I have an active PFC type power supply (Enermax 465AX-VE) which is causing electrical noise on both of my sound cards at the frequency of the mains (50 Hz) and multiples of that.
I have already tried the normal "tricks" suggested by many sites:
- move the cards furthest away from the PSU
- use an UPS (damn the good ones are acoustically noisy!)
- use grounded mains outlet (I'm running that as we speak)
- only have single ground contact (yes, always)
- shield the card with tinfoil wrapping (tried it, didn't affect
my readings)
- try another sound card (have tried RME DIGI 96/8 PAD and
Creative Audigy 2 Platinum eX, both of which some other
people have been able to make more immune to mains noise)
Is there something else I can do?
I'm sorry that I'm not an electricity guru and am at my wits end trying to solve this on my own.
I can't afford to have a separate regulated mains line installed at my home "studio" and to be honest, I don't even know if that would help.
I'd really like to hear other people's ideas on this, especially if you've tackled PC based electrical noise sources before.
best regards,
Halcyon
I have already tried the normal "tricks" suggested by many sites:
- move the cards furthest away from the PSU
- use an UPS (damn the good ones are acoustically noisy!)
- use grounded mains outlet (I'm running that as we speak)
- only have single ground contact (yes, always)
- shield the card with tinfoil wrapping (tried it, didn't affect
my readings)
- try another sound card (have tried RME DIGI 96/8 PAD and
Creative Audigy 2 Platinum eX, both of which some other
people have been able to make more immune to mains noise)
Is there something else I can do?
I'm sorry that I'm not an electricity guru and am at my wits end trying to solve this on my own.
I can't afford to have a separate regulated mains line installed at my home "studio" and to be honest, I don't even know if that would help.
I'd really like to hear other people's ideas on this, especially if you've tackled PC based electrical noise sources before.
best regards,
Halcyon
Other than opening up the powersupply and adding some filtering of your own, i really dunno...
I once had a cd unit that was very noisy when playing audio. The solution? I built a small molex extension cable with decoupling for both the +5v and +12v rails
Worked wonders! Something like that could do it, but soldering on your beloved soundcard would be insane...
I once had a cd unit that was very noisy when playing audio. The solution? I built a small molex extension cable with decoupling for both the +5v and +12v rails
Worked wonders! Something like that could do it, but soldering on your beloved soundcard would be insane...
Hi!
What is your soundcard connected to? Active Speakers or amplifier?
Try using a headphone, is the noise still there? If not, you probably have a ground loop in your chain.
Always try to use the same power line for all the components on the chain (computer, amplifier, speakers...)
If you use an amplifier, try disconnecting all other sources from the amp (when I used a normal - not DIY - integrated amp, I had two computers hooked to the amp, and as long as both were connected - even when not turned on - the running computer produced noise because of ground loops)...
I must not necessarily be the power supply....
Bye,
Arndt
What is your soundcard connected to? Active Speakers or amplifier?
Try using a headphone, is the noise still there? If not, you probably have a ground loop in your chain.
Always try to use the same power line for all the components on the chain (computer, amplifier, speakers...)
If you use an amplifier, try disconnecting all other sources from the amp (when I used a normal - not DIY - integrated amp, I had two computers hooked to the amp, and as long as both were connected - even when not turned on - the running computer produced noise because of ground loops)...
I must not necessarily be the power supply....
Bye,
Arndt
I need to fill in some more details, I was too sparse in my original post
The noise is not easily audible, it's still -80 dB below signal, but it is very annoying for trying to do archive quality recording for later processing, when you know that your hardware is capable of better (that's why it was bought) and that other people get 20 dB better rejection with the same sound card
I guess I will try another power supply.
Are there any special features or abilites I should be on the look for?
Like total metal casing (no fans)?
non-switched mode psu (are they even possible for PCs)?
something else?
Should I try decoupling the sound cards from the case, so that they have contact only to the motherboard and not to the casing back panel?
How about mu-metal shielding (expensive)?
regards,
Halcyon
PS It's not a ground loop. Single ground contact. However, our house wiring is old and the grounding is done via the neutral wire, so there might be potential fluctuations in the neutral wire.
The noise is not easily audible, it's still -80 dB below signal, but it is very annoying for trying to do archive quality recording for later processing, when you know that your hardware is capable of better (that's why it was bought) and that other people get 20 dB better rejection with the same sound card
I guess I will try another power supply.
Are there any special features or abilites I should be on the look for?
Like total metal casing (no fans)?
non-switched mode psu (are they even possible for PCs)?
something else?
Should I try decoupling the sound cards from the case, so that they have contact only to the motherboard and not to the casing back panel?
How about mu-metal shielding (expensive)?
regards,
Halcyon
PS It's not a ground loop. Single ground contact. However, our house wiring is old and the grounding is done via the neutral wire, so there might be potential fluctuations in the neutral wire.
halcyon said:I need to fill in some more details, I was too sparse in my original post
The noise is not easily audible, it's still -80 dB below signal, but it is very annoying for trying to do archive quality recording for later processing, when you know that your hardware is capable of better (that's why it was bought) and that other people get 20 dB better rejection with the same sound card
---------------------------
You should do much better. Why not do a spectral analysis of the output of your PS and compare this to your soundcard output.
Hmm…
It is not necessarily your PSU. If it is digital noise, your mainboard also plays a major role. On many mainboards the low CPU voltages are taken from the +12V by switching regulators. I am using an MSI mainboard now and it generates a lot (20dB) less noise than my old Abit board with the same PSU (el cheapo Powerman).
Also take into account that mains cables generate a lot of EMI. Keep your audio cables far away from it. Preferably route them straight away along the sides of you PC. Use good screened cables.
To get some idea what the performance of your card is really is, download a program like “Audiotester”. Make a loop through cable of 10cm between input and output and see what you see. Also take some FFT’s with longer cables. You will be amazed what the influence on routing is concerning induced mains related noise.
Cheers
It is not necessarily your PSU. If it is digital noise, your mainboard also plays a major role. On many mainboards the low CPU voltages are taken from the +12V by switching regulators. I am using an MSI mainboard now and it generates a lot (20dB) less noise than my old Abit board with the same PSU (el cheapo Powerman).
Also take into account that mains cables generate a lot of EMI. Keep your audio cables far away from it. Preferably route them straight away along the sides of you PC. Use good screened cables.
To get some idea what the performance of your card is really is, download a program like “Audiotester”. Make a loop through cable of 10cm between input and output and see what you see. Also take some FFT’s with longer cables. You will be amazed what the influence on routing is concerning induced mains related noise.
Cheers
Maybe the power supply isn't the souece of the noise. Maybe it's a grounding issue, or something? Have you tried other outlets in your house? Is the noise present with no input connected to the card?
The reason I suggest this is that it's unlikely that you would get 50Hz noise from a switch mode power supply, I think. The way they work is to rectify the mains, create a high frequency (KHz) wave, then rectify it again. It seems unlikely that the 50Hz mains frequency would make it through this process.
The reason I suggest this is that it's unlikely that you would get 50Hz noise from a switch mode power supply, I think. The way they work is to rectify the mains, create a high frequency (KHz) wave, then rectify it again. It seems unlikely that the 50Hz mains frequency would make it through this process.
Even more data
Ok, here's an RMAA 5.0 test of my sound cards non-balanced outputs to non-balanced inputs. The cable was shielded and 5 centimeters long. No mains cable nearby (I also have a very good shielded Supra mains cable).
There is nothing connected to my sound card during the testing, no amp, no speakers, no headhpones. Only the soundcard is connected to itself via the external outputs/inputs.
Of course I have KB, mouse and monitor connected. Monitor is in the same ground contact as my computer. I have tried only two outlets that I have in my room, but the results are the same.
http://www2.uiah.fi/~samu/rmaa/RME_digi_96-8_pad.html
As you can see, there's quite a peak at 50Hz and then less so at multiples of that (mains frequency).
My other sound card is slightly more immune to this noise and produces slightly better results.
Unfortunately I don't have equipment to measure my power supply noise characteristics directly (no scope).
Maybe it's not PSU noise, I'm not sure anymore. I still welcome additional ideas on how to proceed & debug.
And please forgive me if I*m a little slow on the catch up. I may not have the usual level of knowledge people in this forum seem to have
regards,
Halcyon
Ok, here's an RMAA 5.0 test of my sound cards non-balanced outputs to non-balanced inputs. The cable was shielded and 5 centimeters long. No mains cable nearby (I also have a very good shielded Supra mains cable).
There is nothing connected to my sound card during the testing, no amp, no speakers, no headhpones. Only the soundcard is connected to itself via the external outputs/inputs.
Of course I have KB, mouse and monitor connected. Monitor is in the same ground contact as my computer. I have tried only two outlets that I have in my room, but the results are the same.
http://www2.uiah.fi/~samu/rmaa/RME_digi_96-8_pad.html
As you can see, there's quite a peak at 50Hz and then less so at multiples of that (mains frequency).
My other sound card is slightly more immune to this noise and produces slightly better results.
Unfortunately I don't have equipment to measure my power supply noise characteristics directly (no scope).
Maybe it's not PSU noise, I'm not sure anymore. I still welcome additional ideas on how to proceed & debug.
And please forgive me if I*m a little slow on the catch up. I may not have the usual level of knowledge people in this forum seem to have
regards,
Halcyon
Hi!
While generally disliking this device (not supported by kxDriver...), you could try using the Soundblaster Extigy device, since it can be put anywhere, has it's own power supply, no you could exclude any influences from computer internals... Only the digitalized signals get passed through USB to the computer...
If you can try it for free , do so...
Bye,
Arndt
While generally disliking this device (not supported by kxDriver...), you could try using the Soundblaster Extigy device, since it can be put anywhere, has it's own power supply, no you could exclude any influences from computer internals... Only the digitalized signals get passed through USB to the computer...
If you can try it for free
, do so...Bye,
Arndt
Thanks for the tip.
I don't have Extigy and it doesn't do what I want, so it wouldn't solve my issue.
However, I know by looking at other people's measurement graphs that they have been able to get much lower noise figures with the exaxt same model sound cards that I have.
So, it's something in my system or mains electricity (or something else pertaining to my setup) that is causing the problem.
I'm hoping I could find and alleviate/eliminate it rather than keep on switching to inferior sound cards
cheers,
Halcyon
I don't have Extigy and it doesn't do what I want, so it wouldn't solve my issue.
However, I know by looking at other people's measurement graphs that they have been able to get much lower noise figures with the exaxt same model sound cards that I have.
So, it's something in my system or mains electricity (or something else pertaining to my setup) that is causing the problem.
I'm hoping I could find and alleviate/eliminate it rather than keep on switching to inferior sound cards
cheers,
Halcyon
waaaaaaaaaaaaaait!!!
Wait!
I had a similar problem (I suppose), since I changed home.
On my new house I had a humm problem with the sound card that I never had.
Instead of fiddling, you need to think simple first.
If everything is the same (except the house
) why does it hum??!
I noticed it while recording from vinyl, there was a noticeable noise that I never had (with loud music playing it's hard to notice).
Do you know what I did?
I DISCONNECTED THE EARTH FROM MY PC!
Crappy earth, noise for sure.
It's better not to have.
Wait!
I had a similar problem (I suppose), since I changed home.
On my new house I had a humm problem with the sound card that I never had.
Instead of fiddling, you need to think simple first.
If everything is the same (except the house
) why does it hum??!I noticed it while recording from vinyl, there was a noticeable noise that I never had (with loud music playing it's hard to notice).
Do you know what I did?
I DISCONNECTED THE EARTH FROM MY PC!
Crappy earth, noise for sure.
It's better not to have.
carlosfm,
you're probably onto something here.
I checked... our building has very old wiring.
Usually it's done with three phase wires, one ground wire and one neutral wire.
In our aparment we only have two phases and no ground wire, so grounding has been done using the neutral wire (perfectly legal to do and recommended for old buildings, done by a professional/certified company it seems).
So, it could be that the problem with the ground (or neutral) is the cause here?
I really need an electricity quality meter now
regards,
Halcyon
you're probably onto something here.
I checked... our building has very old wiring.
Usually it's done with three phase wires, one ground wire and one neutral wire.
In our aparment we only have two phases and no ground wire, so grounding has been done using the neutral wire (perfectly legal to do and recommended for old buildings, done by a professional/certified company it seems).
So, it could be that the problem with the ground (or neutral) is the cause here?
I really need an electricity quality meter now
regards,
Halcyon
Try this
www.soundtechnology.com
Download the trial of Spectralab, and do some FFT fo the sound card o/p for analysis.
A common problem is to leave unused inputs enabled when doing high-quality work. The mic input is often the worst - are you sure the noise is not mixed from other inputs?
Try muting all inputs except the one you're using.
Andy.
www.soundtechnology.com
Download the trial of Spectralab, and do some FFT fo the sound card o/p for analysis.
A common problem is to leave unused inputs enabled when doing high-quality work. The mic input is often the worst - are you sure the noise is not mixed from other inputs?
Try muting all inputs except the one you're using.
Andy.
halcyon,
It's easy to disconnect the earth wire from the plug and try.
I think it's the first thing to do.
You may be surpised, because I can bet the problem is the earth.
It can be good for fridges and washing machines, but not good for audio.
A defective earth is really not good for audio.
It's easy to disconnect the earth wire from the plug and try.
I think it's the first thing to do.
You may be surpised, because I can bet the problem is the earth.
It can be good for fridges and washing machines, but not good for audio.
A defective earth is really not good for audio.
Ground loops ?
Ehm Andy, Halcyon did post already some pics of FFT’s. The pics were taken in loop through with nothing connected to the soundcard. That means that there are no ground loops in that case. Still the hum is –80dB FS. That means the hum can only come from inside the PC. I think Halcyon is quite close with his first post: The hum comes from the PSU itself. IMHO it is worth trying another PSU and see if this make any difference. No mains cable running internal along the mainboard, to a frontpanel mains switch for instance?
With my internal M-Audio PCI card the mains hum is well below –120 dB. Here I already did post some pics of the M-Audio card, scroll to the bottom of the page. RME is well known for its excellent audio cards.
Ehm Andy, Halcyon did post already some pics of FFT’s. The pics were taken in loop through with nothing connected to the soundcard. That means that there are no ground loops in that case. Still the hum is –80dB FS. That means the hum can only come from inside the PC. I think Halcyon is quite close with his first post: The hum comes from the PSU itself. IMHO it is worth trying another PSU and see if this make any difference. No mains cable running internal along the mainboard, to a frontpanel mains switch for instance?
With my internal M-Audio PCI card the mains hum is well below –120 dB. Here I already did post some pics of the M-Audio card, scroll to the bottom of the page. RME is well known for its excellent audio cards.
Ok, update time.
I removed my PC from the mains ground contact for testing.
Then I ran the RMAA 5.0 tests again and the results are almost exactly the same.
So, it seems that the crap grounding wasn't the main cause of the 50 Hz noise.
I have found a decent power supply (in terms of acoustic noise) that I'm going to try out as soon as it's available at local resellers.
I'll report back after that.
regards,
Halcyon
I removed my PC from the mains ground contact for testing.
Then I ran the RMAA 5.0 tests again and the results are almost exactly the same.
So, it seems that the crap grounding wasn't the main cause of the 50 Hz noise.
I have found a decent power supply (in terms of acoustic noise) that I'm going to try out as soon as it's available at local resellers.
I'll report back after that.
regards,
Halcyon
Most of the time the ripple noise in sondcards is due to the mainboard, and sometimes due to the PSU. Remember the CPU is hogging a lot of power, and a mainboard with poor CPU power regulation - essentially from the 5 and 3.3V rails - will be throwing a lot of noise back onto the rails. Please try out one of each the PS and mobo.
Sometimes even the drivers will make a difference. I have an SB Live! installed which has a lot of noise in 98 using the Creative default drivers, and under 2000 with the Kx driver is quiet as a mouse - or as quiet as a Live! can be, which rally isn't all that great, but does my job well enough.
Sometimes even the drivers will make a difference. I have an SB Live! installed which has a lot of noise in 98 using the Creative default drivers, and under 2000 with the Kx driver is quiet as a mouse - or as quiet as a Live! can be, which rally isn't all that great, but does my job well enough.
More update time:
I don't think it's the drivers in this case, because other people get much better readings with the same card using same drivers. However, I do agree that especially Creative drivers are well known for their ability of mungling the sound (especially due to resampling artifacts).
I got a Fluke Electricity Quality analyzer and a friend who works at an electricity testing lab to hop over.
We quickly tested the ground potential (very close to nominal zero, no constant voltage there), the voltage (very nice sine waveform, below 2% THD, almost constant 235V as it should be) and some other measures.
Of course, some loads on the mains cause the current to break down into ugly waveform (for instance a television does this), but I have the problems even if I don't have these aforementioned devices connected to the same wire.
Haven't yet measured the current waveform for the computer, so I will do that after I've run a bit of logging at my main hifi rig first.
Preliminary testing however suggest that it is not the ground nor the mains electricity that is causing the problem. I'll be hopefully wiser after I've run more tests (and for a longer duration).
I've already ordered a new PSU and will do measurements of the sound card when I get it / have it installed.
More updates to follow, in case anyone's interested
regards,
Halcyon
I don't think it's the drivers in this case, because other people get much better readings with the same card using same drivers. However, I do agree that especially Creative drivers are well known for their ability of mungling the sound (especially due to resampling artifacts).
I got a Fluke Electricity Quality analyzer and a friend who works at an electricity testing lab to hop over.
We quickly tested the ground potential (very close to nominal zero, no constant voltage there), the voltage (very nice sine waveform, below 2% THD, almost constant 235V as it should be) and some other measures.
Of course, some loads on the mains cause the current to break down into ugly waveform (for instance a television does this), but I have the problems even if I don't have these aforementioned devices connected to the same wire.
Haven't yet measured the current waveform for the computer, so I will do that after I've run a bit of logging at my main hifi rig first.
Preliminary testing however suggest that it is not the ground nor the mains electricity that is causing the problem. I'll be hopefully wiser after I've run more tests (and for a longer duration).
I've already ordered a new PSU and will do measurements of the sound card when I get it / have it installed.
More updates to follow, in case anyone's interested
regards,
Halcyon
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.