Hum on one channel of hybrid amp

[Breezio]

The resistance between grounds should be basically zero (maybe a few tenths of an ohm just because of oxidation on metal surfaces and probes). Check resistance along the way. Does it increase across a particular solder joint?

When the amp is on I measure 13 ohm and when it's off the then values (for every point) go back to the tenths of ohm.
What is that supposed to mean?

 

[scott2000]

Any chance you can make a drawing showing how you implemented your regulator into this psu design???

 

[Breezio]

The regulator sits between the B+ coming from the psu and the connectors L/R which power the tube modules.
I just interrupted the B+ supply and put the rgulator inbetween.
I will try to provide a picture.

However, the hum problem is present also when I take the regulator out and revert back to the original supply

 

[Bo Deadly]

When the amp is on I measure 13 ohm and when it's off the then values (for every point) go back to the tenths of ohm.
What is that supposed to mean?

That's actually normal. When you take resistance measurements, you must do it with the power off. Any current between the points being measured will throw off the measurement.

So if it's tenths of an ohm with the power off, then that suggests that the ground connection is fine.

However, you said previously that the supply comes through a main board. Does ground also go through that connector? The general rule is that supply currents should be matched by return currents. Meaning if there's a supply wire going to another board, the ground wire should return in close physical proximity so that electromagnetic fields generated by said currents cancel each other out. If there is a ground wire running back to the supply separately, that would be an error in the design because it would make a physical loop that can emit noise.

 

[Breezio]

That's actually normal. When you take resistance measurements, you must do it with the power off. Any current between the points being measured will throw off the measurement.

So if it's tenths of an ohm with the power off, then that suggests that the ground connection is fine.

However, you said previously that the supply comes through a main board. Does ground also go through that connector? The general rule is that supply currents should be matched by return currents. Meaning if there's a supply wire going to another board, the ground wire should return in close physical proximity so that electromagnetic fields generated by said currents cancel each other out. If there is a ground wire running back to the supply separately, that would be an error in the design because it would make a physical loop that can emit noise.

YES!!!!!!

I was not aware of this golden rule - thanks a lot for bringing it up!
I reviewed the regulator mod and added ground wires following the same path of the unregulated B+ to the additional board and back. Guess what? The hum is finally gone!

Squarewave, thanks again for your invaluable help.

Not only I got rid of a very annoying problem but, most important, I learn something very useful for the future.

 

[Bo Deadly]

Ah, so I finally got something wr-wr-right. Glad to help. Yes, that is a very important principle that pervades many areas of EE design.