Capacitor on "ground" interconnection to block "ground" DC voltage
Sounds strange, I know, and I'm considering this a line-level question since I intend to ask with regards to and solve it by playing with the input circuitry.
This started with an idea to power stuff using the computer's 12V, because most modern PSUs have 30A too much. So I had a molex-to-barrel cable made. Then one day I had a Bose that has its PSU fried and I tried to power up with the computer. Discovered it cannot work because
1) If I power the Bose using the computer and play it using an MP3 player / smartphone / etc. it works
2) If I play the Bose using the computer and use something else for power (i.e. the normal method) it works
3) If I power the Bose using the computer and play it using the computer it goes brapruiosrtaouypastiukp
Conclusion is that this is due to grounding issues - computers have their audio jack's ground at 0V, bias the output to some voltage, then have a dc-blocking cap. Some amplifiers running off a single-rail supply do that too I guess. But then there are those that use virtual ground / rail splitter or something, in which case their ground becomes at midpoint of supply rails which in this case becomes +6V, connecting the two equipment together shorts that +6V to ground. The PSU won't trigger overcurrent until 30A it reached which would happen after the cable melts and the rail splitter circuit dies, meanwhile since the rail splitter's output would be connected to the amplifier's input that sound is literally the rail splitter dying.
So now, how can the circuit be changed so that the computer PSU can still be used while using the sound card of the same computer?
Using a capacitor sounds straightforward. But what about the specifics? If caps on ground do little damage to sound while providing foolproofness surely it would have been used more often already?
Sounds strange, I know, and I'm considering this a line-level question since I intend to ask with regards to and solve it by playing with the input circuitry.
This started with an idea to power stuff using the computer's 12V, because most modern PSUs have 30A too much. So I had a molex-to-barrel cable made. Then one day I had a Bose that has its PSU fried and I tried to power up with the computer. Discovered it cannot work because
1) If I power the Bose using the computer and play it using an MP3 player / smartphone / etc. it works
2) If I play the Bose using the computer and use something else for power (i.e. the normal method) it works
3) If I power the Bose using the computer and play it using the computer it goes brapruiosrtaouypastiukp
Conclusion is that this is due to grounding issues - computers have their audio jack's ground at 0V, bias the output to some voltage, then have a dc-blocking cap. Some amplifiers running off a single-rail supply do that too I guess. But then there are those that use virtual ground / rail splitter or something, in which case their ground becomes at midpoint of supply rails which in this case becomes +6V, connecting the two equipment together shorts that +6V to ground. The PSU won't trigger overcurrent until 30A it reached which would happen after the cable melts and the rail splitter circuit dies, meanwhile since the rail splitter's output would be connected to the amplifier's input that sound is literally the rail splitter dying.
So now, how can the circuit be changed so that the computer PSU can still be used while using the sound card of the same computer?
Using a capacitor sounds straightforward. But what about the specifics? If caps on ground do little damage to sound while providing foolproofness surely it would have been used more often already?
Last edited:
^ Agreed. (I s'pose making a balanced input stage would be pushing things for what seems to be a PC speaker system or somesuch.) The issue is a ground loop rather than differences in potential.
This, btw, probably is why PC power has a bad reputation for being dirty. Understandable if you consider where the ground loop runs in this case.
This, btw, probably is why PC power has a bad reputation for being dirty. Understandable if you consider where the ground loop runs in this case.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.