Ok don't laugh as I didn't really know what to show and how to show it on this picture but the green wires are earths, the stripboard has a few strips together acting as GND and underneath strips of wire going along the GND plane for extra peace of mind, as i was worried the strips may not be good enough.
Try to get all your grounds to meet close to a single point if you can. The principle you have to follow is that every wire carries a current - in either direction - and thus there is actually nothing that physically behaves as 'ground'. If you need it to behave like a single point, you have to implement it as a single point otherwise it's a loop.
My observation from your diagram is that the positive supply board "ground" pole does not have a close enough connection to either amplifier. You would want to get both the supply common connections to one point and take it from there for both boards.
Also, I am not familiar with the PCBs you are using for the build - but you may want to use this power supply ground to get the speaker negative, instead of the PCB. Also from the same point, a return connection to signal ground on the amplifier PCB. If bypass caps are installed it is important that any ground connection to the PCB be of thicker and lower resistance wire than the signal ground connection else the decoupling currents will flow through the signal ground. It is probably not OK to share a single wire for both purposes.
I notice you're using a signal selector (maybe a volume?) on the actual build. The ground connection of that component is not trivial either. I normally isolate noisy sources with a small 0.1 ohm resistor on the ground lead, if needed (typically not needed).
Frankly - the chosen method for your build, of a single supply in one chassis with two amplifiers, is probably worst case scenario for any ground problems. But there are plenty people who've battled the issue and come out on top. I have a gaming amp which has 7 amplifier sections of differing ancestry in one chassis, and 5 of them run off one supply, without even the slightest trace of hum - because I knew it would happen if I wasn't careful. Redo the grounding and you should be fine.
My observation from your diagram is that the positive supply board "ground" pole does not have a close enough connection to either amplifier. You would want to get both the supply common connections to one point and take it from there for both boards.
Also, I am not familiar with the PCBs you are using for the build - but you may want to use this power supply ground to get the speaker negative, instead of the PCB. Also from the same point, a return connection to signal ground on the amplifier PCB. If bypass caps are installed it is important that any ground connection to the PCB be of thicker and lower resistance wire than the signal ground connection else the decoupling currents will flow through the signal ground. It is probably not OK to share a single wire for both purposes.
I notice you're using a signal selector (maybe a volume?) on the actual build. The ground connection of that component is not trivial either. I normally isolate noisy sources with a small 0.1 ohm resistor on the ground lead, if needed (typically not needed).
Frankly - the chosen method for your build, of a single supply in one chassis with two amplifiers, is probably worst case scenario for any ground problems. But there are plenty people who've battled the issue and come out on top. I have a gaming amp which has 7 amplifier sections of differing ancestry in one chassis, and 5 of them run off one supply, without even the slightest trace of hum - because I knew it would happen if I wasn't careful. Redo the grounding and you should be fine.
post2050 shows the inputs at two opposite ends of the back plate.
These connect in a big U shaped wiring arrangement.
Now add the interconnect and the Source.
You will find that you have a very big loop that includes the Source.
When you disconnect one interconnect you are breaking that very big loop.
1.) you must minimise the current in that big loop
2.) you must separate the Signal Ground from the Power Ground for each amplifier in a common Chassis.
3.) you must create two completely separate PSUs, OR adopt a single Main Audio Ground
These connect in a big U shaped wiring arrangement.
Now add the interconnect and the Source.
You will find that you have a very big loop that includes the Source.
When you disconnect one interconnect you are breaking that very big loop.
1.) you must minimise the current in that big loop
2.) you must separate the Signal Ground from the Power Ground for each amplifier in a common Chassis.
3.) you must create two completely separate PSUs, OR adopt a single Main Audio Ground
The problems when the Signal wires of left and right touch and not the earth of the RCA
there's 137mv of AC from the power supplys +/-
Between the AC IEC ground in and the output of the rectifier and thermistor there's about 7ohm. i presume this is normal.
electrically all ground connections seem spot on checked with the DMM no measurable extra resistance.
With Transformer lifted off chassis, no difference
speakers earth taken from various locations has no effect, taken from the chassis there is less noise (probably due to the 7ohm resistance!)
I have GND and Link on the official Diyaudio boards Jumpered
There is no signal selector or volume on at the moment for testing purposes.
sangram:
"Frankly - the chosen method for your build, of a single supply in one chassis with two amplifiers, is probably worst case scenario for any ground problems."
Unfortunately i settled for stripboard PSU that i made myself but Im now thinking of ordering a couple of PCB's
CRC Power Supply PCB for Class A Power Amplifiers | eBay
Atleast the ground plane is solid and i was thinking of using 2 of them one for each channel, but may end up with the same problem if its not that.
for now im going to try get all grounds at a single point.
Andrew:
" 2.) you must separate the Signal Ground from the Power Ground for each amplifier in a common Chassis."
Dont both signal grounds end up at ground anyway ? Or i just don't understand how signal ground can be isolated from power ground as electrically both seem the same.
I may create two seperate PSU but not have two seperate transformers., im going to start rewiring this amp now if fail then buy those PCB's
Cheers
there's 137mv of AC from the power supplys +/-
Between the AC IEC ground in and the output of the rectifier and thermistor there's about 7ohm. i presume this is normal.
electrically all ground connections seem spot on checked with the DMM no measurable extra resistance.
With Transformer lifted off chassis, no difference
speakers earth taken from various locations has no effect, taken from the chassis there is less noise (probably due to the 7ohm resistance!)
I have GND and Link on the official Diyaudio boards Jumpered
There is no signal selector or volume on at the moment for testing purposes.
sangram:
"Frankly - the chosen method for your build, of a single supply in one chassis with two amplifiers, is probably worst case scenario for any ground problems."
Unfortunately i settled for stripboard PSU that i made myself but Im now thinking of ordering a couple of PCB's
CRC Power Supply PCB for Class A Power Amplifiers | eBay
Atleast the ground plane is solid and i was thinking of using 2 of them one for each channel, but may end up with the same problem if its not that.
for now im going to try get all grounds at a single point.
Andrew:
" 2.) you must separate the Signal Ground from the Power Ground for each amplifier in a common Chassis."
Dont both signal grounds end up at ground anyway ? Or i just don't understand how signal ground can be isolated from power ground as electrically both seem the same.
I may create two seperate PSU but not have two seperate transformers., im going to start rewiring this amp now if fail then buy those PCB's
Cheers
Last edited:
If the Signal Ground is separate from the Power Ground then the two signal Grounds becomes separate from each other when the Power Grounds are connected. If you have ONE PSU with a Centre tap then That Centre Tap Must be connected to BOTH Power Grounds.
Don't buy PSU PCBs yet. The problem is in the amplifiers and interconnects.
Thanks people 
Reduced the hum a little more, solid core wire seems to reduce the hum more the wire you get in the house cavity. thick solid core
I have
60mv bias and heatsinks over 50 degrees
and the metal washers on the mosfets are over 60 degrees, seems a bit high for only 0.06 bias
Reduced the hum a little more, solid core wire seems to reduce the hum more the wire you get in the house cavity. thick solid core
I have
60mv bias and heatsinks over 50 degrees
and the metal washers on the mosfets are over 60 degrees, seems a bit high for only 0.06 bias
Last edited:
all earths now going to a central point, same hum, also took the speaker from that point and tried the RCA from that point.
personally i think my veroboard wiring and layout is flawed, having one board for -v and one for +V is surely not a good idea, at the time I thought it was good but the tracks cant be as good as proper PCB.
This projects for a friend but I want it out of the way and done after so many hours of working on it. the sound tho the sound is so lovely I cant believe it, i think it possibly sounds better than the original F5 I built.
Pitty the diyaudio - Universal power supply boards are not selling right now as I like the look of them plus they have the section for the diodes too!
personally i think my veroboard wiring and layout is flawed, having one board for -v and one for +V is surely not a good idea, at the time I thought it was good but the tracks cant be as good as proper PCB.
This projects for a friend but I want it out of the way and done after so many hours of working on it. the sound tho the sound is so lovely I cant believe it, i think it possibly sounds better than the original F5 I built.
Pitty the diyaudio - Universal power supply boards are not selling right now as I like the look of them plus they have the section for the diodes too!
