markiemrboo said:
Ok, thanks very much Andrew!
Unfortunately I will have to get back to you a little later on to tell you if this gets rid of the hum (though I am sure it will, you seem to know your stuff!). My brother is playing his music at the minute, so it's quite hard to listen out for such a quiet humming
I wonder if my grounding mess might also be the reason why one regulator board was getting so much hotter than the other... hmm... probably not, but time shall tell!
Ok. I have separated the input and speaker return / 0v so both boards are identical again. I also removed the connection joining the RCA grounds at the RCA sockets, so this should be as you asked.
The humming is reduced from when I originally posted, but there's still something there. It's very quiet though, and I can't hear it unless my ears right up against the woofer. It's a sort of buzzing clicking noise actually, rather than a hum.
It now looks like this as another rough diagram...
Attachments
markiemrboo said:
Further thinking, this could just be the lack of a wire from the PSU boards to the disconnecting network / split up the 0v and speaker return on the amp board? The clicking / buzzing noise might be what people refer to as "charging pulses" perhaps!?
I didn't do this before as I was getting no humming even without those. I will do this, probably tomorrow now, and let you know if it makes the noise completely disappear. I have a feeling it will...
markiemrboo said:
Ok, I cut the 0v and speaker return trace on one amp board and ran a wire to the disconnecting network for 0v. It's even quieter, and I think it's changed it's sound again
It seems less clicking and more humming.I think I might rejoin the trace and move the wire to the PSU board to see if that makes any difference...
Nordic said:
Good call Nordic. It seems even quieter now that I moved the RCA lead (from preamp to GC) away from running over not one but TWO mains power cables
I think I can still detect the very very slightest noise with my ear basically glued to the woofer. It really is so barely audible I dont think it's worth worrying about at all. Heck, it's so quiet I might even be just imagining it...
Seems like it's just a case of getting some shielded cables now, hopefully!
Thanks ever so much to everyone
Absolutely. Grounding has been quite a battle for me, so I will definately share what has worked for me.
I'll do that tomorrow. There is an audible difference between the separating of the 0v and speaker return for me it seems aswell, the hum is still slightly audible on the board which I didn't separate yet so I want to play about with that a bit before I post.
Also, I dont think the regulators (I posted about this in another thread the other day) on one board are getting quite as hot as before either. I will do some more testing of that tomorrow aswell, as I only have a single speaker connected testing one channel at a time.
I'll do that tomorrow. There is an audible difference between the separating of the 0v and speaker return for me it seems aswell, the hum is still slightly audible on the board which I didn't separate yet so I want to play about with that a bit before I post.
Also, I dont think the regulators (I posted about this in another thread the other day) on one board are getting quite as hot as before either. I will do some more testing of that tomorrow aswell, as I only have a single speaker connected testing one channel at a time.
AndrewT said:
Hi there, I have just built up the ChipAmp LM3886 Dual Mono kits and I run into the same problem. They are currently just mounted on a sheet of MDF, when I connect both RCA leads from my Cambridge Audio pre-amp I get a low freq hum, when I only have one connected it goes away. The boards themselves appear to have a common chassis ground connection, which I have connected to the safety earth at the moment. I did find that if I connected the chassis ground connection to one of the RCA grounds the hum reduced but did not go away.
I am going to try it with a different pre-amp and see if my mileage varies. However, any suggestions would be welcome.
Ian
ianpengelly said:
I am using my Cambridge Audio A5 as a pre-amp. I have a hiss coming from the tweeter, which I think comes from it and not the GC. I have just checked this morning and if I turn the pre-amp all the way up I get some barely audible (ear right up close) buzzing / humming coming from the speakers connected to the Cambridge Audio, I can also hear what I think is the same noise, but actually quieter, coming from the GC. For now I might assume that the remaining hum is coming from the pre-amp (if that's possible?), or maybe it is just something to do with my board layout. I tried adding another ground wire, but on the PSU, going to the disconnecting and it didn't make any difference what-so-ever.
I'm not familiar with the boards you have (I made my own), but I have virtually got rid of my humming by separating, on the amp board:
The signal return and feedback ground
Speaker return, chip pin 7 ( I think) ground and mute ground
Power 0v
(see attached diagram / schematic)
So I have three ground wires coming off of the amp board, and I am running them all to the same point at the disconnecting network. I assume that I could have left the power 0v connected to the speaker return and then instead run this wire from the PSU boards and it would have done the same thing, but it was just easier for me to remove the amp boards from the case now, so I did it there instead.
I'm not actually sure if this is actually a 'true' or 'proper' star ground. I don't know if in a true star ground the grounds are supposed to be totally separate from each other, no trace joining them on the PCB? It seems to have worked well for me though.
Given that it's a kit, I imagine you shouldn't really have to do any modifying of the PCB's to remove hum though, so it might be best to wait for someone with experience using the PCB's you have. Maybe even e-mail the people who make the PCB you have and ask about it?
You might also want to check that your signal cables (both internal amp and the interconnect from pre to GC) aren't running near any power supply or mains cabling, especially if they are unshielded like mine. I had my interconnect laying over the mains cable going to the GC and quite close to my pre-amp mains cable. Moving it out of the way of this removed my last bit of audible hum.
Attachments
Hi,
you have got it right.
I would tidy up a few ends.
1. move the speaker return straight to your off board central star ground (audio ground).
2. move the audio ground to nearer a central point between PCB and PSU.
3. make a longer connection from disconnecting network to the re-located audio ground.
4. all the transformer pairs and rectifier pairs could be twisted together, as pairs, to minimise the externally radiated electromagnetic field (it's that small loop area thing) and it helps if each pair can be kept short.
You have designed your PCB with separated signal ground and power ground.
Most chipamp board designers don't do that and then the problems start because they cannot keep the dirty decoupling returns from contaminating the signal ground.
Carlos' schematic shows all the grounds with the same symbol and most builders do not recognise that all the grounds should be kept separate until they meet at the audio ground. It would be far better if all designers would show differnt symbols for the different types of ground (clean, dirty, very dirty (PSU) and mains safety earth, that's 4. There is an exception or two.
the decoupling should connect to each other by the shortest possible route and the PCB should be designed to allow this very short ground to ground route, particularly for the tiny high frequency caps that should be almost touching each other to allow proper high frequency bypass action. The common point is then connected to the audio ground.
The other exception is the RCA input ground. The barrel connection can go direct to audio ground or to the PCB signal ground. Whichever is quieter with the sources used is the correct one. An ideal situation for a bolted connection that can be swapped over and a pin to pin shorting plug.
you have got it right.
I would tidy up a few ends.
1. move the speaker return straight to your off board central star ground (audio ground).
2. move the audio ground to nearer a central point between PCB and PSU.
3. make a longer connection from disconnecting network to the re-located audio ground.
4. all the transformer pairs and rectifier pairs could be twisted together, as pairs, to minimise the externally radiated electromagnetic field (it's that small loop area thing) and it helps if each pair can be kept short.
You have designed your PCB with separated signal ground and power ground.
Most chipamp board designers don't do that and then the problems start because they cannot keep the dirty decoupling returns from contaminating the signal ground.
Carlos' schematic shows all the grounds with the same symbol and most builders do not recognise that all the grounds should be kept separate until they meet at the audio ground. It would be far better if all designers would show differnt symbols for the different types of ground (clean, dirty, very dirty (PSU) and mains safety earth, that's 4. There is an exception or two.
the decoupling should connect to each other by the shortest possible route and the PCB should be designed to allow this very short ground to ground route, particularly for the tiny high frequency caps that should be almost touching each other to allow proper high frequency bypass action. The common point is then connected to the audio ground.
The other exception is the RCA input ground. The barrel connection can go direct to audio ground or to the PCB signal ground. Whichever is quieter with the sources used is the correct one. An ideal situation for a bolted connection that can be swapped over and a pin to pin shorting plug.
AndrewT said:
Excellent!
Ok.
I have read about and done this as much as I could
Ah! I actually designed it so they were all connected in a straight ground line at the bottom of the PCB, so it was really easy to just cut the trace and separate them up if I needed to. Turned out to be a reasonably smart move on my part obviously!
Yes, it would be quite nice for schematics to show the different grounds for idiots like myself
Thanks again for yet another informative post. Much appreciated!
Hi Markie,
you are not the idiot.
You are the learner.
The idiots are the schematic drawers that take the easy and quick route that happens to suit the CAD systems and particularly the simulators that require all the gorunds to be connected, even though some of them know that the different grounds/references do different jobs.
Well, what about those that don't know? They should stick to chewing toffees and watching the football, whatever it's shape.
you are not the idiot.
You are the learner.
The idiots are the schematic drawers that take the easy and quick route that happens to suit the CAD systems and particularly the simulators that require all the gorunds to be connected, even though some of them know that the different grounds/references do different jobs.
Well, what about those that don't know? They should stick to chewing toffees and watching the football, whatever it's shape.
If you have a pot on the front of your case, try un-hooking it from the case and hang it off to the side and make sure it is not touching the case.
I had a class D amp in a metal case and even though the inputs and outputs from the pot were isolated from the case, the pot shaft was grounded to mains earth and causing the hum.
I had a class D amp in a metal case and even though the inputs and outputs from the pot were isolated from the case, the pot shaft was grounded to mains earth and causing the hum.
Hi Andrew,
Just reading through the advice are you suggesting that with the Chipamp.com LM3886 PCBs, which have a common ground plane that I may need to either cut the traces or run the leads from things like the 0V power in, the speaker return and the signal return to star ground point?
I will feel somewhat let down if I have paid good money for a not very well designed board!
Ian
Just reading through the advice are you suggesting that with the Chipamp.com LM3886 PCBs, which have a common ground plane that I may need to either cut the traces or run the leads from things like the 0V power in, the speaker return and the signal return to star ground point?
I will feel somewhat let down if I have paid good money for a not very well designed board!
Ian
Hi Ian,
don't get over worried about it.
I lived with the same errors in my module kits for nearly 30years before I came to realise how little so many builders and alleged designers know.
The chipampers seem to be the worst and this is made doubly worse because the chip combines some of the clean and dirty routes internally.
don't get over worried about it.
I lived with the same errors in my module kits for nearly 30years before I came to realise how little so many builders and alleged designers know.
The chipampers seem to be the worst and this is made doubly worse because the chip combines some of the clean and dirty routes internally.
Hi Andrew,
Thanks for getting back to me, having done some more searching I finally came across another thread involving a board similar to mine and you actually helped the guy out there as well.
Basically there is a CSG, which is a connection for the chassis ground. As the kit I have features 2 mono boards and rectifiers, the two CSG points need connecting together with a nice short, thick copper wire (I am presuming the thick bit as shortening them considerably reduced the humming, to a much more tolerable level. I am going to go and solder in a 1.5mm2 solid core copper cable and see how I get on with that.
Maybe not such a bad board after all (sorry Viklash!), but it has been driving me nuts trying to figure out why it was humming!
Ian
Thanks for getting back to me, having done some more searching I finally came across another thread involving a board similar to mine and you actually helped the guy out there as well.
Basically there is a CSG, which is a connection for the chassis ground. As the kit I have features 2 mono boards and rectifiers, the two CSG points need connecting together with a nice short, thick copper wire (I am presuming the thick bit as shortening them considerably reduced the humming, to a much more tolerable level. I am going to go and solder in a 1.5mm2 solid core copper cable and see how I get on with that.
Maybe not such a bad board after all (sorry Viklash!), but it has been driving me nuts trying to figure out why it was humming!
Ian
I have just soldered in a connection that goes from the CSG of one board directly to the CSG of the other board, is that what you were suggesting?
There is still a little hum when I put my ear close to the woofers, but until I turn off the computer I won't know how bad it is!
At some point, depending on the chassis they go in (they are mounted on a sheet of plywood at the moment).
There is still a little hum when I put my ear close to the woofers, but until I turn off the computer I won't know how bad it is!
At some point, depending on the chassis they go in (they are mounted on a sheet of plywood at the moment).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.