Question about circuit ground elevated from chassis (Marshall JCM 900)

[jmaf]

A technician here in town just called me to ask about a JCM 900 he has on his bench.

After he fixed whatever problem the musician had complained about, he was looking at the circuit and noticed that the circuit ground was above mains earth by 22 ohms, after looking at the schem he noted it wasn't just a resistor, it's actually 2 diodes, a resistor and a capacitor.

Here is the relevant sector http://i.imgur.com/aPHFv.png

So just for kicks, he shorted the circuit ground to chassis, bypassing that circuitry, and the hum was immense. Take out the shunt and it was quiet....

Can any of you please explain the theory behind this?

That schem was extracted from a PDF mentioned in another discussion here, by Enzo, at this address: http://music-electronics-forum.com/t12351/

I understand why giving the heaters a reference reduces hum, but why would elevating the circuit ground relative to mains earth / chassis reduce hum?

Thanks very much in advance.

[tubeswell]

The potential of the return path is separated from the potential of the chassis to get away from potential ground loops. Lots of commercial cicuits have this nowadays. (Look at all the RI Fenders).

[MWJB]

Hi Tubeswell,

When did Fender start doing this...it's not something I have seen, nor is it mentioned on their RI schems?

[jmaf]

Thanks tubeswell.

So the circuit ground reference above the chassis, the mains earth is down below tied to the chassis....

Could we say the elevated positive ground with respect to earth is an extra push to drive any excess ground currents(hum) down mains earth?

[Merlinb]

Could we say the elevated positive ground with respect to earth is an extra push to drive any excess ground currents(hum) down mains earth?

Nice idea, but 'fraid not.

The circuit ground is at 0V same as the chassis earth. However, the small resistance just reduces any ground current which might try to flow from one appliance, down the interlinking cable, through circuit ground, to earth. By reducing this ground current you therefore reduce any noise voltage (hum) that it produces too.

[tubeswell]

Hi Tubeswell,

When did Fender start doing this...it's not something I have seen, nor is it mentioned on their RI schems?

Okay well so far the only evidence I can produce is that they are in the 63 reverb unit. I'm sure I've seen them in other Fender schematics. (Naturally I'm lookin fer others as we speak, but in case I don't find any, I concede on grounds of my own laziness )

[pdf64]

I've noticed that the RI Bassman, for one, has that arrangement between it's input socket and chassis, though the main power supply ground is directly tied to the chassis.
As merlin advises, it will be to mitigate ground currents within the amp chassis.
The reverb unit, as a line powered 'effect' before the amp input, will be particularly prone to ground loop problems (eg via instrument cable and power cables), so more mitigation may be beneficial.

[jmaf]

Thanks, Merlin. I don't have much of a choice here, I've always soldered a ground bus bar to my chassis' so I've always had my ground = chassis potential. I've taken care to use shielded leads, proper dress, etc, so hum has never been a real problem, but I wonder how much quieter it'd be if I had known this before.

[bob p]

I was going to make the comment that Tubeswell made about the RI reverb unit. I noticed the design in that unit when I was thinking about building one from the vintage schematic, and looked at the new schematic for a better parts list.

The idea of elevating chassis ground above earth potential is very common -- you'll see it in any audio gear that has a ground lift switch on the back. Some people think that a ground lift is accomplished by disconnecting the chassis from the ground lug on the AC cord. that's not the case. you're NEVER supposed to do that. the ground lift disconnects the connection between the circuit ground plane and the chassis which remains at earth potential. the metal chassis has to remain earthed for electrical safety.