Thanks, I did try a 33 Ohm signal GND/0v to chassis but will go through again using the board link to the transformer 0v board location (green)
The worst spot.
Good luck, we tried to no avail. The idea to have a plastic bushing to isolate the mounting plate of the toroid (and also its bolt to chassis connection) is a good one. Had this a few times and I decided that I could not explain it (no toroid manufacturer gives bushings in the mounting kit) but it worked 🙂 It worked with floating Audio GND that is, the transformer often is the cause for leakage.
Good luck, we tried to no avail. The idea to have a plastic bushing to isolate the mounting plate of the toroid (and also its bolt to chassis connection) is a good one. Had this a few times and I decided that I could not explain it (no toroid manufacturer gives bushings in the mounting kit) but it worked 🙂 It worked with floating Audio GND that is, the transformer often is the cause for leakage.
Last edited:
I think the location recommended is the orange X, better than the red X ? and this goes to the chassis via a 33 (10-100) Ohm 2W resistor ?
For relocating the signal out (dark blue) where is best the red X for GND ?
For relocating the signal out (dark blue) where is best the red X for GND ?
No you choose the point where the charging currents run so the absolute worst point. 0V is not 0V just like that. You could reread posts or check Rod Elliotts website if you want to to be guided by a name. In this rare case one has chosen right, Rod is a real guru.
Many good tips have been given, maybe you read them only partly?! A hum solver like a few here are solves this in less than an hour.
Many good tips have been given, maybe you read them only partly?! A hum solver like a few here are solves this in less than an hour.
Last edited:
So make sure the bottom bolt that is attached to the chassis is not making contact ? I did wonder if there was any advantage for using a non magnetic stainless bolt ?
Have you stepped through slides 68 and 69 in the Ground Loops presentation? This may help isolate the specific problem.
It’s important to know whether you have a ground loop, cross channel ground loop or some form of common impedance coupling. The noise mechanisms are all different.
Usually, as long as you use the circular rubber mats supplied by the toroidal manufacturer (you can buy them from RS as well) and the plate washer, isolating the mounting bolt is not necessary because there is no chance of a shorted turn through the mounting bolt.
It’s important to know whether you have a ground loop, cross channel ground loop or some form of common impedance coupling. The noise mechanisms are all different.
Usually, as long as you use the circular rubber mats supplied by the toroidal manufacturer (you can buy them from RS as well) and the plate washer, isolating the mounting bolt is not necessary because there is no chance of a shorted turn through the mounting bolt.
I don't think I made it to page 68, more reading ! Yes the transformer is connected to the chassis at the base with the ability to connect but the top has a rubber washer under the metal hat
I'm guessing where I have marked with a green "G" is the output of the power supply filter. But, you should examine the circuit and verify it. However, it is definitely a better than the transformer ground location as a clean ground connection.
I'm guessing that the speaker ground return location is contributing to the noise that you hear. I would definitely move it.
In terms of the potential transformer shorted turn issue, if the head of the mounting bolt is in direct contact with the bottom of the metal chassis and if there is a top cover plate which would be in direct electrical contact with the chassis, then if the nutted top end of the bolt contacts the cover plate, that creates an one turn secondary winding which is a short circuit. This will most likely cause the line fuse to blow.
However, if you don't have a cover plate, then it will not happen in this way.
I'm guessing that the speaker ground return location is contributing to the noise that you hear. I would definitely move it.
In terms of the potential transformer shorted turn issue, if the head of the mounting bolt is in direct contact with the bottom of the metal chassis and if there is a top cover plate which would be in direct electrical contact with the chassis, then if the nutted top end of the bolt contacts the cover plate, that creates an one turn secondary winding which is a short circuit. This will most likely cause the line fuse to blow.
However, if you don't have a cover plate, then it will not happen in this way.
Attachments
G is quite close to the input as it comes in on the yellow circle on the image above. the two pins at the top right are signal in and signal GND, so the location would be between the signal in and big smoothing caps
I did muck about with the transformer (not much room as the connections are quite restricting), but not much changes for small adjustments
I did muck about with the transformer (not much room as the connections are quite restricting), but not much changes for small adjustments
I have a plan to solder up and move the tracks, thanks Pinholder.
Ive also been looking more closely at the 500W versions (that include a pre and are double insulated). looking at the connections for GND and earth. The chassis is connected to signal GND via a screen that connect to the centre of the output caps but direct. The earth from the plug does go to the board but only connected to the signal GND at the point of speaker output -ve via a 0.1uf 400V polyester film cap, so no direct connection
My current plan Red output transformer GND/Centre, Blue to chassis bond, yellow is signal in I could reduce the impedance here as well ? I'm also thinking about how best to connect the earth to the chassis and intrigued by the 400v 0.1uf cap
The layout below green is silhouettes for the transistors, caps and Transformer.
Any other thoughts about reducing impedance ?
Ive also been looking more closely at the 500W versions (that include a pre and are double insulated). looking at the connections for GND and earth. The chassis is connected to signal GND via a screen that connect to the centre of the output caps but direct. The earth from the plug does go to the board but only connected to the signal GND at the point of speaker output -ve via a 0.1uf 400V polyester film cap, so no direct connection
My current plan Red output transformer GND/Centre, Blue to chassis bond, yellow is signal in I could reduce the impedance here as well ? I'm also thinking about how best to connect the earth to the chassis and intrigued by the 400v 0.1uf cap
The layout below green is silhouettes for the transistors, caps and Transformer.
Any other thoughts about reducing impedance ?
Some progress with changes to one amplifier. I broadly followed this layout
By cutting traces and adding connections, also bulking up the 'T' I was on the limit of my 80W iron, so not the best soldering, but happy there are no dry joints and fully molten where it counted. Anyway I added a GND so I can link the GND to the case bond better than briding from an RCA input.
Top left 0vGND, bottom left signal 0v/GND in (I added a bit more copper), top right transformer 0V GND from secondaries, bottom right 0v GND speaker out
But alas, no better, possible the noise has less 'bass'. I will try to measure using REW and take a trace of the noise.
With the chassis connected to the earth pin of the RCA very noisy, ditto when adding the new centre GND to this location. With no chassis GND or earth to chassis then it's just inherent amplifier noise (like hiss), that I cant reduce enough. The inherent amplifier noise is too high to use so still trying to understand what my options are as the 500W versions that have active pre amplifiers quiet as a mouse. I have tried a small pre amplifier in series and this made no difference
The other amplifier is the same so doubt there is a fault
Any other ideas ?
By cutting traces and adding connections, also bulking up the 'T' I was on the limit of my 80W iron, so not the best soldering, but happy there are no dry joints and fully molten where it counted. Anyway I added a GND so I can link the GND to the case bond better than briding from an RCA input.
Top left 0vGND, bottom left signal 0v/GND in (I added a bit more copper), top right transformer 0V GND from secondaries, bottom right 0v GND speaker out
But alas, no better, possible the noise has less 'bass'. I will try to measure using REW and take a trace of the noise.
With the chassis connected to the earth pin of the RCA very noisy, ditto when adding the new centre GND to this location. With no chassis GND or earth to chassis then it's just inherent amplifier noise (like hiss), that I cant reduce enough. The inherent amplifier noise is too high to use so still trying to understand what my options are as the 500W versions that have active pre amplifiers quiet as a mouse. I have tried a small pre amplifier in series and this made no difference
The other amplifier is the same so doubt there is a fault
Any other ideas ?
Do not connect the RCA earth to chassis.
Connect the ground wire from the amplifier to the chassis through a ground loop breaker. This would be the only connection from the amp circuit to chassis.
Connect the ground wire from the amplifier to the chassis through a ground loop breaker. This would be the only connection from the amp circuit to chassis.
Bill, thanks, more testing using a connection from the speaker out to RCA connection with no signal input to the amplifier, when listening the BPSP 500 is silent you have to get within 1 foot to know its on in a silent room.
I took measurements using the RTA analyser in REW. Actual figures are not super helpful as feeding in 1V into the DAC showed 0.74V in REW (Too much effort to calibrate)
The Beringer input DAC (a very cheap A>D) shows some noise bottom blue trace this is with no connections or lead attached. The Green is the BK1000 with mods above, no signal input connected directly to the DSP output. The violet trace is the BPSP 500 connected in the normal way as this is near silent as noted. BTW the Green is the BK1000 connected directly to the DSP in the same way the BPSP500 is. This is not super loud but just loud enough to hear at the seat.
The pink is the same as green but the case earthed to the power lead as a safety earth. I have about 12 different readings and thinking perhaps there is something in the 500 circuit that reduces higher frequency output. Still thinking here
All traces at 1/6 smoothing
Oh and no clue how I solve the safety earth noise problem
I took measurements using the RTA analyser in REW. Actual figures are not super helpful as feeding in 1V into the DAC showed 0.74V in REW (Too much effort to calibrate)
The Beringer input DAC (a very cheap A>D) shows some noise bottom blue trace this is with no connections or lead attached. The Green is the BK1000 with mods above, no signal input connected directly to the DSP output. The violet trace is the BPSP 500 connected in the normal way as this is near silent as noted. BTW the Green is the BK1000 connected directly to the DSP in the same way the BPSP500 is. This is not super loud but just loud enough to hear at the seat.
The pink is the same as green but the case earthed to the power lead as a safety earth. I have about 12 different readings and thinking perhaps there is something in the 500 circuit that reduces higher frequency output. Still thinking here
All traces at 1/6 smoothing
Oh and no clue how I solve the safety earth noise problem
I will try these two options and the 100nf GND to case for the RCA inlet. Still seems odd to me why earthing an unconnected case and not signal can be noisy
The confusion for me is why when I have this connection as shown below it's noisy ? The signal GND is picked up via the incoming RCA the same as every other power amp in the system and the majority of the power amps have the same connection with the case earthed and no connection to GND at all from the local earth or case.
The comments suggest that this solves a problem when the earth is connected to the chassis AND the 0v signal. In my case the noise is when I connect the earth to the case and there is NO connection to the 0v line. It is the same problem when the the chassis, earth and 0v are connected
"The loop breaker works by adding a resistance in the earth return circuit. This reduces circulating loop currents to a very small value, and thus 'breaks' the loop. The capacitor in parallel ensures that the electronics are connected to the chassis for radio frequency signals, and helps to prevent radio frequency interference. Finally, the diode bridge provides the path for fault currents. The use of a large chassis mounting (35A) type is suggested, since this will be able to handle the possibly very high fault currents that may occur without becoming open circuit. Note the way the bridge is wired, with the two AC terminals shorted, and the two DC terminals shorted. Other connection possibilities are dangerous, and must be avoided."
"The loop breaker works by adding a resistance in the earth return circuit. This reduces circulating loop currents to a very small value, and thus 'breaks' the loop. The capacitor in parallel ensures that the electronics are connected to the chassis for radio frequency signals, and helps to prevent radio frequency interference. Finally, the diode bridge provides the path for fault currents. The use of a large chassis mounting (35A) type is suggested, since this will be able to handle the possibly very high fault currents that may occur without becoming open circuit. Note the way the bridge is wired, with the two AC terminals shorted, and the two DC terminals shorted. Other connection possibilities are dangerous, and must be avoided."