x22uf MLCC DAC coupling makes the high a little bit too bright and tiring, especially when listen to violin with lots of highs. Might switch back to 3x22u.

Maybe meet sometime next week at HSC again?
 
Mar 17, 2008 at 12:51 AM Post #1,570 of 2,194
Quote:

Originally Posted by eboy2003 /img/forum/go_quote.gif
bichi
"...I start feeling 4x22uf MLCC DAC coupling makes the high a little bit too bright and tiring, especially when listen to violin with lots of highs. Might switch back to 3x22u..." Maybe meet sometime next week at HSC again?



Yeah, I get the same "impression" from time to time, depending on recording/source.
- always thought stock opamp feedback circuit would need adjustment, but never took the time, until now.
- desoldered C53, C60, C63 and measured values, inserted into TI TINA Simulator and did AC tranfer and noise runs.
- stock results kinda show why it might be "bright" at high end.
- values run: C60/62 = 700pf (stock); C60/62 = 1.0nf; C60/62 = 1.5nf

Will install 1.0nf NPO and see what acutal effects occur and post RMAA's later this evening.
- for now, gotta make a quick run to Fry's before they close...

ummm, HSC might be tough this week, but maybe Thurs, Mar 20, 5:30pm...

REF:
SB0550: U7 schematic:
X-FI-MOD-112-tina - eSnips, share anything

SB0550: C60 & C62 AC and Noise Simulation Results:
X-FI-MOD-113-effect - eSnips, share anything

SB0550: TI TINA TSC sim file, zipped:
X-FI-MOD-114-tina TSC - eSnips, share anything
 
Mar 17, 2008 at 5:41 AM Post #1,571 of 2,194
Mar 17, 2008 at 3:01 PM Post #1,573 of 2,194
Quote:

Originally Posted by hyyam85 /img/forum/go_quote.gif
anyone has compared this against the x-fi prelude?


doesn't the prelude have swappable dip-chip opamps?
 
Mar 18, 2008 at 10:58 PM Post #1,575 of 2,194
Doing something with my computer now so its temporarily inoperable. Instead, I am using my Zune as source.

All I cay say is, wow.

The 30gb Zune sounds like horse**** compared to AD8599 X-Fi. Sounds so unnatural and wrong. Soundstage has no depth anymore, instrument precision and that sweet, sweet midrange is gone. Bass impact is reduced, slightly boomy. Can't say anything about highs, my music doesn't feature enough for comment. Well, actually, as I just typed that there was a high part of a electric guitar solo, it sounded like it was wearing a wool sweater. Yes, thats the only way I can describe it.

Summary: The modded X-Fi is FANTASTIC!
 

Users who are viewing this thread

Back
Top