Hi Community !
I build a new amplifier (Inspirated of F7 Amplifier /Thank you Mr PASS), but issue with it.
Noise (50Hz) on both channel when I plug the protective earth on the chassis.
Without protective earth ... Dead Silence on outputs.
After few mesurment ... Without protective earth connected, i've ~60VAC on the aluminium enclosure !
i've try to reverse 240V input plug ... same issue.
However, no electic contact between amp boards , capacitor bank , softstart , transformers, outputs or inputs AND the aluminium enclosure ... My questions :
- Why 60VAC present on my enclosure if no electric contact with other parts ?
- Why with protective earth connected i've a noise of 50hz on my speaker outputs (both channel) ?
I've try to disasemble the enclosure , and split the heatsinks with amp board to the power supply section ... and it's ok ... I don't understand.
Wiring diagram and photos
Thank you in advance for your help
I build a new amplifier (Inspirated of F7 Amplifier /Thank you Mr PASS), but issue with it.
Noise (50Hz) on both channel when I plug the protective earth on the chassis.
Without protective earth ... Dead Silence on outputs.
After few mesurment ... Without protective earth connected, i've ~60VAC on the aluminium enclosure !
i've try to reverse 240V input plug ... same issue.
However, no electic contact between amp boards , capacitor bank , softstart , transformers, outputs or inputs AND the aluminium enclosure ... My questions :
- Why 60VAC present on my enclosure if no electric contact with other parts ?
- Why with protective earth connected i've a noise of 50hz on my speaker outputs (both channel) ?
I've try to disasemble the enclosure , and split the heatsinks with amp board to the power supply section ... and it's ok ... I don't understand.
Wiring diagram and photos
Thank you in advance for your help
The leakage you are measuring most probably comes through the mains transformer. A common issue.
So you need to connect the PE without getting any current diversion through it. The usual way is to connect to PE through some impedance, like back-to-back diodes or a capacitor.
Others here will have suggestion for that.
Jan
So you need to connect the PE without getting any current diversion through it. The usual way is to connect to PE through some impedance, like back-to-back diodes or a capacitor.
Others here will have suggestion for that.
Jan
I've try to add a ground loop breaker, like this :
With this quick fix : No 60VAC on the chassis, No Humm 50Hz on my speaker output, but in my opinion, it's not a great solution ...
I am not convinced by the connection of the 0 V power supply unit to earth through this "Earth loop breaker" ...
Another possible solution according to you
With this quick fix : No 60VAC on the chassis, No Humm 50Hz on my speaker output, but in my opinion, it's not a great solution ...
I am not convinced by the connection of the 0 V power supply unit to earth through this "Earth loop breaker" ...
Another possible solution according to you
It's a bit mysterious why separating the amplifiers would eliminate the hum... I can only guess capacitive coupling of the power wiring to the environment?
I posted about this just earlier today.
In DIY power amps you may run across the problem of too much capacitive coupling to the mains in the transformer (easily 500-900 pF, orientation dependent). A Class II design above a certain power level (let's say 20-30 VA transformer or so) pretty much requires a shield winding in the transformer. Otherwise hum on open inputs is quite likely as amplifier ground floats at part of mains voltage but its surroundings do not.
If your transformer is devoid of such luxury, what you can do is using a 3-conductor mains cable to bring in the protective earth connection, which then is used only for a safety capacitor (4.7-10 nF Y2, typ) going to secondary-side ground. (Build the device to Class II standards otherwise.)
This connection is quite low impedance compared to transformer leakage capacitance so largely eliminates the hum, but high impedance compared to our unbalanced connection shields so would still be unproblematic in terms of ground loops.
In a case like this it would be possible to float just the amplifier circuitry while leaving the chassis connected to PE, however then you need isolated input jacks, and in terms of RF it's always best to have RCA ground bonded to the back panel.
I posted about this just earlier today.
In DIY power amps you may run across the problem of too much capacitive coupling to the mains in the transformer (easily 500-900 pF, orientation dependent). A Class II design above a certain power level (let's say 20-30 VA transformer or so) pretty much requires a shield winding in the transformer. Otherwise hum on open inputs is quite likely as amplifier ground floats at part of mains voltage but its surroundings do not.
If your transformer is devoid of such luxury, what you can do is using a 3-conductor mains cable to bring in the protective earth connection, which then is used only for a safety capacitor (4.7-10 nF Y2, typ) going to secondary-side ground. (Build the device to Class II standards otherwise.)
This connection is quite low impedance compared to transformer leakage capacitance so largely eliminates the hum, but high impedance compared to our unbalanced connection shields so would still be unproblematic in terms of ground loops.
In a case like this it would be possible to float just the amplifier circuitry while leaving the chassis connected to PE, however then you need isolated input jacks, and in terms of RF it's always best to have RCA ground bonded to the back panel.
Last edited:
RFI filter usually it contains two capacitors between line - if you disconnect it from mains side but on filter exit its connected to chassis - you will get some voltage (approx half of mains) on chassis...
but once i had hum when source connected - and the reason was faulty C13 cable..
but once i had hum when source connected - and the reason was faulty C13 cable..
Attachments
I think you have two problems here.
My first suggestion is that you remove the EMI filter. You most likely have Y caps As shown by osscar above that will cause a problem like this if the PE is not connected (which it should be for safety!)
Once you’ve done this, you can start to debug the 50 Hz noise problem you’ve got.
Can you confirm your speaker 0V are not connected to the chassis?
My first suggestion is that you remove the EMI filter. You most likely have Y caps As shown by osscar above that will cause a problem like this if the PE is not connected (which it should be for safety!)
Once you’ve done this, you can start to debug the 50 Hz noise problem you’ve got.
Can you confirm your speaker 0V are not connected to the chassis?
If the noise only appears when you mechanically connect the heatsinks, which have the amp modules attached.
To me it suggests you have a mechanical issue.
I'd be running a multi meter between the heatsink and each board connection and also the output transistor leads.
Do I also see a pcb stand off that holds the amp boards to a steel plate that goes over the transformers ?
Id also investigate that as a potential circuit connection.
To me it suggests you have a mechanical issue.
I'd be running a multi meter between the heatsink and each board connection and also the output transistor leads.
Do I also see a pcb stand off that holds the amp boards to a steel plate that goes over the transformers ?
Id also investigate that as a potential circuit connection.
Last edited:
I don't think the mains filter is really a problem here, although these have a habit of being screwed to the chassis and establishing a PE connections like that (but obviously not so in this case). Looks like the mains filter capacitors go straight to PE, which should be present if the mains lead and outlet are OK, so their leakage currents are promptly disposed of.
patrick101's 10 ohm resistors are another variation of the "low impedance compared to this, high impedance compared to that" theme. That's another way of breaking the internal ground loop, used by a fair few amplifiers. Using diodes is another, similar trick. You can also combine both, as the ground loop breaker seen above does.
But as I said, I'd rather have the amplifier as a whole more or less floating, with only a weakish PE connection (xx nF, || 100R-1k if you insist). This will require some more care in construction of the primary side be be up to Class II standard (e.g. using a regular double-insulated mains cable internally, heatshrinking connections, keeping mains away from signal stuff etc.) but it's by no means unmanageable. You'll have less trouble with problematic sources this way.
patrick101's 10 ohm resistors are another variation of the "low impedance compared to this, high impedance compared to that" theme. That's another way of breaking the internal ground loop, used by a fair few amplifiers. Using diodes is another, similar trick. You can also combine both, as the ground loop breaker seen above does.
But as I said, I'd rather have the amplifier as a whole more or less floating, with only a weakish PE connection (xx nF, || 100R-1k if you insist). This will require some more care in construction of the primary side be be up to Class II standard (e.g. using a regular double-insulated mains cable internally, heatshrinking connections, keeping mains away from signal stuff etc.) but it's by no means unmanageable. You'll have less trouble with problematic sources this way.
Hi everybody
And thank you for your replys
I've investigated , and try different things … without succes.
- First : I’ve try to remove RFI Filter and replace it with direct connection on the softstart board : Same problem
- Second : I removed the metal plate with softstart just over the two transformers.
Before remove this plate, i notice the center fixation of the right transformer is very hot ! And the two transformers vibrate.
After remove this metal plate : transformers stop vibrated and the centrer fixation is cold now.
But the Noise is still present on the output when PE is connected to chassis
- 3th : Adding Ground Loop Breaker like this :
Great improvment in term of noise, but the ears very near to the test speakers : 50hz is still present …
Thank you for your precious help
And thank you for your replys
I've investigated , and try different things … without succes.
- First : I’ve try to remove RFI Filter and replace it with direct connection on the softstart board : Same problem
- Second : I removed the metal plate with softstart just over the two transformers.
Before remove this plate, i notice the center fixation of the right transformer is very hot ! And the two transformers vibrate.
After remove this metal plate : transformers stop vibrated and the centrer fixation is cold now.
But the Noise is still present on the output when PE is connected to chassis
- 3th : Adding Ground Loop Breaker like this :
Great improvment in term of noise, but the ears very near to the test speakers : 50hz is still present …
Thank you for your precious help
The original photos didn't show it clearly enough.
But obviously (now) you had that top plate bolted down with the transformer bolts.
This would represent a shorted turn to the transformers.
Hence the heat and mechanical buzz in the transformers.
I wouldn't be suprised if the transformers are now damaged.
They might now be generating electrical noise that is getting into the amplifier modules via the wiring.
My suggestions.
Remove that top metal plate. Permanently.
Don't use the amplifier module circuit board mounting hole. It's a small light board and doesn't really need it.
Get a different transformer and test wire it outside the chassis.
If that fixes the speaker noise, mount the two new transformers.
Search the forums here as I've seen others with the same issue where all sorts of wiring changes made no improvement and was finally cured with different brand of transformer.
But obviously (now) you had that top plate bolted down with the transformer bolts.
This would represent a shorted turn to the transformers.
Hence the heat and mechanical buzz in the transformers.
I wouldn't be suprised if the transformers are now damaged.
They might now be generating electrical noise that is getting into the amplifier modules via the wiring.
My suggestions.
Remove that top metal plate. Permanently.
Don't use the amplifier module circuit board mounting hole. It's a small light board and doesn't really need it.
Get a different transformer and test wire it outside the chassis.
If that fixes the speaker noise, mount the two new transformers.
Search the forums here as I've seen others with the same issue where all sorts of wiring changes made no improvement and was finally cured with different brand of transformer.
Looks like it may indeed have been a shorted turn issue. That would explain why separating the the heatsinks from the chassis stopped the noise - the short would be broken in that case.
Normally - if you DO NOT have a shorted turn, when you put the cover on an amplifier, the noise will go down. This it because the radiated field will prefer to flow in the chassis (higher permeability) rather than through space where it will of course induce hum/noise into the wiring. So, I’d try that as a next step.
Also, try rotating the transformers - often you will find a null point where the noise is 8-10 dB lower which is a significant improvement.
Normally - if you DO NOT have a shorted turn, when you put the cover on an amplifier, the noise will go down. This it because the radiated field will prefer to flow in the chassis (higher permeability) rather than through space where it will of course induce hum/noise into the wiring. So, I’d try that as a next step.
Also, try rotating the transformers - often you will find a null point where the noise is 8-10 dB lower which is a significant improvement.
It sounds like a cross channel ground loop...
http://www.updatemydynaco.com/documents/GroundingProblemsRev1p4.pdf
Possibly another problem to overcome - getting an amplifier quiet is like peeling an onion. you switch it on full of anticipation but half an hour later you are in tears
Best test for a cross channel ground loop (CCGL) this is to connect the left and right channels with a cable. With a CCGL it will hum/buzz. When you unplug the cable it will stop. Always fit a hum breaking resistor to stop this kind of problem, locate the input connectors next to each other and pay attention to how you route the wiring inside the amp - the shortest connection is not always the quietest.
Best test for a cross channel ground loop (CCGL) this is to connect the left and right channels with a cable. With a CCGL it will hum/buzz. When you unplug the cable it will stop. Always fit a hum breaking resistor to stop this kind of problem, locate the input connectors next to each other and pay attention to how you route the wiring inside the amp - the shortest connection is not always the quietest.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.