Hmm... I would look at mains grounding as a loop culprit. I've seen this before, where both the preamp and power amp have their circuit grounds connected to the ground at the mains outlet. By connecting signal ground between the two pieces of equipment, you'll get a great big loop. I've seen a case where two power amps, each connected to mains ground, are then further connected through the signal grounds going back to the preamp. Lifting the mains ground in the preamp may help. When lifting mains ground connections, a good idea might be to use a thermistor instead of something like 10R fixed. Nelson Pass does this on some of his designs. It provides better safety if there is a ground fault as it's resistance will drop when loaded.
I don't know if splitting the PSU will solve the problem. I'd save that as a last-ditch effort. Focus on ground loops and high current wiring first. From your initial description, I think you're clearly picking up power supply rectification pulses as signal noise. Rectification pulses in the power supply are high current bursts which can produce high magnetic field strength. When dealing with magnetic fields, everything pretty much boils down to electrical loops, and their physical size, or area. Current flowing around any electrical loop will "transmit" magnetic flux, and any circuit which forms a loop will pick up a current when immersed in a changing magnetic flux. So if your power supply is producing pulses of high current which flow around a loop, then any ground loop or circuit which picks up this magnetic flux will show these pulses. I haven't seen the pictures, but one thing you might try is twisting power supply wires together in pairs, where the current will flow in equal and opposite amounts through each wire... leads from trafo to rectifier bridge can be twisted together,same with leads from bridge to PSU caps, and likewise leads from PSU to amp circuit. Keeping all these wires short will help too. I'd like to draw a diagram, but I'm at work right now... maybe this evening I'll whip up a sketch for you.
I don't know if splitting the PSU will solve the problem. I'd save that as a last-ditch effort. Focus on ground loops and high current wiring first. From your initial description, I think you're clearly picking up power supply rectification pulses as signal noise. Rectification pulses in the power supply are high current bursts which can produce high magnetic field strength. When dealing with magnetic fields, everything pretty much boils down to electrical loops, and their physical size, or area. Current flowing around any electrical loop will "transmit" magnetic flux, and any circuit which forms a loop will pick up a current when immersed in a changing magnetic flux. So if your power supply is producing pulses of high current which flow around a loop, then any ground loop or circuit which picks up this magnetic flux will show these pulses. I haven't seen the pictures, but one thing you might try is twisting power supply wires together in pairs, where the current will flow in equal and opposite amounts through each wire... leads from trafo to rectifier bridge can be twisted together,same with leads from bridge to PSU caps, and likewise leads from PSU to amp circuit. Keeping all these wires short will help too. I'd like to draw a diagram, but I'm at work right now... maybe this evening I'll whip up a sketch for you.
I think you mentioned the portable Cd player earlier but I just didn't pay attention I suppose..
The cable you are using, does it have the ground connected in the 3.5 mm plug to both left and right channel cables? That might be a problem since that is a long loop. However from your posts it seems the problems are there even when inputs are grounded.
/UrSv
The cable you are using, does it have the ground connected in the 3.5 mm plug to both left and right channel cables? That might be a problem since that is a long loop. However from your posts it seems the problems are there even when inputs are grounded.
/UrSv
Yea, it has both channels grounded in the cable, but the noise is still there when the inputs are just grounded.
Fixed some pics of the waveforms at the speaker terminals:
Right chan, 50mV/div
http://daviruz.homelinux.org/pics/amp/right-50mVdiv.jpg
Right chan, 5mV/div
http://daviruz.homelinux.org/pics/amp/right-5mVdiv.jpg
Left chan, 5mV/div
http://daviruz.homelinux.org/pics/amp/left-5mVdiv.jpg
Fixed some pics of the waveforms at the speaker terminals:
Right chan, 50mV/div
http://daviruz.homelinux.org/pics/amp/right-50mVdiv.jpg
Right chan, 5mV/div
http://daviruz.homelinux.org/pics/amp/right-5mVdiv.jpg
Left chan, 5mV/div
http://daviruz.homelinux.org/pics/amp/left-5mVdiv.jpg
What about bringing the thing up on a high bridge or an overpass, letting it go? ;=)
Just kidding. Kind of frustrating, even for us, though. Hope you sort it out soon.
I'm not sure if I'm not paying attention, but what happens when you completely disconnect one board? Tried both boards by themselves?
//magnus
Just kidding. Kind of frustrating, even for us, though. Hope you sort it out soon.
I'm not sure if I'm not paying attention, but what happens when you completely disconnect one board? Tried both boards by themselves?
//magnus
If your chassis is plastic that means your earth ground is already lifted, try connect the star point of PSU to Earth with a jump wire to see what it does to the hum.
What I meant is from the speaker terminal -ve to the star point of PSU and Earth to the star point of PSU and the input -ve to the Earth point. Use jump wire with clips so less soldering for now. Make sure you only jump grounds( clean and dirty grounds only).
Good Luck.
What I meant is from the speaker terminal -ve to the star point of PSU and Earth to the star point of PSU and the input -ve to the Earth point. Use jump wire with clips so less soldering for now. Make sure you only jump grounds( clean and dirty grounds only).
Good Luck.
swede: Hm, i have tried the channels separatly, can't remember the resulits though, i could test it again.
If the hum came from the bridge, wouldn't it spread to both channels then? I mean, one channel is very clean and the other one is totally ****ed up
chris ma: Im sorry, but i dont know what you mean
-ve, is that ground? In that case i have tried everything you suggested.
What do you mean by "jump grounds"?
If the hum came from the bridge, wouldn't it spread to both channels then? I mean, one channel is very clean and the other one is totally ****ed up
chris ma: Im sorry, but i dont know what you mean
-ve, is that ground? In that case i have tried everything you suggested.
What do you mean by "jump grounds"?
I got the same problem with my amp about a year ago; i solved it connecting al the grounds to a start point obatined with a plated block, like those used in car system to collect all the grounds from the amps, and then connecting it with a 4/5 inch cable to the caps. Avoid to connect directly any ground return (either supply or signal or speaker) between to the caps or to their terminals because their charge and discharge cycles can cause a great amount of noise.
DaViruz,
Sorry about my expression, jump grounds, like jump start a car with cables, connect dirty grounds together, connect signal grounds together, and jump/connect the two gathering points by a single wire.
Last time I solved my gound problem was to leave the amp as is, used a hook up wire experiment the results by shorting/ connecting different ground points, notice which connection would increase or decrease the ground noise. I used a old disposable 8 ohm speaker as load and listen to the noise.
Sorry about my expression, jump grounds, like jump start a car with cables, connect dirty grounds together, connect signal grounds together, and jump/connect the two gathering points by a single wire.
Last time I solved my gound problem was to leave the amp as is, used a hook up wire experiment the results by shorting/ connecting different ground points, notice which connection would increase or decrease the ground noise. I used a old disposable 8 ohm speaker as load and listen to the noise.
I just tried running the channels separate, the ledt one was silent and the right one still made that annoying sound. I was pretty sure botch channels were silent before, but that was appearantly wrong.
Oh, and here is the board i designed for this amp, it isn't exactly the same though, this is an enhanced version i made after i made the acrual boards for the amp. The only difference is that i moved around the components.
http://daviruz.homelinux.org/pics/amp/poweramp.png
Oh, and here is the board i designed for this amp, it isn't exactly the same though, this is an enhanced version i made after i made the acrual boards for the amp. The only difference is that i moved around the components.
http://daviruz.homelinux.org/pics/amp/poweramp.png
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.