or Connect
Head-Fi.org › Forums › Equipment Forums › Computer Audio › Hotrodding the X-Fi: A Layman's Guide (No 56k)
New Posts  All Forums:Forum Nav:

Hotrodding the X-Fi: A Layman's Guide (No 56k) - Page 74

post #1096 of 2194
Got a hold of a "free" X-FI ExtremeGamer, SB0703, installed in second machine, P5W64 WS Pro, QX6700 @ 3.0g.
- Creative X-FI Driver: v5.12.6.1187, dated 08/17/06. (same driver used in P5WDH, Q6700 @ 3.0g, running X-FI SB0460)
- both machines running XP Pro SP2
- RMAA v6.0.5 (exe dated: 06/01/07)

- same NJR4556 for Line-Out (smaller package) and same TI M33078 for MIC-IN, as SB0460.
- caps are a combo of G-Luxon 604H(M), Jamicon SS, WinCap SM.
- DC converter cap is G-Luxon 605B(M), 220uf @ 25vdc, 105c.

- has the same sonic quality of pre-mod SB0460, "muddy" bass with "harsh, low energy, tinny" highs.
- RMAA results of stock, SB0703, both 44k/16b and 96k/24b posted below.

- plan to update with lower Vs (+/-17vdc), LME49720 as opposed to previously used LME49860.
- update DC converter cap to same Panasonic FK, 1500uf @ 16vdc.

Stock: 44khz / 16b
http://www.esnips.com/doc/84ea4532-4...I-MOD-015b-XFG

Stock: 96khz / 24b
http://www.esnips.com/doc/16c3f3df-1...I-MOD-016b-XFG


RMAA tests done by Creative, with RMAA v5.5:
http://audio.rightmark.org/downloads...MAA%20v5.5.pdf
- as additional comparison reference, prior to modifications:

44/16
http://www.esnips.com/doc/dbcedbb4-8...I-MOD-018b-XFG

96/24
http://www.esnips.com/doc/3d339799-6...I-MOD-017b-XFG

X-FI SB0703 (ExtremeGamer)
http://www.esnips.com/doc/25625c6e-c...I-MOD-014b-XFG
post #1097 of 2194
Quote:
Originally Posted by bichi View Post
Got a hold of a "free" X-FI ExtremeGamer, SB0703, installed in second machine, P5W64 WS Pro, QX6700 @ 3.0g.
- Creative X-FI Driver: v5.12.6.1187, dated 08/17/06. (same driver used in P5WDH, Q6700 @ 3.0g, running X-FI SB0460)
- both machines running XP Pro SP2
- RMAA v6.0.5 (exe dated: 06/01/07)

- same NJR4556 for Line-Out (smaller package) and same TI M33078 for Line-In as SB0460.
- caps are a combo of G-Luxon 604H(M), Jamicon SS, WinCap SM.
- DC converter cap is G-Luxon 605B(M), 220uf @ 25vdc, 105c.

- has the same sonic quality of pre-mod SB0460, "muddy" bass with "harsh, low energy, tinny" highs.
- RMAA results of stock, SB0703, both 44k/16b and 96k/24b posted below.

- plan to update with lower Vs (+/-17vdc), LME49720 as opposed to previously used LME49860.
- update DC converter cap to same Panasonic FK, 1500uf @ 16vdc.

44khz / 16b
http://www.esnips.com/doc/84ea4532-4...I-MOD-015b-XFG

96khz / 24b
http://www.esnips.com/doc/16c3f3df-1...I-MOD-016b-XFG

X-FI SB0703 (ExtremeGamer)
http://www.esnips.com/doc/25625c6e-c...I-MOD-014b-XFG
Your sound assesment agrees with mine. good portion of the treble defects solved with D.C. coupling here. The rest was bypassing the power sypply caps near the DAC for me. No opamp change reqired at my end here.

By the way I like to say you gave an excellent tutoriial though if someone did want to chng the opamps though
post #1098 of 2194
Quote:
Originally Posted by LawnGnome View Post
But that doesn't explain why a better spec'd opamp would lead to lower results.

A stock card measures better than a card with the upgraded opamp.

This leads me to think that the amp stage is more complicated that people expect, and a simple drop in part needs more research.



The card SHOULD measure better after the opamp upgrade, but it doesn't. The ADC or DAC wont matter, since they stay unchanged.
Some of the things that might explain a possable reason for lower performance of the "better" part is the "better" part was designed to work with 600 or more ohms at the output. This resistance needs to be as close to the output of the opamp as possable to shield it from excessive capacitance that it might see with the cabling. This is if indeed there is a decrease in performance specwise compared to his card stock.

The X-Fi soundcards are designed to be able to drive headphones through the line-out & as a consequence do not have the required resistance for the "beter" opamp to work into so the 'better" opamp is being stressed beyond it's capabilities current wise hence an increase in distortion is possable. The opamp that came with the X-Fi was designed to be able to drive headphones at least at modest levels so the sensitivity to capacitance in the cabling may be less than the "better" part.

This is one of the primary reasons I didn't change my opamps out. The other is I didn't have the confidence to be able to do the job myself in this regard though after reading Bichi's excellent tutorial I do believe I could have done it, but I'm not going to for the first reason.
post #1099 of 2194
Updated post #1096 above
- added RMAA v5.5 tests done by Creative as additional "pre-mod" references.
- note test result deviations, but all within nominal "tolerance-stackup" ranges.
- (technical deviations are less than .01%, on average, which are excellent, given mass-produced, consumer grade audio)
- but the real object is to improve "subjective" audio quality, or at least "con" ourselves into "feeling," the improvements are real....
- loads of afternoon fun and worth the entertainment value, eh?

Quote:
Originally Posted by germanium View Post
"...This is one of the primary reasons I didn't change my opamps out. The other is I didn't have the confidence to be able to do the job myself in this regard though after reading Bichi's excellent tutorial I do believe I could have done it, but I'm not going to for the first reason...."
- chicken****! (LOL)
- I've got 5yrs on yer butt and still did it....
post #1100 of 2194
Quote:
Originally Posted by bichi View Post
Updated post #1096 above
- added RMAA v5.5 tests done by Creative as additional "pre-mod" references.
- note test result deviations, but all within nominal "tolerance-stackup" ranges.
- (technical deviations are less than .001%, on average, which are excellent, given mass-produced, consumer grade audio)
- but the real object is to improve "subjective" audio quality, or at least "con" ourselves into "feeling," the improvements are real....
- loads of afternoon fun and worth the entertainment value, eh?
Believe me if the mods are good mods the improvement is not a con, they are real. By the way I dicided to go back & rebypass the +&- 12 volt caps on the Elite Pro & ran the tests again . I found the the bypasses reduced IM distortion above 15KHz. It dropped from -102db to -106db but only above 15KHz. The sound is slightly brighter as I reported before but I'm finding it livable for a time till I can get bigger metalyzed film bypass caps. Something on the order of 10uf. Right now it is only 3.3uf on each. This should solidify the bass more & have more body throughout the spectrum without losing the beautifull detail I now have. The detail & soundstage is excellent especially with the bypasses on the +/- 12volt rails.

No, not chicken**** but still not going to do it for the first reason. If I was chicken I wouldn't have done any mods but then I couldn't live with myself knowing there was more potential.
post #1101 of 2194
Looked at Creatives RMAA results for my card & mine are exacly the same in frequency response & all distortion specs. I get more than 3db better signal to noise & dynamic range. My noise spectrum in the distortion tests are radically cleaner looking. In the THD graph my second harmonic distortion is higher but everything else is lower. Thier IMD is a quite a bit higher than mine above 20Khz. This could be audible as IMD products tend get dumped back into the audible range. My stereo seperation is better by a lot, 10db in fact.

Thanks for the link to Creatives tests Bichi.
post #1102 of 2194
Quote:
Originally Posted by germanium View Post
"...Believe me if the mods are good mods the improvement is not a con, they are real. By the way I dicided to go back & rebypass the +&- 12 volt caps on the Elite Pro & ran the tests again . I found the the bypasses reduced IM distortion above 15KHz. It dropped from -102db to -106db but only above 15KHz. The sound is slightly brighter as I reported before but I'm finding it livable for a time till I can get bigger metalyzed film bypass caps. Something on the order of 10uf. Right now it is only 3.3uf on each. This should solidify the bass more & have more body throughout the spectrum without losing the beautifull detail I now have. The detail & soundstage is excellent especially with the bypasses on the +/- 12volt rails.
- damn bastid! What started out as a "quicky" is turning into a project.
- might have to slap in a PCI extender and probe around with scope...
- (nagging thoughts, back of head, about possible ringing at upper freqs with wider-band op amp, inside unknown circuit design)
- at the "obsessive" end, might have to buy/build SOIC socket/flipchips and try different op amps (have AD8599s staring at me)

Yeah, DC supply impulse performance is another curiosity.
- but do agree with Cyril Bateman's and Rod Elliott's theories on parallel power-cap bypassing and will probably replace with larger values, as opposed to adding smaller bypass caps across larger valued caps.
- counter-intuitive to me, since my standard practice is to add 1uf tantalum power bypass, close to target chip and/or local DC regulator, in the digital world...

Quote:
Originally Posted by germanium View Post
"...No, not chicken**** but still not going to do it for the first reason. If I was chicken I wouldn't have done any mods but then I couldn't live with myself knowing there was more potential...

- so you let me be the "sacrificial" rat-boi, eh?
- still think yer a chicken-poo, LOL! (its much easier with solder-paste with SOIC sized parts, than with wire solder...)
- I see yer located in Tukwila. Might have to look you up, next time I'm at Microsoft campus...
- prime rib dinner, on me, IF you bring your X-FI with modded SOIC op amp...
post #1103 of 2194
Quote:
Originally Posted by bichi View Post
- damn bastid! What started out as a "quicky" is turning into a project.
- might have to slap in a PCI extender and probe around with scope...
- (nagging thoughts, back of head, about possible ringing at upper freqs with wider-band op amp, inside unknown circuit design)
- at the "obsessive" end, might have to buy/build SOIC socket/flipchips and try different op amps (have AD8599s staring at me)

Yeah, DC supply impulse performance is another curiosity.
- but do agree with Cyril Bateman's and Rod Elliott's theories on parallel power-cap bypassing and will probably replace with larger values, as opposed to adding smaller bypass caps across larger valued caps.
- counter-intuitive to me, since my standard practice is to add 1uf tantalum power bypass, close to target chip and/or local DC regulator, in the digital world...
Larger electrolytics are counterproductive generally I've found but bigger bypasses with metalyzed film works wonders for me every time. Note in the post above that my IMD is lowered compared to stock especially above 15KHz. Even below 15KHz it is slightly better. Bigger electrolytics may actually increase IMD in this case unless it is a stacked electrolytic & these are very expensive & usually quite large & industrial. Too large for this application. Polymer electrolytics may help though. They are claimed by Sanyo to have near metalyzed film properties.

You should at least try the bypass method, It is easily reversed unlike the opamp swap.
post #1104 of 2194
Quote:
Originally Posted by germanium View Post
"...Polymer electrolytics may help though. They are claimed by Sanyo to have near metalyzed film properties..."
Agreed
- only consideration for me was low-ESR, solid electrolytic types, possibly at higher wvdc for lower res freqs, if physical size remains reasonable....
- will have to compare results, now that we have a common, third-party standard.
- caps and bypass only, compared with op amp, no-bypass and larger value, power cap solid electrolytics

For my information:
- what X-FI driver version?
- OS?
- RMAA rev?
- MB model, processor and speed?
- PSU?
post #1105 of 2194
Quote:
Originally Posted by bichi View Post
Agreed
- only consideration for me was low-ESR, solid electrolytic types, possibly at higher wvdc for lower res freqs, if physical size remains reasonable....
- will have to compare results, now that we have a common, third-party standard.
- caps and bypass only, compared with op amp, no-bypass and larger value, power cap solid electrolytics

For my information:
- what X-FI driver version?
- OS?
- RMAA rev?
- MB model, processor and speed?
- PSU?
2_13_0012
Windows Vista Ultimate 64 bit version on 750GB RAID 0 drive. I also have Windows XP sp2 on a 400GB RAID 0 drive.
RMAA 6.05
Abit AB9Pro, Core2Duo E6600 overclocked to 3.006 GHZ & 4GB memory.
Pure power 480

By the way I have the Elite Pro which is a different card altogether than you have. Different DACs & ADCs
post #1106 of 2194
Quote:
Originally Posted by germanium View Post
2_13_0012
Windows Vista Ultimate 64 bit version.
RMAA 6.05
Abit AB9Pro, Core2Duo E6600 overclocked to 3.006 GHZ & 4GB memory.
Pure power 480
More additional work for me....
- will have install Vista 64 Ultimate so our tests are "apples-apples..."
- don't know for sure, but rumors say Vista series have "cleaner" audio core than XP SP2?

Quote:
Originally Posted by germanium View Post
By the way I have the Elite Pro which is a different card altogether than you have. Different DACs & ADCs
- yep, assumed that to be the case from your past posts... SB0550, right?
post #1107 of 2194
Quote:
Originally Posted by bichi View Post
More additional work for me....
- will have install Vista 64 Ultimate so our tests are "apples-apples..."
- don't know for sure, but rumors say Vista series have "cleaner" audio core than XP SP2?
Makes no difference though if you use the bit perfect playback option though but yes it is cleaner than XP sp2. You can also control the core sample rate in Windows Vista but make sure that both the recording & the playback are the same sample rate or you will see massive degradation. Test with RMAA only in music creation mode or you could pay the price of crashing window sound & have to reinstall the driver like I had to once. Test will not work at all in entertainment mode & Mine crashed the driver in gaming mode.
post #1108 of 2194
Quote:
Originally Posted by germanium View Post
Makes no difference though if you use the bit perfect playback option though but yes it is cleaner than XP sp2. You can also control the core sample rate in Windows Vista but make sure that both the recording & the playback are the same sample rate or you will see massive degradation. Test with RMAA only in music creation mode or you could pay the price of crashing window sound & have to reinstall the driver like I had to once. Test will not work at all in entertainment mode & Mine crashed the driver in gaming mode.
Interesting info, thanks...
- no "audio creation" mode, eh?
- all my RMAA tests were done in "audio-creation" mode, with all effects "off"
- off to dinner and maybe some more X-FI fun, later this evening...
post #1109 of 2194
Quote:
Originally Posted by bichi View Post
More additional work for me....
- will have install Vista 64 Ultimate so our tests are "apples-apples..."
- don't know for sure, but rumors say Vista series have "cleaner" audio core than XP SP2?



- yep, assumed that to be the case from your past posts... SB0550, right?
My tests come out exactly the same in windows Vista & Windows XP sp2 if the bit perfect box is checked but radically worse in XPsp2 if not.

Doesn't list the model number in Windows Vista but yes that is the card.
post #1110 of 2194
Quote:
Originally Posted by bichi View Post
Interesting info, thanks...
- no "audio creation" mode, eh?
- all my RMAA tests were done in "audio-creation" mode, with all effects "off"
- off to dinner and maybe some more X-FI fun, later this evening...
Audio creation is what I meant, sorry. I was just trying the different modes but audio creation did in fact work.

Have a good dinner ok.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Computer Audio
Head-Fi.org › Forums › Equipment Forums › Computer Audio › Hotrodding the X-Fi: A Layman's Guide (No 56k)