Page 1 of 1

Update 04.13.09

Posted: Mon Apr 13, 2009 10:44 am
by Chuck
First, if you haven't done your taxes now would be a good time ;-)

Second, have any of you ever run into this problem; computer noises (for lack of a better term) coming through the speakers. We've found that if you use a cheap computer power supply that the soundblaster port emits noises that are in sync with computer events (moving a mouse, hard disk access, etc.) I found this issue with our amp and then found it happens with any amp (even my expensive alpine.) We are going to add additional filtering components to our audio amp to try and eliminate this noise. I just wondered if anyone had experience with it using mame boxes or home stereos hooked up to a computer. The easy fix is buying a better power supply.

Re: Update 04.13.09

Posted: Mon Apr 13, 2009 3:40 pm
by Jazman
Yea, if you're talking about the on-board sound on the motherboard, you can run into this. As it is, the motherboard based chips are noisy simply because the analog amp shares the same power supply as all the high-speed switching stuff. Even if you have an external amp, there is still an amplifier to drive the external connection. So, the noisier the power supply feeding the motherboard is, the noisier the whole system is.

Have you tried playing around with the grounding of the system? If you've got ground loops in the system, you will have a higher susceptibility to picking up the noise and amplifying it further. Filtering it out is certainly one fix, but the preferable way is to prevent it from getting there. Remember that grounding something every place you can isn't necessarily the best solution as this can lead to ground loops.

Jaz

Re: Update 04.13.09

Posted: Mon Apr 13, 2009 3:50 pm
by Chuck
Yeah, I've been in bands for the last 20 years and have a lot of experience with ground loops ;-) I did try to make sure everything was grounded and then tried to isolate a few things. I think there is just no good way to get around a junk power supply.

On a positive note my Intel Core2 based test machine is running really well. The case I ordered with it was garbage (no way to know until you try it.) I swapped the new components into my old nucore test machine and will take that to PATZ this week and test it under extreme load. Jaz, you coming to PATZ this weekend?

Re: Update 04.13.09

Posted: Mon Apr 13, 2009 4:04 pm
by Jazman
Yea, there is an inherent ground loop in your scheme that is hard to eliminate. You have the ATX power cable (5V, 12V, GND) with the ground running back to the power supply and connecting to multiple places (ie: hard drives, motherboard, etc) which may have chassis connections. Then, you have the common (also usually ground referenced) coming from the audio output for your signal. So, the 2 ground connections are both necessary and have unknown connections. The only thing you can try to do really is figure out a better way to route them. Remember that the susceptibility of the ground loop is directly related to the area inside the loop. So, for example, if you could somehow get power from the motherboard or motherboard connection and run it in parallel with the audio output, then the loop area would be minimized as would the issue.

Of course, if a better power supply fixes it in all cases, it will probably be the easiest way to address it.

Re Zoo: Man, I wish I could. It's my nieces birthday this weekend and the missus has told me I will be there. :)

Jaz