Selmer Treble N Bass Mk III - Ground bus question.

[CrimsonGhost]

Just picked up a Selmer amp and I noticed the ground bus terminates on the centre of the PI valve socket, but the tab on the valve socket that it shares a relationship with is empty and instead is uses a ground connection from the transformer, which is daisy chained along the 2 EL34's, the PI and then on to the three pre amp valves (the brown wire).

Any reason for this ground to terminate here?

[BrianS]

I can't tell you the exact reason why they chose to ground at this point, but I think it's fair to say that if it is working, don't change it! In my direct experience with repairing wacky old tube amps, changing anything with layouts or grounding schemes can have unintended, negative consequences.

[SEL49]

...I noticed the ground bus terminates on the centre of the PI valve socket... Any reason for this ground to terminate here?

That connection is mainly used just to support the end of the ground buss. That center pin is very useful in RF circuits but really has no application at audio frequencies.

[CrimsonGhost]

Just adding a picture of the amp with a top down view for anyone interested, you can get a better idea where the ground comes from.

[CrimsonGhost]

Im just correcting a false statement above but also looking for an explanation as to what's happening.

In my above post I wanted to know why the ground bus terminated at the valve socket and why the corresponding pin was unused. I wrongly stated that the brown wire was ground from the transformer. I had done some continuity checks with chassis ground while the amp was turned off and it beeped anywhere I probed it.

I pull the amp back out today because the bass channel doesn't work and I spend more time looking at it, I realised the brown wire I mention in my above post must be the heater circuit, so I fire it up and I measure 3.2vac on the brown wire, But when I turn the amp off the brown wire has continuity with chassis ground again.

Can anyone explain why that happens?

[tubeswell]

I measure 3.2vac on the brown wire, But when I turn the amp off the brown wire has continuity with chassis ground again.

I’m seeing several brown wires in that gutshot. Can you markup the picture to show where you had the VAC probes?

[CrimsonGhost]

This one here.

Edit: to further clarify, black meter probe was connected to chassis ground and the red probe was on the brown wire you see in the photo.

[pdf64]

... I measure 3.2vac on the brown wire, But when I turn the amp off the brown wire has continuity with chassis ground again.

Can anyone explain why that happens?

'Continuity' is an indeterminate parameter, anything up to a few hundred ohms might 'beep' the meter, it's much better to report actual resistance readings rather than that.
Whatever, my guess is that the mains transformer heater winding has a centre tap that's connected to the chassis; or there's a pair of resistors eg 100R from each heater leg to the chassis; can you confirm?
If so the yellow heater wire should have the exact same resistance to the chassis.

[CrimsonGhost]

When I measure from chassis to the brown or yellow wire using the lowest ohms setting(600ohm) on my meter I get 001ohms on both wires(one probe on chassis,one on either the brown or yellow wire). When I probed the other wires on the valve socket all show OL on the meter.

[HotBluePlates]

... I fire it up and I measure 3.2vac on the brown wire, But when I turn the amp off the brown wire has continuity with chassis ground again.

Can anyone explain why that happens?

Normal.

The heater winding has a center-tap that is connected to Chassis/Ground.  The heater winding is a few-turns of fat-wire, and has low resistance from either end to the center-tap.  Therefore, your measurements are normal.

[Merlin]

Just picked up a Selmer amp and I noticed the ground bus terminates on the centre of the PI valve socket, but the tab on the valve socket that it shares a relationship with is empty and instead is uses a ground connection from the transformer, which is daisy chained along the 2 EL34's, the PI and then on to the three pre amp valves (the brown wire).
Any reason for this ground to terminate here?

I don't really understand your question. The ground bus is connected to chassis at one end, the other end is soldered to the valve socket spigot as a convenient tie point, so the ground bus is suspended in the air. Why would you expect the valve socket tab to be connected to anything else?

[CrimsonGhost]

Thanks very much to everyone who helped.

I never suggested it should be connected anywhere in particular so Im not quite sure what your talking about?

I only asked why it terminated at the centre of the valve socket and why the corresponding tab that it shares a relationship with was empty.

I didn't suggest it should or should not be connected anywhere else, Im just asking a simple question. What's your problem?

[AlNewman]

There's several ground references in that amp.  They all go to the same place.

[HotBluePlates]

I only asked why it terminated at the centre of the valve socket and why the corresponding tab that it shares a relationship with was empty.

... That center pin is very useful in RF circuits but really has no application at audio frequencies.

SEL49 gave you the answer, but more explanation might have helped.

A "capacitor" is just a couple of conductors separated by an insulator.  So if you look at the socket-pins, you might observe there is capacitance between the different pins.

You can find the formula that explains how the parts of the capacitor influence capacitance, but suffice to say the socket's pins have a very small capacitance between them.  Too small really to imagine a "phantom cap" connecting the pins at audio frequencies, but at radio frequencies... Here, it turns out the capacitance is big enough to matter for RF circuits.

How to combat unwanted phantom-capacitor connections at RF?  That center-spigot splits the distance between pins across the socket from each other in half.  2 capacitors in-series create an even smaller total capacitance, which would move the affected frequencies higher.

Except socket-manufacturers gave another, better option:  Ground the spigot, either directly or by using a connected tab as a wire tie-point.  Although we could envision 2 cross-socket "phantom caps" in-series, we can instead imagine a cap-to-ground from "pin to center-spigot" that prevents signal on one pin from being coupled into another pin where it doesn't belong.

Tube manufacturers eventually considered the end-user's possible use of sockets with these center-spigots, and also possible coupling among tube-pins when they design the tube and decide what-circuit-element will be brought out to what-pin, and what that implies for unwanted coupling and end-user circuit-wiring.