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 145

post #2161 of 2185

I will try that. Thank you.

post #2162 of 2185

 Hey,

so I put everything back to its place, exchanged the 0,1uF capacitor for the new one and.. nothing. Its dead :( Something's blown up I suppose. Thanks for your help anyway.

post #2163 of 2185

I did this mod by following the guide by ramachandra, so I used Silmic II 1500/16 for the main power filtering cap, but I also ordered the Nichicon KG 2200/63 that was recommended in this guide. How in the world do you fit that thing onto a sound card?!?!? It looks like it would be impossible. It was hard enough to fit the Silmic II (my father helped me with it, he found a small area with no components or resistors and  glued the cap sideways).

 

So my question is, have any of you actually put the Nichicon KG 2200/63 on this?

post #2164 of 2185

I did the mod, but the volume is extremely low.  sound is coming from both sides, but I can hardly hear it on maximum.  What I can hear though, sounds great and hope I can get this fixed.  Does anyone know why the volume would be low?  I've only installed the opamp.  It appears it's seated fine.  I saw someone earlier in the thread who fixed this problem, which seems identical to mine:

 

"It seems part of the AD8599 v- pin was somehow cut off even though I did not do anything like that. I just added more solder to connect that joint and it now works out perfectly"

 

but, i don't understand what this means.

 

 

edit: 

 

 

I think this pad might have come off:

 

slabeledpl3.jpg

 

I cannot find a diagram in this thread showing where it goes to.  I have no multimeter, and don't know how to use it.  Can someone just tell me where it goes?

 

edit:  found a diagram, and i don't think the V- was the problem.  I have no idea what to do.


Edited by lolwatpear - 2/5/12 at 3:08am
post #2165 of 2185

edit


Edited by Baphomet - 2/4/12 at 7:16pm
post #2166 of 2185

Hey guys, kinda skipped to the end here has there are 150 pages on this thread. Can somebody give me specs for the caps they used to replace the power filter and decoupling caps?

 

I heard 1000uf 16v cap on the power filter was good, is the correct?. On the first page there is a photo showing 4 caps that should be shorted. I would rather replace these, what spec caps should I use for them?

 

Thanks in advance...

post #2167 of 2185

Hi, I'm interested in removing the high-pass/coupling cap on the X-fi's  (Xtreme Music) line-in input.  Could anyone direct me to which cap this is?  150 pages is too much to wade through!

post #2168 of 2185

Ok, I've bypassed C19, C48, and C49, which gives the line-in a direct connection (after 20k series resistance) to the ADC.  However, I've found from the ADC spec sheet that there is a digital highpass filter that can only be disabled through software.  Is anyone aware of a way to disable this, say through modded drivers?

 

(I wish to use the soundcard as an A/D board, but my signals are too low frequency.)

post #2169 of 2185

I have had my modded card for a couple of years now, its amazing. My first card i over heated and fu*t it up. My second card i still use today. I had never soldered before but i managed.My advise is buy the really thin solder and you need to read all of this thread, here is the link   http://www.head-fi.org/t/226975/hotrodding-the-x-fi-a-laymans-guide-no-56k

post #2170 of 2185

I have been an electronics tech for several decades and I was horrified to read page 1 of this thread.

 

http://www.head-fi.org/t/226975/hotrodding-the-x-fi-a-laymans-guide-no-56k

 

Never, never, never short a capacitor. That is sheer madness. Here's why.

 

Capacitors for audio come in two basic types: electrolytic, which have larger capacitance and are polarized, and disk/paper/polyester/polystrene type, which are smaller and non-polarized.

 

On the SB0550 for example, the caps are mainly electrolytic and will have a + sign to mark where they connect to a voltage which is more positive than the -ve terminal. In the case of current flow. Direct current (electrons) flows into the capacitor plates via the -ve terminal and are stored on the plates till they are discharged back out the -ve terminal.

 

Normally, for filter caps and decoupling caps, that means the +ve goes to +5 volts, or +3.3 volts, and the -ve side goes to ground.

 

The caps provide two services, they act as filters to smooth power supplies and bypass high frequency noise spikes or they couple signals from one stage to another. I saw a reference to tantalum capacitors which are not intended to be used in place of electrolytics. Tantalums are normally used in parallel with a much larger capacitor. Being of smaller capacitance and physical size, they will react to and bypass spikes which are not affected by larger capacitors. Tantalums shunt noise spikes past the larger electrolytic filters.

 

If you short a capacitor that acts as a filter, you short the power supply and you blow something, usually the power supply or the regulator supplying that section. If you short a coupling capacitor that couples audio to the outside world, like an auxiliary out or a speaker, you expose the driving transistor, which may be in a chip, to damage.

 

Part of the purpose of coupling capacitors is keeping higher voltages from driving currents through transistors or chips that they cannot handle. They are even referred to as DC blockers in that case. Output coupling capacitors in modern direct-coupled transistor stages are polarized to carry currents from transistor collectors to aux outs, speakers or headphones. If you short the cap, you defeat the function of the cap by which it returns stored current to the circuit. That will degrade the signal and make it sound raspy, it it doesn't blow a chip.

 

It is claimed on page 1 that a higher quality chip (IC) replacement makes it OK to bypass the capacitor. That is wrong.

 

While I'm at it. let's lay this nonsense to rest about replacing capacitors with higher quality capacitors, especially larger ones. Higher quality capacitors will only help when it comes to the insulators (dielectrics) between the capacitor plates. A capacitor is nothing more than two metallic plates, usually foil, separated by an insulator called a dielectric. When a voltage is impressed across the cap, current runs into the plate connected to the negative supply but only until it is charged. It will sit there fully charged till the supply voltage is lowered or reversed.

 

The only advantage of a higher quality cap is cutting down on the likelihood of leakage through the dielectric from plate to plate. Higher quality caps will make no difference whatsoever to sound quality. In fact, when used to couple signals with capacitors, any capacitor will degrade the signal by altering the signal phase.

 

That's the problem with passive devices like capacitors (C) and inductors (L) and it's why modern audio circuits use direct coupling to eliminate inductors and capacitors as coupling devices. L and C devices change the relationship of signal current to signal voltage therefore they distort signals.

 

High quality audio power amps eliminate the output coupler but they use a fuse in series with the speakers or they have a current limiting circuit to prevent short circuit conditions if the output is shorted. With low signal circuits like auxiliary outs, there is no way to tell what the chip output is seeing so it's far safer to include a cap in series with the output to protect the chip. As a technician of longstanding, I find it insane to bypass output coupling caps. In fact, I find it insane to replace all those capacitors period.

 

As an example, if you have a cap coupling audio from a chip to headphones, if the headphones short, or if anything shorts while plugging the headphones into the jack, the chip might blow. It probably will.

 

Making caps larger serves no purpose. As filter components, capacitors are designed to work at 60 hertz for half wave rectifiers and 120 hertz for full wave rectifiers. The filtering action depends on the charge and discharge time of the capacitor which depends on it's size and the resistance through which it charges and discharges. They are designed to control ripple when an alternating current is rectified and the size must be fairly precise.

 

Changing the size could insert distortion by affecting the ripple frequency of the filtering. Power supply caps have an optimal size and making them larger does nothing. Some circuits use regulator chips after the capacitor filter anyway.

 

Making decoupling capacitors larger solves nothing either. Decoupler caps are placed all over the circuit board to decouple noise and spurious signals to ground. Making them larger messes with the kind of noise they can bypass. If you short them, you short the power supply, which is super dumb.

 

Making output couplers larger can blow a chip. The amount of current through a chip's transistor junction is limited, or the junction will burn. The larger you make a capacitor, the more it appears to a rising signal edge through a junction as a short circuit, and the transient conditions may be too large for the junction to handle.

 

Capacitors have the equivalent of a resistors resistance and it is called capacitive reactance. It is expressed in ohms. The reactance, Xc, is the inverse of the capacitance, so the larger you make a capacitor, the smaller the resistance is represents to an alternating signal. How could making the resistance lower possibly affect the fidelity of a signal? The fidelity depends on how well the signal is reproduced across a wide frequency band (20 -20,000 Hz for audio) and that depends on other matters like circuit design, distortion introduced by transistors, as well as other distortions like phase distortions from passive devices. The phase distortion from capacitors has nothing to do with capacitor quality, it is an in-built phenomenon of all inductors and capacitors related to the way they deal with currents running through them and into them.

 

I have already pointed out that a capacitor distorts the signal by altering the phase relationship of the signal current to the signal voltage. That distortion is frequency dependent. Changing the cap quality does nothing to improve that phase distortion and it certainly cannot affect any other parameter to improve fidelity.

post #2171 of 2185
Quote:
 

I know this is an old thread but for anyone coming across it, here's some stuff. Thanks to the poster of the previous post for the detailed info.

 

This thread is http://www.head-fi.org/t/226975/hotrodding-the-x-fi-a-laymans-guide-no-56k/1395, so find it and look at the previous thread.



>1) Given Line-OUT coupling capacitors have a 2.5vdc BIAS @ 1.55ma, shorting four will cause additional 6.2ma load on the VR1 regulator. Shorting all 16 Line-Out coupling caps would >add 24ma. Each of the four CS4398's nominally consume 25ma, totaling 100ma. The 78M05 is rated at MAX 500ma and nominally would be loaded at under 125ma to remain "cool to 

>the touch," given alotted PCB space. Also note measured temperatures of regulators.

 

Capacitors do not have a bias and if you are reading a current of 1.55 ma direct current through a capacitor, you have a leaky capacitor. It's more likely that you are reading an average DC current based on an alternating current signal and it would depend on the type of meter used.

 

Measuring DC voltage across a capacitors tells you nothing of value. Again, you are likely getting an average AC. Measuring AC across it is pointless since the AC varies.

 

Shorting out the output caps can lead to far more serious problems than an overloaded regulator. Since the coupler is normally in series with the upper leg of a push-pull output transistor  in a LINE OUT chip circuit, and the output, like headphones, shorting the cap allows a higher-than-design DC current to run through the chip's output transistor. That's a potentially serious situation, and that could cause a regulator to run hot. It's more likely to pop the transistor in the chip.

 

There is another purpose for the output cap. It charges in one direction through the headphones on one half of the output cycle and discharges through the headphones on the other half. The cap is an integral part of the audio output signal path. If you short it, you lose half the audio signal and most certainly you will overload the output transistor. By shorting the cap, you will introduce distortion for sure.

>I'm still not sure which regulator is getting "hot," so maybe use the reference picture in post above and indicate by "VR" number?

 

If a regulator is getting hot (normally meaning you can't hold a finger on it without getting burned) it is not being caused by current through an output coupling cap. A reg running too hot indicates too much current being drawn which means the load across it has increased. That means a lowered resistance than normal.

 

I would suspect shorting out coupling capacitors before anything else, unless while soldering a spatter of solder got caught across a trace.

post #2172 of 2185
Some of what you are saying is false The coupling capacitors are in fact not needed in these circuits in order to function properly. Data sheets for the DACs will bare this out. Yes there is 2.5 volt on these coupling capacitors but when you bypass them by shorting the caps the 2.5 volts positive gets sent to BOTH the Plus & Minus input of the OPAMP. Since the same positive voltage is applied to both inputs the positive D.C. voltage is cancelled. Only the undistorted audio signal remains. Distortion remains low & the output takes on a fresh openness that eluded the cap coupled output section.
 
Current in these output circuits to the  From the DAC to the opamp is much less than 1.5ma. Otherwise a much larger cap would be needed than 10uf to prevent loss of bass frequencies.
 
I've been doing these mods for years with no failure of opamps DAC's or voltage regulators. The sound improvement is always the same. Improved soundstage , more openness & better sounding bass with more harmonic detail that can tell you the type instrument being played including the brand & model if one is astute to the sound of those particular instruments. Much of this sound is lost in the coupling caps found on these P.C. Audio cards. Not all coupling caps sound bad but to me Electrolytic coupling caps sound almost universally bad. Metallyzed film cap are much better but won't fit in the confines of a computer sound card. Since good quality caps won't fit then direct coupling is in fact the best option from a sound point of view.
post #2173 of 2185
Quote:
Originally Posted by germanium View Post
 
>>>Some of what you are saying is false. The coupling capacitors are in fact not needed in these circuits in order to function properly. Data sheets for the DACs will bare this out.
 
Just want to clarify that my post was not intended as a flame. I appreciate the amount of work that went into the mod and some of the related information was of benefit to me.
 
I was not talking about DAC output. I was referring to this quote from the original post:
 
"Opamps and capacitors are the two worst offenders of sound quality. With any soundcard you can change these opamps to far better ones, remove capacitors no longer needed due to the better opamp, and give it more power by increasing the size of the power supply filter capacitors on the sound card".
 
This statement has several inaccuracies. Opamps by themselves are highly unlikely to contribute to poor fidelity. Most modern opamps have ideal circuitry and are far superior to circuits built from discrete components. Furthermore, you cannot simply change one opamp for another unless it is a fairly direct replacement.
 
The audio output opamps used on the X-Fi are already low noise, high performance opamps:
 
 
It's high unlikely that replacing them with the LM4562 will make that much of a difference. Sound quality is often relative to the ears of those listening.
 
Capacitors do degrade sound fidelity but the degradation is probably not noticeable to the average ear. Caps only degrade the phase shift of the signal between its current and voltage components. However, that kind of degradation is typical of any amplifier since their frequency response is frequency dependent.
 
In the statement above it claims you can replace caps 'no longer needed' due to the better opamp. That is simply incorrect. The cap is an integral part of the push-pull output stage used in opamps. Output caps are used even in high quality power amps.
 
A push-pull stage uses two transistors. In the bipolar junction type, they use either two NPNs or a complimentary paired NPN-PNP combo. The NPN pair is more common since it's easier to match two NPNs than it is an NPN to a PNP. When you have a pair operating, only one transistor is running at one time. The input signal is split between them so that the +ve going signal part drives one transistor and the -ve going side drives the other. That is called a Class AB1 amplifier.
 
With AB1, there is a transition zone as the +ve going output signal changes over to the -ve going signal. The output capacitor helps smooth that transition, otherwise the distorted signal transition sounds raspy through the speakers. It actually sounds like bees, and the distortion region as viewed on an oscilliscope look like 'bees knees'.
 
Besides that function, the capacitor is required for operation with a single ended power supply. During one half of the signal cycle, the cap stores charge from one transistor, then releases the charge through the other transistor on the other half cycle. Without the cap, the circuit wont work correctly.
 
The quote above also claims that increasing the size of the filter caps will increase power. Not true. Increasing the size of the caps may even reduce it. The caps are used to smooth out the rectified signal when a rectifier bridge converts the alternating supply voltage to a pulsating +ve voltage. The caps are designed to hold enough charge so that the charge will be released through the load a certain amount before the next rectified pulse. If the cap is too large, it will no longer function with the correct RC time constant.
 
In another part of the article:
 
"Replace the sound card's power supply capacitors with larger ones that are at least several times the original value. Quality matters, especially specs like ripple current".
 
Not true and not worth it.
 
And again:
 
"...some of them [opamps] (are??) extremely esoteric, requiring its own special power supply. Some of them plain unstable".
 
There is no such thing as a modern opamp that is unstable. If you use them in the wrong circuit they might be, however. There is a big difference between an opamp using bipolar junction transistors and one using MOS technology (CMOS). You just can't take any old opamps and solder it into the circuit where another has been removed, just as you can't replace any transistor with any other transistor. Opamps have specialized uses and it's little wonder the author found some unstable.
 
It takes a lot of experience in electronics and a lot of research to replace one component with another.
 
Then this:
 
"I don't have the schematics to the X-Fi so I don't know how to apply them. If others discover new mods I can add them to the list.

Short the 22uF caps near each opamp...."

 

The author admits not having a schematic then advises the reader to short a component. That is just not done in electronics. Before I ventured into an X-Fi circuit, I would make my own schematic and I would not replace or short components till I understood exactly what they did. Even at that, it is not always possible to understand the original design. They may have had trouble with stability during the design phase and added a cap for stability.

 

Shorting anything is never a good idea in electronics unless one knows exactly what he/she is doing.

 

From the article:

 

"...you bypass the opamps completely and use a discrete analog output stage..."

 

Makes no sense whatsoever. Opamps are superior to discrete-component amps because they can be built on the same silicon wafer. The problem with discrete components is matching the transistors and other components. In an opamp, transistors and components can be matched to near perfection

 

Bypassing an opamp is not as easy as it sounds. There are components built into the sound card that depend on having the opamp be there as a load or input. Removing the opamp without disabling the circuits could lead to instability. For example, I bypassed a circuit in a TV once by removing it's input and output connectors and disabling a regulator supplying voltage to the circuit. While I was doing something else, with the TV powered off, the sound circuit suddenly started operating. Nothing else was operating, just the sound circuit. It made no sense but such is life in the world of electronics.

 

Finally....

 

"Passive Output Mod - like in the discrete mod, you bypass the opamps completely, and use either a transformer or a bandpass filter

Alternatively, you can simply use a RC highpass filter..."

 

Why would anyone want to introduce a transformer into a high fidelity circuit? The inductance in the transformer distorts the output signals phase just as the coupling capacitor does.

 

There is an argument among so-called purists that tube amps are superior to transistor amps. That is nonsense, and I have repaired tube amps at one time. The problem with tube amps, besides the unreliable nature of tubes, due to the heat they produce, is the large output transformer required. It acts as a filter that distorts the tube output signal, making it sound soft. People are mistaking a distorted sound for better quality.

 

You can do the same, if you like, by taking the near perfect fidelity of a well designed transistor amp and running it through a filter.

 

Why, why, why would anyone attach an RC highpass filter to an audio amp to improve the sound? All you are doing is cutting the lows and passing the noise.

 

*********************

 
>>>>Yes there is 2.5 volt on these coupling capacitors but when you bypass them by shorting the caps the 2.5 volts positive gets sent to BOTH the Plus & Minus input of the OPAMP.
 
You are likely talking about the DAC output here. I can tell you one thing, almost for sure. If the coupling caps were not required they would not be there. Inter-stage coupling caps are used to keep direct currents from one circuit from flowing through the circuits of other components. Do you think engineers would use them if direct coupling was a superior option?
 

>>>I've been doing these mods for years with no failure of opamps DAC's or voltage regulators.

 

As I said, I have no intention of flaming and I applaud your effort. However, I have several decades experience in electronics, having spent several years on the bench repairing every kind of amp from tube amps to transistor amps. I am also highly trained with several years at university studying electrical engineering. All I am trying to do is pass on information that might prevent an inexperienced user from blowing his/her sound card.

 

I would never attempt your mods blindly without first making a schematic and spending hours researching the effects of the mods.

 

I do have an interest in modding the connections between various Creative products to make the drives from one work with another.

 

For example, some card have the Ad_Link connector on the front of the card. I have an SB0550 used with the Elite Pro external model and I'd like to see if the SB0460, with the same Ad_link connector will work with the Elite Pro. I wont try any mods till I am absolutely sure they wont blow anything or compromise the integrity of the original design.

post #2174 of 2185
Quote:
Originally Posted by germanium View Post
 
>>>Some of what you are saying is false. The coupling capacitors are in fact not needed in these circuits in order to function properly. Data sheets for the DACs will bare this out.
 
Just want to clarify that my post was not intended as a flame. I appreciate the amount of work that went into the mod and some of the related information was of benefit to me.
 
I was not talking about DAC output. I was referring to this quote from the original post:
 
"Opamps and capacitors are the two worst offenders of sound quality. With any soundcard you can change these opamps to far better ones, remove capacitors no longer needed due to the better opamp, and give it more power by increasing the size of the power supply filter capacitors on the sound card".
 
This statement has several inaccuracies. Opamps by themselves are highly unlikely to contribute to poor fidelity. Most modern opamps have ideal circuitry and are far superior to circuits built from discrete components. Furthermore, you cannot simply change one opamp for another unless it is a fairly direct replacement.
 
The audio output opamps used on the X-Fi are already low noise, high performance opamps.It's high unlikely that replacing them with the LM4562 will make that much of a difference. Sound quality is often relative to the ears of those listening.
 
Capacitors do degrade sound fidelity but the degradation is probably not noticeable to the average ear. Caps only degrade the phase shift of the signal between its current and voltage components. However, that kind of degradation is typical of any amplifier since their frequency response is frequency dependent.
 
In the statement above it claims you can replace caps 'no longer needed' due to the better opamp. That is simply incorrect. The cap is an integral part of the push-pull output stage used in opamps. Output caps are used even in high quality power amps.
 
A push-pull stage uses two transistors. In the bipolar junction type, they use either two NPNs or a complimentary paired NPN-PNP combo. The NPN pair is more common since it's easier to match two NPNs than it is an NPN to a PNP. When you have a pair operating, only one transistor is running at one time. The input signal is split between them so that the +ve going signal part drives one transistor and the -ve going side drives the other. That is called a Class AB1 amplifier.
 
With AB1, there is a transition zone as the +ve going output signal changes over to the -ve going signal. The output capacitor helps smooth that transition, otherwise the distorted signal transition sounds raspy through the speakers. It actually sounds like bees, and the distortion region as viewed on an oscilliscope look like 'bees knees'.
 
Besides that function, the capacitor is required for operation with a single ended power supply. During one half of the signal cycle, the cap stores charge from one transistor, then releases the charge through the other transistor on the other half cycle. Without the cap, the circuit wont work correctly.
 
The quote above also claims that increasing the size of the filter caps will increase power. Not true. Increasing the size of the caps may even reduce it. The caps are used to smooth out the rectified signal when a rectifier bridge converts the alternating supply voltage to a pulsating +ve voltage. The caps are designed to hold enough charge so that the charge will be released through the load a certain amount before the next rectified pulse. If the cap is too large, it will no longer function with the correct RC time constant.
 
In another part of the article:
 
"Replace the sound card's power supply capacitors with larger ones that are at least several times the original value. Quality matters, especially specs like ripple current".
 
Not true and not worth it.
 
And again:
 
"...some of them [opamps] (are??) extremely esoteric, requiring its own special power supply. Some of them plain unstable".
 
There is no such thing as a modern opamp that is unstable. If you use them in the wrong circuit they might be, however. There is a big difference between an opamp using bipolar junction transistors and one using MOS technology (CMOS). You just can't take any old opamps and solder it into the circuit where another has been removed, just as you can't replace any transistor with any other transistor. Opamps have specialized uses and it's little wonder the author found some unstable.
 
It takes a lot of experience in electronics and a lot of research to replace one component with another.
 
Then this:
 
"I don't have the schematics to the X-Fi so I don't know how to apply them. If others discover new mods I can add them to the list.

Short the 22uF caps near each opamp...."

 

The author admits not having a schematic then advises the reader to short a component. That is just not done in electronics. Before I ventured into an X-Fi circuit, I would make my own schematic and I would not replace or short components till I understood exactly what they did. Even at that, it is not always possible to understand the original design. They may have had trouble with stability during the design phase and added a cap for stability.

 

Shorting anything is never a good idea in electronics unless one knows exactly what he/she is doing.

 

From the article:

 

"...you bypass the opamps completely and use a discrete analog output stage..."

 

Makes no sense whatsoever. Opamps are superior to discrete-component amps because they can be built on the same silicon wafer. The problem with discrete components is matching the transistors and other components. In an opamp, transistors and components can be matched to near perfection

 

Bypassing an opamp is not as easy as it sounds. There are components built into the sound card that depend on having the opamp be there as a load or input. Removing the opamp without disabling the circuits could lead to instability. For example, I bypassed a circuit in a TV once by removing it's input and output connectors and disabling a regulator supplying voltage to the circuit. While I was doing something else, with the TV powered off, the sound circuit suddenly started operating. Nothing else was operating, just the sound circuit. It made no sense but such is life in the world of electronics.

 

Finally....

 

"Passive Output Mod - like in the discrete mod, you bypass the opamps completely, and use either a transformer or a bandpass filter

Alternatively, you can simply use a RC highpass filter..."

 

Why would anyone want to introduce a transformer into a high fidelity circuit? The inductance in the transformer distorts the output signals phase just as the coupling capacitor does.

 

There is an argument among so-called purists that tube amps are superior to transistor amps. That is nonsense, and I have repaired tube amps at one time. The problem with tube amps, besides the unreliable nature of tubes, due to the heat they produce, is the large output transformer required. It acts as a filter that distorts the tube output signal, making it sound soft. People are mistaking a distorted sound for better quality.

 

You can do the same, if you like, by taking the near perfect fidelity of a well designed transistor amp and running it through a filter.

 

Why, why, why would anyone attach an RC highpass filter to an audio amp to improve the sound? All you are doing is cutting the lows and passing the noise.

 

*********************

 
>>>>Yes there is 2.5 volt on these coupling capacitors but when you bypass them by shorting the caps the 2.5 volts positive gets sent to BOTH the Plus & Minus input of the OPAMP.
 
You are likely talking about the DAC output here. I can tell you one thing, almost for sure. If the coupling caps were not required they would not be there. Inter-stage coupling caps are used to keep direct currents from one circuit from flowing through the circuits of other components. Do you think engineers would use them if direct coupling was a superior option?
 

>>>I've been doing these mods for years with no failure of opamps DAC's or voltage regulators.

 

As I said, I have no intention of flaming and I applaud your effort. However, I have several decades experience in electronics, having spent several years on the bench repairing every kind of amp from tube amps to transistor amps. I am also highly trained with several years at university studying electrical engineering. All I am trying to do is pass on information that might prevent an inexperienced user from blowing his/her sound card.

 

I would never attempt your mods blindly without first making a schematic and spending hours researching the effects of the mods.

 

I do have an interest in modding the connections between various Creative products to make the drives from one work with another.

 

For example, some card have the Ad_Link connector on the front of the card. I have an SB0550 used with the Elite Pro external model and I'd like to see if the SB0460, with the same Ad_link connector will work with the Elite Pro. I wont try any mods till I am absolutely sure they wont blow anything or compromise the integrity of the original design.

post #2175 of 2185
Quote:
Originally Posted by germanium View Post
 
>>>Some of what you are saying is false. The coupling capacitors are in fact not needed in these circuits in order to function properly. Data sheets for the DACs will bare this out.
 
Just want to clarify that my post was not intended as a flame. I appreciate the amount of work that went into the mod and some of the related information was of benefit to me.
 
I was not talking about DAC output. I was referring to this quote from the original post:
 
"Opamps and capacitors are the two worst offenders of sound quality. With any soundcard you can change these opamps to far better ones, remove capacitors no longer needed due to the better opamp, and give it more power by increasing the size of the power supply filter capacitors on the sound card".
 
This statement has several inaccuracies. Opamps by themselves are highly unlikely to contribute to poor fidelity. Most modern opamps have ideal circuitry and are far superior to circuits built from discrete components. Furthermore, you cannot simply change one opamp for another unless it is a fairly direct replacement.
 
The audio output opamps used on the X-Fi are already low noise, high performance opamps. It's high unlikely that replacing them with the LM4562 will make that much of a difference. Sound quality is often relative to the ears of those listening.
 
Capacitors do degrade sound fidelity but the degradation is probably not noticeable to the average ear. Caps only degrade the phase shift of the signal between its current and voltage components. However, that kind of degradation is typical of any amplifier since their frequency response is frequency dependent.
 
In the statement above it claims you can replace caps 'no longer needed' due to the better opamp. That is simply incorrect. The cap is an integral part of the push-pull output stage used in opamps. Output caps are used even in high quality power amps.
 
A push-pull stage uses two transistors. In the bipolar junction type, they use either two NPNs or a complimentary paired NPN-PNP combo. The NPN pair is more common since it's easier to match two NPNs than it is an NPN to a PNP. When you have a pair operating, only one transistor is running at one time. The input signal is split between them so that the +ve going signal part drives one transistor and the -ve going side drives the other. That is called a Class AB1 amplifier.
 
With AB1, there is a transition zone as the +ve going output signal changes over to the -ve going signal. The output capacitor helps smooth that transition, otherwise the distorted signal transition sounds raspy through the speakers. It actually sounds like bees, and the distortion region as viewed on an oscilliscope look like 'bees knees'.
 
Besides that function, the capacitor is required for operation with a single ended power supply. During one half of the signal cycle, the cap stores charge from one transistor, then releases the charge through the other transistor on the other half cycle. Without the cap, the circuit wont work correctly.
 
The quote above also claims that increasing the size of the filter caps will increase power. Not true. Increasing the size of the caps may even reduce it. The caps are used to smooth out the rectified signal when a rectifier bridge converts the alternating supply voltage to a pulsating +ve voltage. The caps are designed to hold enough charge so that the charge will be released through the load a certain amount before the next rectified pulse. If the cap is too large, it will no longer function with the correct RC time constant.
 
In another part of the article:
 
"Replace the sound card's power supply capacitors with larger ones that are at least several times the original value. Quality matters, especially specs like ripple current".
 
Not true and not worth it.
 
And again:
 
"...some of them [opamps] (are??) extremely esoteric, requiring its own special power supply. Some of them plain unstable".
 
There is no such thing as a modern opamp that is unstable. If you use them in the wrong circuit they might be, however. There is a big difference between an opamp using bipolar junction transistors and one using MOS technology (CMOS). You just can't take any old opamps and solder it into the circuit where another has been removed, just as you can't replace any transistor with any other transistor. Opamps have specialized uses and it's little wonder the author found some unstable.
 
It takes a lot of experience in electronics and a lot of research to replace one component with another.
 
Then this:
 
"I don't have the schematics to the X-Fi so I don't know how to apply them. If others discover new mods I can add them to the list.

Short the 22uF caps near each opamp...."

 

The author admits not having a schematic then advises the reader to short a component. That is just not done in electronics. Before I ventured into an X-Fi circuit, I would make my own schematic and I would not replace or short components till I understood exactly what they did. Even at that, it is not always possible to understand the original design. They may have had trouble with stability during the design phase and added a cap for stability.

 

Shorting anything is never a good idea in electronics unless one knows exactly what he/she is doing.

 

From the article:

 

"...you bypass the opamps completely and use a discrete analog output stage..."

 

Makes no sense whatsoever. Opamps are superior to discrete-component amps because they can be built on the same silicon wafer. The problem with discrete components is matching the transistors and other components. In an opamp, transistors and components can be matched to near perfection

 

Bypassing an opamp is not as easy as it sounds. There are components built into the sound card that depend on having the opamp be there as a load or input. Removing the opamp without disabling the circuits could lead to instability. For example, I bypassed a circuit in a TV once by removing it's input and output connectors and disabling a regulator supplying voltage to the circuit. While I was doing something else, with the TV powered off, the sound circuit suddenly started operating. Nothing else was operating, just the sound circuit. It made no sense but such is life in the world of electronics.

 

Finally....

 

"Passive Output Mod - like in the discrete mod, you bypass the opamps completely, and use either a transformer or a bandpass filter

Alternatively, you can simply use a RC highpass filter..."

 

Why would anyone want to introduce a transformer into a high fidelity circuit? The inductance in the transformer distorts the output signals phase just as the coupling capacitor does.

 

There is an argument among so-called purists that tube amps are superior to transistor amps. That is nonsense, and I have repaired tube amps at one time. The problem with tube amps, besides the unreliable nature of tubes, due to the heat they produce, is the large output transformer required. It acts as a filter that distorts the tube output signal, making it sound soft. People are mistaking a distorted sound for better quality.

 

You can do the same, if you like, by taking the near perfect fidelity of a well designed transistor amp and running it through a filter.

 

Why, why, why would anyone attach an RC highpass filter to an audio amp to improve the sound? All you are doing is cutting the lows and passing the noise.

 

*********************

 
>>>>Yes there is 2.5 volt on these coupling capacitors but when you bypass them by shorting the caps the 2.5 volts positive gets sent to BOTH the Plus & Minus input of the opamp.."
 
You are likely talking about the DAC output here. I can tell you one thing, almost for sure. If the coupling caps were not required they would not be there. Inter-stage coupling caps are used to keep direct currents from one circuit from flowing through the circuits of other components. Do you think engineers would use them if direct coupling was a superior option?
 

>>>I've been doing these mods for years with no failure of opamps DAC's or voltage regulators.

 

As I said, I have no intention of flaming and I applaud your effort. However, I have several decades experience in electronics, having spent several years on the bench repairing every kind of amp from tube amps to transistor amps. I am also highly trained with several years at university studying electrical engineering. All I am trying to do is pass on information that might prevent an inexperienced user from blowing his/her sound card.

 

I would never attempt your mods blindly without first making a schematic and spending hours researching the effects of the mods.

 

I do have an interest in modding the connections between various Creative products to make the drives from one work with another.

 

For example, some card have the Ad_Link connector on the front of the card. I have an SB0550 used with the Elite Pro external model and I'd like to see if the SB0460, with the same Ad_link connector will work with the Elite Pro. I wont try any mods till I am absolutely sure they wont blow anything or compromise the integrity of the original design.

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)