Hi - I tried searching but am only finding esoteric and complicated discussions... I am putting together (past tense, really) one of briangt's (chipamp.com) lm3886 kits. I managed to wade through a lot of challenges and have the amp nicely mounted to heatsinks in an oak box with a toroidal transformer, etc.. a couple of questions.
1) I have some hum. The amplifier PCB schematic has an output for "chassis ground" which the instructions suggest to connect to my "star ground" but since the chassis itself isn't conductive, I'm not really sure what this means. The power supply has two "grounds" as it produces +25vdc and -25vdc, do these get connected to the star ground? What about my RCA plugs... they each have a "ground" that are summed where they connect into my potentiometer... do I need to wire up the RCA input "ground" to the star ground? I don't really get it... sorry for such newbie questions.
2) Sort of in relation to #1 - the earth ground from my AC mains... do I connect that to my star ground? I didn't think it made much sense to ground the chassis as it's made out of oak.
I'm amazed the damn thing works so far, to be honest.
1) I have some hum. The amplifier PCB schematic has an output for "chassis ground" which the instructions suggest to connect to my "star ground" but since the chassis itself isn't conductive, I'm not really sure what this means. The power supply has two "grounds" as it produces +25vdc and -25vdc, do these get connected to the star ground? What about my RCA plugs... they each have a "ground" that are summed where they connect into my potentiometer... do I need to wire up the RCA input "ground" to the star ground? I don't really get it... sorry for such newbie questions.
2) Sort of in relation to #1 - the earth ground from my AC mains... do I connect that to my star ground? I didn't think it made much sense to ground the chassis as it's made out of oak.
I'm amazed the damn thing works so far, to be honest.
Well, I read both articles and I'm not sure I understood them. I've tried wiring up a star ground in several ways... with and without a connection to earth-ground, with the star at the RCA inputs, with a star at the PS outputs, with the star at the earth ground junction, and none of it seems to get rid of the 60hz hum... additionally - there are two forms of 60hz hum...
1) with no RCA connections (only speakers plugged in) there is a very faint 60hz hum that does not depend on the value of the potentiometer. It also dissappears the instant the connection to the transformer is severed (when the amp is just draining the capacitors)
2) with RCA's, there is a 60hz hum that varies in audibility with the potentiometer and persists after the transformer is off.
I'm guessing #1 has to do more with the power supply design and physical proximity of the transformer to the amplifier boards... however, one of my amplifier boards is very close to the transformer where the other is a significant distance away... yet the intensity of the hum is nearly equivalent between the outputs....
I'm also guessing #2 has more to do with the balanced/unbalanced stuff in those articles but no amount of ground configurations seems to alleviate this problem.
For what it's worth, the two RCA inputs do sum their grounds at the potentiometer... so maybe that's a problem? Should I only star ground them and avoid summation anywhere else? (I think I pulled that out of article #2). Note that this is a non-trivial modification at this point, due to the way I constructed this amp... so I'd rather not just loosely experiment unless it's actually a problem.
1) with no RCA connections (only speakers plugged in) there is a very faint 60hz hum that does not depend on the value of the potentiometer. It also dissappears the instant the connection to the transformer is severed (when the amp is just draining the capacitors)
2) with RCA's, there is a 60hz hum that varies in audibility with the potentiometer and persists after the transformer is off.
I'm guessing #1 has to do more with the power supply design and physical proximity of the transformer to the amplifier boards... however, one of my amplifier boards is very close to the transformer where the other is a significant distance away... yet the intensity of the hum is nearly equivalent between the outputs....
I'm also guessing #2 has more to do with the balanced/unbalanced stuff in those articles but no amount of ground configurations seems to alleviate this problem.
For what it's worth, the two RCA inputs do sum their grounds at the potentiometer... so maybe that's a problem? Should I only star ground them and avoid summation anywhere else? (I think I pulled that out of article #2). Note that this is a non-trivial modification at this point, due to the way I constructed this amp... so I'd rather not just loosely experiment unless it's actually a problem.
You should wire your RCA screen/grounds from RCA to pot bottom to pcb ground. Don't connect that to any 'star ground'. The confusion always is that RCA ground really isn't any ground. It is just the reference for the signal coming from the source, and should stay that way past the volume pot (if you have one) onto th pcb where the signal and screen get connected. On a well designed pcb there should be a connection to it.
Then the 'references' of the supply (the 'ground' points of the reservoir caps and any center taps) should be connected to a star point, together with the load return, which in a power amp is the speaker 'ground' lead. Don't connect the speaker return to the pcb. The main issue is to keep ground currents away from 'reference' ground points.
If you have a feedback and/or input resistor that has to grounded, do it also at the pcb, at the same point where you connected those RCA screen/ground leads, NOT at the heavy current star ground.
This all assumes a correctly designed pcb in the first place, and then it is a sure fire way to avoid hum & interference. Doing it this way I left these problems behind decades ago.
Jan Didden
Then the 'references' of the supply (the 'ground' points of the reservoir caps and any center taps) should be connected to a star point, together with the load return, which in a power amp is the speaker 'ground' lead. Don't connect the speaker return to the pcb. The main issue is to keep ground currents away from 'reference' ground points.
If you have a feedback and/or input resistor that has to grounded, do it also at the pcb, at the same point where you connected those RCA screen/ground leads, NOT at the heavy current star ground.
This all assumes a correctly designed pcb in the first place, and then it is a sure fire way to avoid hum & interference. Doing it this way I left these problems behind decades ago.
Jan Didden
janneman said:
OK - so just have RCA---->pot---->amplifier PCB ... no chassis/star grounding...
So - I don't think my toroid transformer is a center tap so I guess that doesn't apply here...
What about the "ground" of the power supply output... as in, the +vdc ground and the -vdc ground? They should go to star ground, along with the negative (ground) terminal of the speaker output? Along with earth ground, I'm assuming?
Got it - well, I didn't design the PCB but I'm assuming others have had success with it. I know these are common problems and so I do appreciate the help. I think I understand, but if you could "check my work" before I go hacking away on this thing again, I'd appreciate it
alright well I don't think it's my grounding topology... placing the wires in various positions in the chassis seem to decrease (or increase) the hum intensity... I think I should have designed a case with the amplifier boards further from the transformer!
edit - yes this is definitely the problem. The letdown is that I can't completely redesign the architecture of the case... I'll just have to keep this in mind for future reference... no wonder even low power amps seem to use large chassis - the further the PS is from the amplifier, the better, right?
edit - yes this is definitely the problem. The letdown is that I can't completely redesign the architecture of the case... I'll just have to keep this in mind for future reference... no wonder even low power amps seem to use large chassis - the further the PS is from the amplifier, the better, right?
john blackburn said:
I've re-routed the wires some, and although if they bend / stray from their current position, the hum returns, for the time being, the hum is all but gone, so i can now enjoy.
The sound is, frankly, not much different than most other class a/b amps i've ever listened to...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.