The reason I say that is that looking at #5 some of the flux of the mains transformer will induce a small voltage across the power supply and the black wires. This will cause the two input circuits to have a small voltage between their grounds. The coax inputs being further away will have less induced voltage. Basically you have a magnetic circuit with the black wires and the coax. Using one input tube eliminates this issue as all is in one place. Do you have a trace of the hum its a good indication of the nature of the issue?
If I was going to make a simple mod wrt #5 remove the black wire going to the cathode resistor/220uF cap and grid resistor from the power supply, and reroute instead over the coax cable to the same place on the right turret board.
With 50hz power that explains 100hz hum. How it only effects one channel is a probably a component too close to the PS section would be my guess. Something is picking up the 100hz noise and the PS section is where the 100hz is being produced. Do try the virtual ground and see if it helps. I would use at least 2 watt maybe even 3 watt to start with. You can then measure voltage drop across one of the resistors to determine wattage actually needed using ohms law.
Your noise (hum) comes form the bottom inductor, cause is coupling the ripple that is meant to filter directly to the left output transfomer via coupling, the only way to reduce it is to fisically move the inductor away from the OPT or to play with its magnetic path orientation (rotate the inductor in its place) to reduce magnetic coupling, another alternative way is using a RLC to filter ripple, but will cause heat and will reduce the output power if not designed properly, also can you can use a solid state regulator plastered around to a chassis bolt if you cant live with the noise of the supply.
If you will you can test and play around for the best result.
But when the left 12AX7 is removed the hum goes away. If it were choke to OPT the hum would stay, right?
The hum/noise may introduce itself to the loop cause the partial feedback (output anode to 12ax7 resistor) but it will be surely noted with more loop gain, the 12ax7 has the most gain in the amplifier itself, so the hum may not disappear, but just be almost 50 times less in magnitude.
1. Grounding one end of a 6.3V secondary . . . Puts 6.3VAC at one end of the filament.
Wrong way to do it.
Using a Pseudo Center Tap (two 100 Ohm resistors to ground, and the other ends to the 6.3V wires) . . .
Puts 3.15VAC at each end of the filament (-6dB versus grounding one end of the secondary).
That probably is not the cause of your problem.
2. But it is beginning to sound like you have an un-controlled input tube circuit ground loop, or other un-controlled ground loop.
Often, a bad solder connection will do this, but if not, then . . .
Are both the Left and Right channel RCA input connectors insulated from the chassis, is one shorted to the chassis, and then are those RCA connectors return conections connected to the bottom of the 12AX7 self bias networks?
Are the input tube's grid resistors, Rg returned to that same point?
Are the return wire of the volume pot returned to the bottom of the 12AX7 self bias networks.
After you have created this correctly controlled Local ground loop, then you should connect the bottom of those self bias networks to the amplifier central ground point, which should also should be the connection to the chassis.
B+
There should be a short local loop from the anodes of the 1N1007 to the negative of the first filter cap negative. Then use a second wire from the first filter cap negative to the second filter cap negative. Then use a third wire from the second filter cap negative to the central ground.
Ground loops need to be controlled, it is hard work, but the rewards are less hum and hash.
Wrong way to do it.
Using a Pseudo Center Tap (two 100 Ohm resistors to ground, and the other ends to the 6.3V wires) . . .
Puts 3.15VAC at each end of the filament (-6dB versus grounding one end of the secondary).
That probably is not the cause of your problem.
2. But it is beginning to sound like you have an un-controlled input tube circuit ground loop, or other un-controlled ground loop.
Often, a bad solder connection will do this, but if not, then . . .
Are both the Left and Right channel RCA input connectors insulated from the chassis, is one shorted to the chassis, and then are those RCA connectors return conections connected to the bottom of the 12AX7 self bias networks?
Are the input tube's grid resistors, Rg returned to that same point?
Are the return wire of the volume pot returned to the bottom of the 12AX7 self bias networks.
After you have created this correctly controlled Local ground loop, then you should connect the bottom of those self bias networks to the amplifier central ground point, which should also should be the connection to the chassis.
B+
There should be a short local loop from the anodes of the 1N1007 to the negative of the first filter cap negative. Then use a second wire from the first filter cap negative to the second filter cap negative. Then use a third wire from the second filter cap negative to the central ground.
Ground loops need to be controlled, it is hard work, but the rewards are less hum and hash.
Last edited:
A side note: Due to the schematics, only one of the triodes in each 12AX7 is used. Is this correct? Then I suggest one of the following: Either disconnect the heater supply of the unused triodes, which would save about 2 watts of power consumption. Or connect both triodes of each tube in parallel and decrease the cathode and plate resistors to half of their actual values. This will reduce noise by 3 dB. Or use one 12AX7 for both channels and omit the 2nd tube.
Best regards!
Best regards!
So I modified the ground plane as you told me (I applied the same treatment for EL34)
, and it works. The noise on the left channel has disappeared...
I think the ground plane of this amplifier is really bad...
What ground plan would be ideal?
Modification like this ____
Last edited:
If it works use that. The problem I think is the ground cables were picking up some of the magnetic field of the mains transformer inducing hum between the grid and cathode of the first stage. The EL34 will make little difference as the signal is much bigger. It is a problem I had. I think anything else needs you to just use one 12ax7 and both halves. Having good photos was great. Its not a good arrangement to have the left on on side and the right on the other for the input stage. It opens up a big magnetic loop around the ground and the coax inputs.
Last edited: