Quiet amp hums with signal source or input directly tied to star earth??

[bigwill]

I am troubleshooting an amp for someone and noticed that although it's pretty quiet (normal hiss with hum only just audible with your ear on the dustcap) with no input, it hums if you connect a source, even a floating one such as an iPod, or even if you directly connect it to the amp's earth star!

The input valve's grid is tied to ground via 100k, and goes through the PCB to the star earth directly on its own trace.

I made the PCB for this amp and it was very quiet even during testing with a nest of clip leads.

The wiring all looks sane and I can't see anything that would cause a loop, and surely grounding the grid via a clip lead wouldn't cause a loop through the high impedance grid circuit.

Can anyone shed some light on what might be happening? I've never encountered this with amps I've wired myself, but this was done by a friend and he may have made a mistake...

[Serge66]

Hi,
Is the power supply clean? How much ripple do you have?

It could also come from a magnetic coupling (output transformer too close to the mains transformer or smoothing choke).

If it hums when you connect a source, it could well be a loop somewhere.

Good luck.

Cheers,

Serge

[bigwill]

Originally Posted by Serge66

Hi,
Is the power supply clean? How much ripple do you have?

It could also come from a magnetic coupling (output transformer too close to the mains transformer or smoothing choke).

If it hums when you connect a source, it could well be a loop somewhere.

Good luck.

Cheers,

Serge

The power supply is fairly clean - it's just a CLC network fed with a GZ34 though. The amp is very quiet when the input is left floating, but as soon as you connect it to anything - there's a hum!

What I can't quite get my head around is the fact that if you connect a clip lead to the exact point the 100K grid resistor is earthed, poke a 33k resistor into the other clip and touch it to the input, it hums! How can you have a loop with 133k of resistance?? This seems to defy logic!

[bigwill]

It seems even if I short the input terminal with a 33k resistor, it hums.

I cannot figure this out. Why are my rats nests so much quieter than this fairly neat amp??

[SGregory]

Is the Hum 50 or 100Hz?

[bobodioulasso]

Have you tried one channel only hooked to the source. If hum stops, it's a ground loop.

[waltube]

This is a problem with a ground loop. The loop is closed with a source connected to the amp due to a signal cables
Try to put a 10 ohm 1 w resistor in series from the circuit ground to the chassis ground of one (or both) channel.
I assume that the input pin jack are isolated from chassis

Ciao

Walter

[Arnulf]

Does this mystery amp of yours happen to use active hum cancellation ? If it does, this would explain the apparent absence of hum (incorrectly set null) with high-Z input and apparent presence with overcompensation using low-Z input.

If it doesn't, short input tube's grid to the point just below its cathode resistor. Does hum still appear or not ?

[bigwill]

Originally Posted by SGregory

Is the Hum 50 or 100Hz?

It's non-raspy 50Hz, although it doesn't sound like a pure sine.

Originally Posted by bobodioulasso

Have you tried one channel only hooked to the source. If hum stops, it's a ground loop.

No. It's a monoblock

Originally Posted by waltube

I assume that the input pin jack are isolated from chassis

Yep, with nylon washers that came with the phono sockets

Originally Posted by Arnulf

Does this mystery amp of yours happen to use active hum cancellation ?

Nope

Originally Posted by Arnulf

If it doesn't, short input tube's grid to the point just below its cathode resistor. Does hum still appear or not ?

Hmm, I haven't tried this, although I have tried bypassing the 100k grid resistor to earth and the hum appeared. It did not appear with a 33k resistor carefully placed in parallel by hand though...

[Wavebourn]

Looks like rectifier's current goes to the ground. I would suggest that the PS has own ground bus from rectifier to the last filter cap, and the cap is connected to the star earth (not a center tab of a power tranny!)

Also, both power amp and a signal source may have some power filters that create a ground loop. Screw them. They are needed only to protect power lines from ripples if your amp has SMPS that generate them. However, if you can't do that line level transformers may help, but it is an expensive option.