it seems my problems are related to the grounding of my minidsp's and not related at all to the kits from audiosector. i will nevertheless go back to my "old" psu solution with one battery psu for left and another for the right channel, and add an adittional battery bank for the minidsp's, just to make sure. i am confident that allthough i had problems separating the sound quality from a regular transformer based psu and a battery based one on two sets of lm3875, it will not hurt to hunt for perfection.
i have nothing but good things to say about the 5 or 6 lm4780 kits and the 2-3 lm3875 kits i have purchased, they work wonderfully.
edit:
i am just mentioning this because it might also help you out as well Dean. i do not know how you are doing the grounding of the minidsp but sometimes itbis better with floating ground on them it seems, a pain to work out right in cases like this.
seeing we have similar setups we also have similar problems, and maybe also similar causes.
i have nothing but good things to say about the 5 or 6 lm4780 kits and the 2-3 lm3875 kits i have purchased, they work wonderfully.
edit:
i am just mentioning this because it might also help you out as well Dean. i do not know how you are doing the grounding of the minidsp but sometimes itbis better with floating ground on them it seems, a pain to work out right in cases like this.
seeing we have similar setups we also have similar problems, and maybe also similar causes.
Last edited:
Another update. This is the photo's of the amp on my test bench. Sorry it took so long.
I have disconnected everything but 1 channel from the transformer so the circuit looks like this:
Fuse - Switch - CL60(x2) - 400x22x2 Transformer - LM4780 board
Photo 1) is the amp connected to the Scope only.
Photo 2) is the amp connected to the Scope and a jumper lead connected between the input (+)(@ RCA end) and the chassis earth block.
Photo 3) is the board under inspection to look for possible wiring issues.
A few notes:
1. The amp makes noise at the speakers with or without an input connected, be it the miniDSP or a CD player.
2. The only earth from the amp/psu goes from the (OG) point on the amp board to the main chassis earth Block. This is the hole between the two output leads marked SG & OG
3. I have isolated and checked the input(RCA) and output (Binding post) connectors from the chassis with plastic through hole washers.
4. Whatever mistake I have made is exactly the same on all 6 boards. I have tested them all with the scope and get the same level of noise at the speakers.
5. Have I actually got the earth lead to the board connected at the correct location? Wouldn't that be a laugh......
I have disconnected everything but 1 channel from the transformer so the circuit looks like this:
Fuse - Switch - CL60(x2) - 400x22x2 Transformer - LM4780 board
Photo 1) is the amp connected to the Scope only.
Photo 2) is the amp connected to the Scope and a jumper lead connected between the input (+)(@ RCA end) and the chassis earth block.
Photo 3) is the board under inspection to look for possible wiring issues.
A few notes:
1. The amp makes noise at the speakers with or without an input connected, be it the miniDSP or a CD player.
2. The only earth from the amp/psu goes from the (OG) point on the amp board to the main chassis earth Block. This is the hole between the two output leads marked SG & OG
3. I have isolated and checked the input(RCA) and output (Binding post) connectors from the chassis with plastic through hole washers.
4. Whatever mistake I have made is exactly the same on all 6 boards. I have tested them all with the scope and get the same level of noise at the speakers.
5. Have I actually got the earth lead to the board connected at the correct location? Wouldn't that be a laugh......
Attachments
Hi Dean, that first waveform looks just like what mine did on my LM3886 when I had the 1000uF filter capacitance in the PS (only I think a bit bigger). It's 100Hz so is more than likely charging pulses from the rectifiers. The question is I guess how is it getting into the amp. Mine reduced in amplitude when I upped the capacitance to 5700uF / rail. I'd have to dig up the posts to see how much of a difference it actually made though. I never got it to go away completely, but then I have a different PSU to you as well.
I created some shorting rca's for shorting the inputs. Just got some rca plugs and connected the earth to the centre pin. They should not act as an antena when used. I'm a bit surprised by the result inj the second scope shot, but it could be that you created a large loop area by running the earth back to the earth block rather than having the short right at the rca plug. That may tend to indicate you are picking up airborne noise from the transformers. Are you able to unmount one of the traffos and move it further away from the amp modlue and see if the amplitude varies? Also you could try turning the traffo on it's axis and seeing if that has any effect on the amplitude of the noise.
Tony.
I created some shorting rca's for shorting the inputs. Just got some rca plugs and connected the earth to the centre pin. They should not act as an antena when used. I'm a bit surprised by the result inj the second scope shot, but it could be that you created a large loop area by running the earth back to the earth block rather than having the short right at the rca plug. That may tend to indicate you are picking up airborne noise from the transformers. Are you able to unmount one of the traffos and move it further away from the amp modlue and see if the amplitude varies? Also you could try turning the traffo on it's axis and seeing if that has any effect on the amplitude of the noise.
Tony.
I never thought of creating a shorting RCA and will do that now then take a photo and post back here. So at least we have something to compare.
Peter stated in an earlier post that larger filter caps will not solve this issue, a reduction to below audible at the listening seat would be great but if it is some simple mistake I have made that can be rectified and remove the noise altogether that would be the best outcome.
If I do end up needing to change the caps is there any you would recommend. This would be the last resort but may be required. I know the Pana FC's only go up to 2200uf or something in the 50V spec.
Cheers
Dean
Peter stated in an earlier post that larger filter caps will not solve this issue, a reduction to below audible at the listening seat would be great but if it is some simple mistake I have made that can be rectified and remove the noise altogether that would be the best outcome.
If I do end up needing to change the caps is there any you would recommend. This would be the last resort but may be required. I know the Pana FC's only go up to 2200uf or something in the 50V spec.
Cheers
Dean
Hi Dean, How efficient are your speakers. ~4mV p2p on your noise measurement it pretty low. I'll hook up my LM3886 and have a look at what mine is. I don't think I'd worked out how to get the 2mV sensitivity on the scope when I did the previous measurements, but I suspect mine is not a lot less than yours, and I have to put my ear right beside the speaker to hear the noise.
I note that the latest scope shot is much lower than the previous ones you posted.
My original PS had 1000uF panasonic FC's in it, and I ended up adding 4700uF Nichicon FW's (I bought twice as many as I needed due to some stupidity when designing my regluated supply so had them surplus). I actually posted the difference earlier in this thread here it is again http://www.diyaudio.com/forums/powe...ding-challenge-please-help-7.html#post2519106 Look at the last two pics. 2nd last one is with 1000uF/rail (scope on 5mv/div) and second is with 5700uF/rail. Not a huge difference, but it did mean I had to put my ear right against the speaker to here it after the capacitance upgrade compared to a few cm away before that. My speakers are probably around the 89db 2.83V/1M mark.
Tony.
I note that the latest scope shot is much lower than the previous ones you posted.
My original PS had 1000uF panasonic FC's in it, and I ended up adding 4700uF Nichicon FW's (I bought twice as many as I needed due to some stupidity when designing my regluated supply so had them surplus). I actually posted the difference earlier in this thread here it is again http://www.diyaudio.com/forums/powe...ding-challenge-please-help-7.html#post2519106 Look at the last two pics. 2nd last one is with 1000uF/rail (scope on 5mv/div) and second is with 5700uF/rail. Not a huge difference, but it did mean I had to put my ear right against the speaker to here it after the capacitance upgrade compared to a few cm away before that. My speakers are probably around the 89db 2.83V/1M mark.
Tony.
I raise two points.
You have kept the PSU PCB attached to the amp PCB?
That brings the dirty electricity closer to the clean electricity.
The cables from PSU to AMP are spread apart.
That increases the loop areas and increases the interference.
I cannot see how the Signal return and Speaker return and Zobel return and Power ground and PSU zero volts meet up at your Main Audio Ground.
You have kept the PSU PCB attached to the amp PCB?
That brings the dirty electricity closer to the clean electricity.
The cables from PSU to AMP are spread apart.
That increases the loop areas and increases the interference.
I cannot see how the Signal return and Speaker return and Zobel return and Power ground and PSU zero volts meet up at your Main Audio Ground.
A few more test.
Pic 1) 240v Lead plugged in but amp OFF
Pic 2) Output (no input short)
Pic 3) Output (input Shorted)
Pic 4) Output (no input short) amp earth lead disconnected
Pic 5) Amp out of case (no input short or board earth)
Pic 6) Amp out of case with speaker connected
Pic 7) Amp out of case with speaker connected with input shorted
Pic 8) Pic of amp out of case
I look forward to anybody's advice on this situation. Cant wait to get this sorted and move onto listening to some music.
Cheers
Dean
Pic 1) 240v Lead plugged in but amp OFF
Pic 2) Output (no input short)
Pic 3) Output (input Shorted)
Pic 4) Output (no input short) amp earth lead disconnected
Pic 5) Amp out of case (no input short or board earth)
Pic 6) Amp out of case with speaker connected
Pic 7) Amp out of case with speaker connected with input shorted
Pic 8) Pic of amp out of case
I look forward to anybody's advice on this situation. Cant wait to get this sorted and move onto listening to some music.
Cheers
Dean
Attachments
Well I tried to get some measurements but whatever the interference that pops up at my place from time to time is, it is here tonight with a vengence. I was getting 40mV p2p noise on the output of the amp. I had the scope plugged into an outlet on the other side of the room, decided to plug it into the same powerboard as the amp, and it dropped to about 8mV p2p weird. Usually I get better results if the scope is plugged into a different circuit. Anyway the noise drowns out anything from the PS so totally useless to try and do any measurements tonight. Its a bit higher than usual, around the 1Mhz mark tonight.
edit: just saw your last post Dean, that first pic is rather puzzling. It is showing some 100Hz blips yet you say the amp is turned off. Do you get that if the mains lead is unplugged?
Also have you taken a measurement with just the amp module with no input wiring connected at all (open and shorted at the amp input)? It looks like your input wire runs right over the top of the rectifiers, is it possible to move it away from them?
Tony.
edit: just saw your last post Dean, that first pic is rather puzzling. It is showing some 100Hz blips yet you say the amp is turned off. Do you get that if the mains lead is unplugged?
Also have you taken a measurement with just the amp module with no input wiring connected at all (open and shorted at the amp input)? It looks like your input wire runs right over the top of the rectifiers, is it possible to move it away from them?
Tony.
Thanks for going the extra mile Tony.
I look forward to your results.
Andrew..
1. I haven't tried separating the boards but will
2. Do your mean the (V-,PG-)+(V+PG+) connections? I have some with a short copper link and some with a 20mm wire loop. They test the same. If I separate the boards wont this extend that link?
3. I am not sure about the returns etc either. Here is a link to Peters Schematics.
DIY Chip Amplifier Kits, PCB's, Components and Information.
The attached Pic shows that I only have 1 earth lead from the PSU/AMP combo to the earth Block. Is this incorrect??
Cheers
Dean
I look forward to your results.
Andrew..
1. I haven't tried separating the boards but will
2. Do your mean the (V-,PG-)+(V+PG+) connections? I have some with a short copper link and some with a 20mm wire loop. They test the same. If I separate the boards wont this extend that link?
3. I am not sure about the returns etc either. Here is a link to Peters Schematics.
DIY Chip Amplifier Kits, PCB's, Components and Information.
The attached Pic shows that I only have 1 earth lead from the PSU/AMP combo to the earth Block. Is this incorrect??
Cheers
Dean
Attachments
Last edited:
Hi Dean just going back to this one. When you have the speaker connected up how close do you need to get to hear the noise in pic3? I would say that is probably about as low as the noise on my LM3886 amp and I have to have my ear pretty much planted against the speaker (maybe a case of my speakers are less sensitive/I'm a bit deaf
) You may not be able to completely eliminate the noise. Once again I'd suggest disconnecting your input lead and shorting at the input to the amp board. What this I think will tell you is whether the noise is being picked up by the input wiring, or whether it is coming from the power supply rails directly into the amp.
Tony.
Good morning Peter & Tony. Lets hope I haven't been chasing my own *** on this one.
Most of my testing has been done with no short at the input. As we saw from yesterday's pics, when shorted the scope noise roughly halves.
When I remove the earth wire completely at either end there is no change.
Pic 3 is with the input shorted which I have only listened to using my shed system. It is quieter but I will have to re-install the amp in my main system for a reference.
If this level is acceptable I will need to look at whats happening at the MiniDSP, like I said, chasing my own *** and learning the hard way.
I will report back asap.
Thanks for the patience gents.
Dean
Most of my testing has been done with no short at the input. As we saw from yesterday's pics, when shorted the scope noise roughly halves.
When I remove the earth wire completely at either end there is no change.
Pic 3 is with the input shorted which I have only listened to using my shed system. It is quieter but I will have to re-install the amp in my main system for a reference.
If this level is acceptable I will need to look at whats happening at the MiniDSP, like I said, chasing my own *** and learning the hard way.
I will report back asap.
Thanks for the patience gents.
Dean
Hi Dean, my own experience is that it is difficult to eliminate noise completely (though I'm not saying it is impossible). My LM3886 based amp with input shorted is very quiet, when I connect up my HTPC not so quiet (but still not detectable from the listening position, I decided to live with it
With input shorted, my playmaster series 200 mosfet amp, is dead silent on one channel but the other one has noise (again at fairly low levels not enough to hear from the listening position). I've been trying to get rid of that for years because I don't like the fact that one channel has noise and the other doesn't but I have not managed to
Lots of variables could come into play as well, the efficiency of your speakers and also the ambient noise levels may be vastly different to mine so what might be completely acceptible to me, may not be to you.
Tony.
With input shorted, my playmaster series 200 mosfet amp, is dead silent on one channel but the other one has noise (again at fairly low levels not enough to hear from the listening position). I've been trying to get rid of that for years because I don't like the fact that one channel has noise and the other doesn't but I have not managed to
Lots of variables could come into play as well, the efficiency of your speakers and also the ambient noise levels may be vastly different to mine so what might be completely acceptible to me, may not be to you.
Tony.
Well, there you have it, I am a NOOB!
I built a passive pre box (ALPS pot in a box with a few in and outs) and connected my cd player through the pot too the LM4780 amp.
AND guess what..... No noise at the speakers.
I did say I was determined to learn the hard way, right.
I then connected the miniDSP in line and hey presto. Noise is back.
So the lesson for today. K.I.S.S. Only pull your amp apart twice over once you have tried a different input source.
My thanks to Peter, Tony, Andrew and KaffiMann for there patience and advice.
I am now looking at different power-supplies/earthing for the miniDSP to defeat the noise at that end.
Cheers all
Dean
I built a passive pre box (ALPS pot in a box with a few in and outs) and connected my cd player through the pot too the LM4780 amp.
AND guess what..... No noise at the speakers.
I did say I was determined to learn the hard way, right.
I then connected the miniDSP in line and hey presto. Noise is back.
So the lesson for today. K.I.S.S. Only pull your amp apart twice over once you have tried a different input source.
My thanks to Peter, Tony, Andrew and KaffiMann for there patience and advice.
I am now looking at different power-supplies/earthing for the miniDSP to defeat the noise at that end.
Cheers all
Dean
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.