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 79

post #1171 of 2194
Well, I have another 40 opamps coming of various specs. Most for my amp though.
post #1172 of 2194
Quote:
Originally Posted by LawnGnome View Post
Well, I have another 40 opamps coming of various specs. Most for my amp though.
Wow, you have a lot of soldering ahead of you Lawn Gnome
post #1173 of 2194
Cheapo SOIC Socket-Hack done.
- no more slobbering on X-FI SMT pads and feeling guilty about cutting legs off "good" op amps. - (weeeeee!)
- SMT pads on board seem to be robust enough to handle insertion forces. (very light with this mod, compared to DIP8)
- will most likely add some epoxy for added safety.

AD8599 is a very nice sounding op amp, compared to LME49860/720
- first impression is, ummm, "more classical" than LME's, with same low and high clearity.

RMAA results show no obvious anomalies and slightly better distortion figures.
- will leave it to Germanium to critique

SOIC socket hack - parts:
http://www.esnips.com/doc/90a2332b-8.../X-FI-MOD-032b

Finished op amp "module:"
http://www.esnips.com/doc/f87daa51-8.../X-FI-MOD-033b

SOIC socket hack - AD 8599 installed in Line-Out position:
http://www.esnips.com/doc/a80ab9fb-c.../X-FI-MOD-034b

SB0460 with AD8599 op amp, Line-Out - sample rate: 96k - tantalum bypass - op amp SOIC "socket hack"
RMAA Results: http://www.esnips.com/doc/202603be-5.../X-FI-MOD-035b
RMAA Detailed: http://www.esnips.com/doc/d13bdb7a-3.../X-FI-MOD-036b
post #1174 of 2194
Audigy 4 (non-pro) hotrod results:

100uf changed to 470uf Samxon
47uf's piggy backed with cheapo 100uf caps.
4x 22uf near front channel opamp changed to Samxons
Opamp changed to LM4562
Rear of card shielded

24bit/96khz is not much different to stock. 16bit/44khz has improved dramatically from stock though (THD and IMD was about twice as much)
post #1175 of 2194
Quote:
Originally Posted by bichi View Post
Cheapo SOIC Socket-Hack done.
- no more slobbering on X-FI SMT pads and feeling guilty about cutting legs off "good" op amps. - (weeeeee!)
- SMT pads on board seem to be robust enough to handle insertion forces. (very light with this mod, compared to DIP8)
- will most likely add some epoxy for added safety.

AD8599 is a very nice sounding op amp, compared to LME49860/720
- first impression is, ummm, "more classical" than LME's, with same low and high clearity.

RMAA results show no obvious anomalies and slightly better distortion figures.
- will leave it to Germanium to critique

SOIC socket hack - parts:
http://www.esnips.com/doc/90a2332b-8.../X-FI-MOD-032b

Finished op amp "module:"
http://www.esnips.com/doc/f87daa51-8.../X-FI-MOD-033b

SOIC socket hack - AD 8599 installed in Line-Out position:
http://www.esnips.com/doc/a80ab9fb-c.../X-FI-MOD-034b

SB0460 with AD8599 op amp, Line-Out - sample rate: 96k - tantalum bypass - op amp SOIC "socket hack"
RMAA Results: http://www.esnips.com/doc/202603be-5.../X-FI-MOD-035b
RMAA Detailed: http://www.esnips.com/doc/d13bdb7a-3.../X-FI-MOD-036b
Slightly higher low order THD on LME49680 but way less high order spurie. Less noise when signal present with LME49680.

definately lower IMD swept distortion on AD8599 & very definately cleaner IMD spectrum. LME49680 still has slighly lower noise floor in spectum analysis of IMD.

Crosstalk worse on AD8599 but only in one direction.

Since IMD distortion is likely to be more audible than THD the AD8599 would seem to have the upper hand though the high order spurie does cause me slight concern on the THD.

Noise with no signal slightly higher on AD8599. Spurie in noise with no signal slightly higher on on LME49680.

The LME49680 appears to have the classic lower negative feedback distortion characteristics. Higher low order distortion & cleaner high order which acording to the accepted thought on feedback & distortion the LME49680 would have the upper hand & sound less clinical. Is that what you mean by discribing the AD8599 as more "classical" as in more clinical like older opamp designs Bichi??
post #1176 of 2194
The fact that the distortion characteristics are plainly visable & different, even with such a low grade ADC handicapping the design, it appears that the LME49680 is not the low distortion part it is claimed to be.

I would try to get the NJM2114 to compare as that is what comes stock on my card. It seems to actually have better distortion characteristics than either of these in this circuit.
post #1177 of 2194
Quote:
Originally Posted by germanium View Post
Yes I was going to try that & see if it helps here . It has gernerally worked before. This is bypassing the 2 -220uf caps by the opamps I tried it before but never looked closely at the IM distortion which improved in good portion. Even though the bass is very solid now, better than stock , I feel there is still room for improvement. Looking to put about 10uf here. You see my tests Ter1?

I guess I have been very busy on here lately haven't I??

Hi Germanium, can you have a summary of all your mod of x-fi elite pro in one post? Of course you can modify it based on your new founds.

Some other questions,

1. I remember you said after you bypassed the 2 220uf caps by the opamps, the highs are too bright, so you don't suggest doing that? You have changed you mind on this? You suggest 1 10uf metalized caps now for the 2?

2. The 2 220uf by the opamps are for power supply? If so, BG will be good choice, right? The original is 25v 220uf, can I replace them with 16v 1000uf BG ones? The voltage is okay?

3. What are the functions of the 2 columns of 25v 100uf and 16v 47uf caps? For power supply use or other use? What caps do you suggest besides BG caps that I can replace them with the same voltage and capacitance?

4. For question 3, do you still suggest "100" be bypassed with 2 3.3uf and "47" be bypassed with 1 3.3uf ?

5. Do you still suggest bypass the 220uf power supply cap near DSP with1-3.3uf metalized film cap?

Thanks! You are so welcome here!~
post #1178 of 2194
Quote:
Originally Posted by germanium View Post
"...The LME49680 appears to have the classic lower negative feedback distortion characteristics. Higher low order distortion & cleaner high order which acording to the accepted thought on feedback & distortion the LME49680 would have the upper hand & sound less clinical. Is that what you mean by discribing the AD8599 as more "classical" as in more clinical like older opamp designs Bichi??..."
Quote:
Originally Posted by germanium View Post
The fact that the distortion characteristics are plainly visable & different, even with such a low grade ADC handicapping the design, it appears that the LME49680 is not the low distortion part it is claimed to be.
I would try to get the NJM2114 to compare as that is what comes stock on my card. It seems to actually have better distortion characteristics than either of these in this circuit.
Ge32,

Thanks for the detailed analysis!
- guess I'll have to send you two sockets and a few op-amp "carriers" so you can enjoy easy op-amp swaps, eh?
- PM with mailing data, if interested.

Yes, "classical" as in clinical. Refrained with accuracy and "blending," as opposed to, ummm, "excited."
- would be interesting to see how a LME49840/760 would spec on your SB0550 and its digital section...
- could mount a LME and NJM2114 on two carriers, provided you manage to get a socket soldered on.

Completed an analysis of various single freq, whitenoise digital playbacks and RMAA.
- see no ill effects of digital section power (tant) and socket modifications.
- beginning to wonder if "technical" improvement limit is the DAC/ADC, and not so much the op-amps.

Have ordered a handfull of NJM2114's and should be here in a week or so.

Also pondering "pinning" coupling cap terminals, for easy shorting or cap coupling with various types.
post #1179 of 2194
Quote:
Originally Posted by ter1 View Post
Hi Germanium, can you have a summary of all your mod of x-fi elite pro in one post? Of course you can modify it based on your new founds.

Some other questions,

1. I remember you said after you bypassed the 2 220uf caps by the opamps, the highs are too bright, so you don't suggest doing that? You have changed you mind on this? You suggest 1 10uf metalized caps now for the 2?

2. The 2 220uf by the opamps are for power supply? If so, BG will be good choice, right? The original is 25v 220uf, can I replace them with 16v 1000uf BG ones? The voltage is okay?

3. What are the functions of the 2 columns of 25v 100uf and 16v 47uf caps? For power supply use or other use? What caps do you suggest besides BG caps that I can replace them with the same voltage and capacitance?

4. For question 3, do you still suggest "100" be bypassed with 2 3.3uf and "47" be bypassed with 1 3.3uf ?

5. Do you still suggest bypass the 220uf power supply cap near DSP with1-3.3uf metalized film cap?

Thanks! You are so welcome here!~
1.Yes going to try the bigger metalyzed film caps on the 2-220uf caps. can only get 12uf caps at the local store so going to pick them up this weekend. Yes I have changed my mind & am going to leave these bypassed. See below for reason.

2. BG might be better if you arn't going to bypass the power caps but a waste of time & money if you are as the film bypasses are probably substantially superior.

3. 4 of the 47uf caps serve as a ref voltage supply & may not be critical. The 100uf caps are to the intermal regulated supply on the DAC & appear to be critical. Do not change these but bypassing seems to help the clearity. The red marked caps are definately coupling caps & can be shorted.

4.yes definately on the 100uf cap not sure whether the 47uf cap bypass matters but leaving mine bypassed with 1 3.3uf cap.

5. Yes but going to try bigger 12uf bypasses there as well as there may be an improvement there. There was definately a drop in IMD swept distortion at 15KHZ & above though even with just the 3.3uf bypass caps there. Sound is slightly brighter but not overwhelmingly bright. There is definately a fair amount more detail there.

If you look at my tests & compare them to Creatives own tests mine are subtantially better. Bypassing input coupling caps to the ADC helps the noise floor of the ADC & slightly flattens the low frequence response of the ADC. Creative poor noise result is likely because of these caps.
post #1180 of 2194
Quote:
Originally Posted by bichi View Post
Ge32,

Thanks for the detailed analysis!
- guess I'll have to send you two sockets and a few op-amp "carriers" so you can enjoy easy op-amp swaps, eh?
- PM with mailing data, if interested.

Yes, "classical" as in clinical. Refrained with accuracy and "blending," as opposed to, ummm, "excited."
- would be interesting to see how a LME49840/760 would spec on your SB0550 and its digital section...
- could mount a LME and NJM2114 on two carriers, provided you manage to get a socket soldered on.

Completed an analysis of various single freq, whitenoise digital playbacks and RMAA.
- see no ill effects of digital section power (tant) and socket modifications.
- beginning to wonder if "technical" improvement limit is the DAC/ADC, and not so much the op-amps.

Have ordered a handfull of NJM2114's and should be here in a week or so.

Also pondering "pinning" coupling cap terminals, for easy shorting or cap coupling with various types.
Sounds like my analysis was correct for the sound of the AD8599 from the graphs you provided.

I also noticed an wild variation of open loop phase distortion in the audio band in the graphs provided by Analog Devices which may be contributing to thier less favorable high order distortion plot but on the other hand this may be a case of AD being more honest than many others in the open loop plots. AD provided substantially more detail about a lot of things including what percentage of thier yields tested to within certain tolereances.

Let me know how thing go in your tests, ok.
post #1181 of 2194
bichi

Yes the DAC is our limit. Our opamp mods or cap mods won't better the RMAA numbers.

BUT, changing the opamp can change qualities RMAA doesn't measure, like detail.

Look for an opamp with higher slew rate. 50-150 V/us, and lower settling time (500 to 50ns). This will help bring out detail.


I have to change this damn OPA2132 though, the sound is nice and detailed, but lacks the impact and brightness I need. Way to mellow for me.
post #1182 of 2194
Last of the power-cap mods for me, then, on to coupling caps (for SB0460):
- changed C101 and C124 to 33uf @ 16vdc TANT from Jamicon 22uf @ 16vdc.
(ref: C1, DVDD and C3, AVDD on WM8775 datasheet, page 37, figure 29)
- changed C115 and C123 to 6.8uf @ 16vdc TANT from Wincap 10uf @ 16vdc.
(ref: C5, ADCREFP and C7 VMIDACD on WM8775 datasheet, page 37, figure 29)

Noticable increased bass definition and power, without being muddy. Highs remain tight, clear...
- listened with both LME49860 and AD8599, installed as Line-Out.

When attempting to install SOIC socket for Line-In op amp and inspecting with eye-loupe, discovered:
- only 1 of two amps is active, where second amp's inputs are grounded.
- appears it's function is mono "mic" preamp only. (original MC33078 op amp on SB0460)
- appears Line-In/Mic jack, when in stereo-mode, goes direct to Wolfson WM8775 ADC, coupled with two Jamicon 22uf NP's
- so careful, most of the assumptions in this thread about Line-In op amp are WRONG

*yawn* - time to make some coffee and get out the door...
- RMAA, pics, later this evening....

REF:
Wolfson WM8775 Datasheet:
http://www.wolfsonmicro.com/uploads/.../en/WM8775.pdf
post #1183 of 2194
Quote:
Originally Posted by germanium View Post
Sounds like my analysis was correct for the sound of the AD8599 from the graphs you provided.

I also noticed an wild variation of open loop phase distortion in the audio band in the graphs provided by Analog Devices which may be contributing to thier less favorable high order distortion plot but on the other hand this may be a case of AD being more honest than many others in the open loop plots. AD provided substantially more detail about a lot of things including what percentage of thier yields tested to within certain tolereances.

Let me know how thing go in your tests, ok.
AtomicNumber32,

- will do....
- interesting you mention "phase distortion" on AD8599
- could have sworn a bassoon "moved" on a fade out, during a listening test (wrote it off as my imagination...)
post #1184 of 2194
Quote:
Originally Posted by LawnGnome View Post
bichi
Yes the DAC is our limit. Our opamp mods or cap mods won't better the RMAA numbers. BUT, changing the opamp can change qualities RMAA doesn't measure, like detail. Look for an opamp with higher slew rate. 50-150 V/us, and lower settling time (500 to 50ns). This will help bring out detail. I have to change this damn OPA2132 though, the sound is nice and detailed, but lacks the impact and brightness I need. Way to mellow for me.
Yep, was always an "assumed," but did not know where the limit was.
- now I know, I think, what the practical bounds are.

Ummm, not sure high-slew instrumentation class op amps would behave in audio circuits, stability and ringing at upper freqs. (beyond audio)
- will take your word on OPA2132 sounding, ummm, "instrumentation" like....

EDIT:
Ahhh, never-mind
- just looked at datasheet and TI mentions "audio," along with instrumentation stuff.....
- FET, eh?
- January 1995?
post #1185 of 2194
Quote:
Originally Posted by bichi View Post
AtomicNumber32,

- will do....
- interesting you mention "phase distortion" on AD8599
- could have sworn a bassoon "moved" on a fade out, during a listening test (wrote it off as my imagination...)
When I said open loop that is without any negative feedback to correct distortion. Having a fair degree of phase distortion open loop can throw some little monkey wrenches into the works so that while the feedback can correct some distortion prducts, it wouldn't be able to correct as much as it would normally without the phase distortion being there. High order distortion would likely be harder to correct than low order in this scenerio.

By the way lower feedback amps generally do have better sound in spite of increased distortion as the distortion tends to be mostly low order. Which likely explains the better sound that people are reporting with the LME49720. It seems to have the classic low order distortion characteristics that are less audible & hence sound more accurate to the ear.

My friend Jacks amps had 0 feedback & only about .02% distortion @ half power which was almost exclusively 3rd order. Second order distortion being cancelled out by the closely matched push-pull output section. This amp was very wide band & not just at milliwatt output as some tube amp builder like to quote but at full power. This amp was very very lively sounding with excellent dynamics that put the likes of Krell & Mark Levinson to shame in direct comparisons. I should know because I was there in those comparisons. The sound of the big tranny amps was restrained dynamically compared to the 0 feedback tube amps Jack built. By the way there was 0 phase distortion in the audio band on these amps & was like only 1-2 degrees off at 30,000 cycles. A far cry from the wild 30 degree variations of the AD8599.
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)