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 86

post #1276 of 2194
Quote:
Originally Posted by germanium View Post
Only signicant harmonic that increased was the second harmonic as expected, All other are either the same or lower.
IMD swept did seem to increase across the board. It seems related to the increased second harmonic distortion however when you look at the IMD spectrum analysis. You will notice that the IMD & IMD swept increased by approximately the same amount as the THD & the THD increase was 100% second harmonic only increase. While I wasn't quite expecting an increase in IMD I was expecting the THD increase & the fact that it was solely second harmonic increase &that the IMD in fact mirrored that increase.
frequency response slightly improved with D.C. coupling.
Crosstalk definately improved toward higher frequencies especially.
Difference in dynamic range though visable is too hard to tell which is better by the graph.
Overall even though there is a slight increase in distortion there is an overall more optimal distortion characteristic than the cap coupled cousin.
Lastly you D.C. offset seems higher than usual at least compared to what I've seen with the stock opamp
Thank you for the excellent analysis, as usual!
- will have to think about the (-0.253vdc) DC bias, with shorted Line-Out coupling caps....
- package in the mail to you tomorrow...
- check out next post... (dang OSCON solids might be a good thing for audio power caps...)
post #1277 of 2194
Sept 14, 2007 - SB0460
- Test OSCON Polymer ultra-low ESR/Impedance for noise reduction over low-ESR type.
- Change: C46, OSCON F8, 330uf @ 6.3vdc from Panasonic FC, 220uf @ 16vdc <-- (see note below)
- Change: C177, OSCON F8, 330uf @ 6.3vdc from Panasonic FK 1500uf SMT @ 16vdc

note: 6.3vdc voltage rating margin too low for +5vdc environment and not recommended. Part was on hand from samples and installed for experimental purposes.


Got a bit of noise and IMD improvement....
And the sound quality, OooLaaLaa!

48/24/48 - Comparsion with OSCON solid-polymer power caps vs., low-ESR, similar to Panasonic FC/FK, same configuration:
Frequency response (from 40 Hz to 15 kHz), dB +0.01, -0.08 (+0.01, -0.09)
Noise level, dB (A) -101.7 (-101.6)
Dynamic range, dB (A) 101.7 (101.5)
THD, % 0.0014 (0.0015)
THD + Noise, dB (A) -92.3 (-91.6)
IMD + Noise, % 0.0029 (0.0033)
Stereo crosstalk, dB -98.4 (-98.3)
IMD at 10 kHz, % 0.0063 (0.0073)

Picture:
http://www.esnips.com/doc/cef37c13-2.../X-FI-MOD-058b

RMAA v6.0.5 Results:
Summary: http://www.esnips.com/doc/daeb977a-8.../X-FI-MOD-059b
Details: http://www.esnips.com/doc/122b8afc-c.../X-FI-MOD-060b
post #1278 of 2194
Okay first time poster and sound card modder here so howdy everyone. I followed the mod changing out the the cap for the 2200uf 16v blackgate and it works noticed a differance imediatly .Im going to wait for burn in before I change the opamps out so I can see/hear for a differance. My other card I just shielded no other mod and I can hear a differance also softer sounding warmer I can crankit without any annoying sounds like b4. I noticed bichi has used differant caps smaller I think panasonics? any differance Im asking because the blackate is huge and I dont like it hanging all out I cant stand it up I have no room.
post #1279 of 2194
Quote:
Originally Posted by the.oldman View Post
Okay first time poster and sound card modder here so howdy everyone. I followed the mod changing out the the cap for the 2200uf 16v blackgate and it works noticed a differance imediatly .Im going to wait for burn in before I change the opamps out so I can see/hear for a differance. My other card I just shielded no other mod and I can hear a differance also softer sounding warmer I can crankit without any annoying sounds like b4. I noticed bichi has used differant caps smaller I think panasonics? any differance Im asking because the blackate is huge and I dont like it hanging all out I cant stand it up I have no room.
You can bend the Black Gate over without any ill effects. However you won't be able to if you've already made the leads very short obviously.
post #1280 of 2194
Well, my card setup is doing really nice right now. The AD8066 sounds great, very quick opamp when compared to the slow slew rate of the recommended national opamp. (180v/µs vs 20v/µs)

I should hopefully have some Chemicon PSA series polymer caps coming soon to put in the analog section.
post #1281 of 2194
Quote:
Originally Posted by the.oldman View Post
Okay first time poster and sound card modder here so howdy everyone. I followed the mod changing out the the cap for the 2200uf 16v blackgate and it works noticed a differance imediatly .Im going to wait for burn in before I change the opamps out so I can see/hear for a differance. My other card I just shielded no other mod and I can hear a differance also softer sounding warmer I can crankit without any annoying sounds like b4. I noticed bichi has used differant caps smaller I think panasonics? any differance Im asking because the blackate is huge and I dont like it hanging all out I cant stand it up I have no room.
I'd save your money and get panasonic FM capacitors instead. The blackgates aren't really needed in most of the positions on the card.

Only possible place you could benefit from them is in the coupling stage.

I'd have to say I doubt the cap change for the APU with the blackgate would make an audible difference, if ANY at all.

That cap smooths the voltage going to the APU, which is 100% digital. The only way the blackgate would make a difference is your psu has bad voltage regulation which was making its way to the card, and it was bad enough to make the APU create processing errors. And if that happened, you would most likely get crashes.

I'm also pretty sure the x-fi APU implements error checking and correction.

The replacement of that cap is more for reliability and not performance.
post #1282 of 2194
Quote:
Originally Posted by the.oldman View Post
"...I noticed bichi has used differant caps smaller I think panasonics? any differance Im asking because the blackate is huge and I dont like it hanging all out I cant stand it up I have no room.
Welcome to the monkey-house!

- Experimental stuff I am doing, so don't pay too much attention to parts I am using, relative to parts recommended in this thread. My parts change every other day or so....

Some notes/suggestions:
- Diameter for C177 (power cap) on an SB0460 is 10mm (0.40inch), for it to fit, standing up.
- PCI interslot spacing is about 15mm (0.60inch), so try to say below 14mm (0.55inch) height.

- Most suggest Blackgates, but even an "N," 470uf @ 16vdc is 16x24mm and "standard" 2200uf @ 16vdc is 16x31.5mm.
- Mounting and keeping leads short as possible is the "creative" part...

- Power capacitor(s) I am currently playing with are Sanyo OSCON series, which are "solid polymer" electrolytics.
- OSCON part with highest value for "drop-in," would be SEP series, 330uf @ 16vdc, F13 size.

Selection of capacitor type is subjective and depends on your tastes, budget, etc...

Theory: (per Sanyo OSCON "Technical Marketechture...")
- objective is to attenuate as much power-supply noise with lowest ESR/Impedance.
- OSCON's pack more capacitance in smaller packages, with extremely low ESR/Impedance, compared to any aluminum or tantalum electrolytics.
- due to polymer electrolytic and its ESR abilities, theory says a 330uf OSCON is more effective than a 2000uf aluminum.
- (thus, all the fun testing...)

As an alternative, Panasonic, Elna, Nippon-Chemicon, etc., all make excellent low-ESR aluminum capacitors.
(Panasonic FK SMT specs below, as an example)
- had a Panasonic FK series, SMT, 1500uf @ 16vdc, (12.5x13.5mm), installed, on its side, prior to current experimental OSCON. (lower right corner) http://www.esnips.com/doc/beb79ee4-b.../X-FI-MOD-055b

REFs:
Blackgate: http://www.kyoto-electro.com/nx.html
OSCON Characteristics: http://us.sanyo.com/industrial/elect...n_noisecap.pdf
OSCON Size Charts: http://us.sanyo.com/industrial/elect...ystemtable.pdf
Panasonic FK SMT: http://industrial.panasonic.com/www-...BA0000CE88.pdf
post #1283 of 2194
Don't forget Nichicon or Rubycon!

The Rubycon MCZ series is supposed to match the ESR performance of OSCONs.
post #1284 of 2194
Hey Lawngnome well it may be my psu I power z5500's and I can set my volume lower yet it seems louder and cleaner. Call me a crack head I know I want improvement .The extreme music card Is my frankenstein my Fatality gamer will be my tru mod with pre testing and such b4 I claim victory. I dwnloaded RMMA 6.0.5 now I just have to figure out how to use it correctly.
Yukon T I did bend it over, b4 doing the cap changeout I measured my space so I knew going in it was going to be tight so I left the leads long to bend over the opposite way to what everyone else has but thats the only way itll fit. Bichi keep up the good work Because Ill probally copy /follow you down the same path of changing and rechanging things out to allways get better. ***actually it may be my motherboard its a chainteck vn4 ultra which I regrett ever getting but Im riding this pony into the ground untill the end of the year when I upgrade**
post #1285 of 2194
Quote:
Originally Posted by the.oldman View Post
Hey Lawngnome well it may be my psu I power z5500's and I can set my volume lower yet it seems louder and cleaner. Call me a crack head I know I want improvement .The extreme music card Is my frankenstein my Fatality gamer will be my tru mod with pre testing and such b4 I claim victory. I dwnloaded RMMA 6.0.5 now I just have to figure out how to use it correctly.
Yukon T I did bend it over, b4 doing the cap changeout I measured my space so I knew going in it was going to be tight so I left the leads long to bend over the opposite way to what everyone else has but thats the only way itll fit. Bichi keep up the good work Because Ill probally copy /follow you down the same path of changing and rechanging things out to allways get better. ***actually it may be my motherboard its a chainteck vn4 ultra which I regrett ever getting but Im riding this pony into the ground untill the end of the year when I upgrade**
If that's the case, then I'd order a bunch of different opamps from texas instruments and analog devices (they both have 100% free samples) and do some swapping of the opamps. Since that will be where the big improvement comes from.

Then, once you have an opamp you like, then start swapping caps. With the cap mods, your squeezing the rest of the performance out of the amp and DAC.
post #1286 of 2194
Quote:
Originally Posted by the.oldman View Post
"...Bichi keep up the good work Because Ill probally copy /follow you down the same path of changing and
rechanging things out to allways get better..."
Yeah? one more note/suggestion:
- Op Amp SOCKET/PLUG!
- got mine from local tech surplus, but can be purchased from Digikey.
- SMT pads are strong enough for SOIC type socket insertion/de-insertion forces, compared to DIP8 type sockets.
- if you have steady hands/tools, buying "headers" with more pins and cutting is cheaper.
- example: Header 20pins ($4.40) = 2 x 8-pin headers. (destroy 2-pins when cutting in half + 2 "awww sheeit" safety pins)

SAMTEC
SOCKET: (.050" (1.27mm))
CLP Series: Positions: 8 - Digikey PN: SAM1154-04-ND ($4.33)

HEADER: (Micro Strips)
FTS Series, Surface Mount: Positions: 8 - Digikey PN: SAM1159-04-ND ($3.80)
FTSH Series, Surface Mount: Positions: 8 - Digikey PN: SAM1161-04-ND ($3.80) a bit longer, but easier to handle

Digikey Catalog, Samtec section: http://dkc3.digikey.com/PDF/T073/0118-0119.pdf

Op Amp SOIC Socket Hack Pictures:
http://www.esnips.com/doc/90a2332b-8.../X-FI-MOD-032b
http://www.esnips.com/doc/f87daa51-8.../X-FI-MOD-033b
http://www.esnips.com/doc/87ab1b96-f.../X-FI-MOD-033c
post #1287 of 2194
Comparison between "common" LM358D op amp and newest LME49860
- a bit of afternoon fun...
- does not sound as bad as RMAA results imply...

48/24/48 - LM358D (2005?) vs., LME49860 (July 2007)
Frequency response (from 40 Hz to 15 kHz), dB +0.01, -0.08 (+0.01, -0.08)
Noise level, dB (A) -100.8 (-101.7)
Dynamic range, dB (A) 100.9 (101.7)
THD, % 0.015 (0.0014)
THD + Noise, dB (A) -75.1 (-92.3)
IMD + Noise, % 0.019 (0.0029)
Stereo crosstalk, dB -97.4 (-98.4)
IMD at 10 kHz, % 0.172 (0.0063)
General performance: Very good (Excellent)

REF: LM358D Datasheet:
http://www.national.com/ds.cgi/LM/LM158.pdf

RMAA TEST Results:
Summary: http://www.esnips.com/doc/6c48c141-1.../X-FI-MOD-061b
Detail: http://www.esnips.com/doc/fa885135-e.../X-FI-MOD-062b
post #1288 of 2194
Quote:
Originally Posted by bichi View Post
Comparison between "common" LM358D op amp and newest LME49860
- a bit of afternoon fun...
- does not sound as bad as RMAA results imply...

48/24/48 - LM358D (2005?) vs., LME49860 (July 2007)
Frequency response (from 40 Hz to 15 kHz), dB +0.01, -0.08 (+0.01, -0.08)
Noise level, dB (A) -100.8 (-101.7)
Dynamic range, dB (A) 100.9 (101.7)
THD, % 0.015 (0.0014)
THD + Noise, dB (A) -75.1 (-92.3)
IMD + Noise, % 0.019 (0.0029)
Stereo crosstalk, dB -97.4 (-98.4)
IMD at 10 kHz, % 0.172 (0.0063)
General performance: Very good (Excellent)

REF: LM358D Datasheet:
http://www.national.com/ds.cgi/LM/LM158.pdf

RMAA TEST Results:
Summary: http://www.esnips.com/doc/6c48c141-1.../X-FI-MOD-061b
Detail: http://www.esnips.com/doc/fa885135-e.../X-FI-MOD-062b

Just a quick question, why would you install that chip?

Its specs don't look good, and it doesn't even list slew rate or settling time.

Was it just to test?
post #1289 of 2194
Quote:
Originally Posted by bichi View Post
Comparison between "common" LM358D op amp and newest LME49860
- a bit of afternoon fun...
- does not sound as bad as RMAA results imply...

48/24/48 - LM358D (2005?) vs., LME49860 (July 2007)
Frequency response (from 40 Hz to 15 kHz), dB +0.01, -0.08 (+0.01, -0.08)
Noise level, dB (A) -100.8 (-101.7)
Dynamic range, dB (A) 100.9 (101.7)
THD, % 0.015 (0.0014)
THD + Noise, dB (A) -75.1 (-92.3)
IMD + Noise, % 0.019 (0.0029)
Stereo crosstalk, dB -97.4 (-98.4)
IMD at 10 kHz, % 0.172 (0.0063)
General performance: Very good (Excellent)

REF: LM358D Datasheet:
http://www.national.com/ds.cgi/LM/LM158.pdf

RMAA TEST Results:
Summary: http://www.esnips.com/doc/6c48c141-1.../X-FI-MOD-061b
Detail: http://www.esnips.com/doc/fa885135-e.../X-FI-MOD-062b
THD & IMD distortion almost off the charts. At least the THD has a general good direction in that it goes down as the order of harmonic goes up but that is about all I can say good about it.

The IMD swept is way out there. It is -60db at 10KHZ but skyrockets from there to -42db at 20KHz & -16db at 44KHz. Seeing as how these are for intermodulation products from such high IMD even that far above the audio band can dump back into the audio band. Fortunately as the IMD spectrum shows they seem to track with the harmonics of the 7KHz signal & continue only upwards & not downward into the heart of the midrange. This may be the reason that Bichi didn't report really bad sound in spite of the outragious distortion pattern. A different IMD test signal though could alter that though.

Noise look resonable but slight increse.

Same for Dynamics, No visable harmonics on either but slight increase in noise.

Very slight improvement in crosstalk at the midband but worse as you go higher.
post #1290 of 2194
Quote:
Originally Posted by bichi View Post
Sept 14, 2007 - SB0460
- Test OSCON Polymer ultra-low ESR/Impedance for noise reduction over low-ESR type.
- Change: C46, OSCON F8, 330uf @ 6.3vdc from Panasonic FC, 220uf @ 16vdc <-- (see note below)
- Change: C177, OSCON F8, 330uf @ 6.3vdc from Panasonic FK 1500uf SMT @ 16vdc

note: 6.3vdc voltage rating margin too low for +5vdc environment and not recommended. Part was on hand from samples and installed for experimental purposes.


Got a bit of noise and IMD improvement....
And the sound quality, OooLaaLaa!

48/24/48 - Comparsion with OSCON solid-polymer power caps vs., low-ESR, similar to Panasonic FC/FK, same configuration:
Frequency response (from 40 Hz to 15 kHz), dB +0.01, -0.08 (+0.01, -0.09)
Noise level, dB (A) -101.7 (-101.6)
Dynamic range, dB (A) 101.7 (101.5)
THD, % 0.0014 (0.0015)
THD + Noise, dB (A) -92.3 (-91.6)
IMD + Noise, % 0.0029 (0.0033)
Stereo crosstalk, dB -98.4 (-98.3)
IMD at 10 kHz, % 0.0063 (0.0073)

Picture:
http://www.esnips.com/doc/cef37c13-2.../X-FI-MOD-058b

RMAA v6.0.5 Results:
Summary: http://www.esnips.com/doc/daeb977a-8.../X-FI-MOD-059b
Details: http://www.esnips.com/doc/122b8afc-c.../X-FI-MOD-060b
Slight improvement in even low order THD ,less improvement on low odd order THD. Mixed bag above 5th order THD. Overall better though.

Too close to call on the noise graphs by sight. Have to take RMAA word on the improvement.

IMD spectrum very close but slightly improved with new caps.

IMD swept is very slightly higher according to the graphs, Strange as RMAA gives the new caps a better score than the old ones.

Getting tired Accidently closed the wrong windows

Hope this is what you were looking for Bichi.
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)