Connection from input jack to V1

[Malcolm Irving]

Often the connection from input jack to V1 is via a screened cable, with the screen connected at one end only. Signal return current from the cathode goes via the cathode resistor through some ground connection back to the jack.

Connecting the screen only at one end prevents a ground loop, and gives screening against electric fields, but provides no screening against magnetic fields.
Interference from nearby heater wiring can be magnetic (since heater wiring is relatively low voltage but high current).

To get magnetic screening, in addition to electric, the outer conductor of the screened cable must carry the signal return current.
So how about the following:

Grid leak 1M at the input jack as usual. Outer conductor of screened cable connected to jack ‘sleeve’. Near V1, the outer conductor is soldered to a wire which continues alongside the grid stopper, does a U-turn near the V1 pins, travels close alongside the wire back from the cathode and alongside the cathode resistor eventually connecting to the ‘ground end’ of the cathode resistor.
The single ground connection to the chassis can be at the jack or at the ‘ground end’ of the cathode resistor.
In the above scheme the signal return current goes via the outer conductor and also the overall ‘loop area’ of signal and return is minimised to reduce magnetic interference. No ground loop is formed.
If there is a cathode bypass cap, the wire should go close to that, rather than close to the cathode resistor, as the AC signal passes through the cap.

I’ve tried it and it works, but I can’t present any practical evidence whether it works better than the usual scheme or not. I think in theory it should be better!

[pdf64]

The (hypothetical) benefit being that the magnetic interference should then act both on signal and its 0V conductors, becoming effectively common mode and cancel out?
If so, it seems a bit dodgy to me; if signal and screen conductors were fed into either side of differential inputs then I can see it would work, but as it's fed into a single ended pre-amp, then I'm a bit sceptical that it would be improve hum performance

[Chuck H]

I almost finished before a vein popped in my head.

[Malcolm Irving]

The (hypothetical) benefit being that the magnetic interference should then act both on signal and its 0V conductors, becoming effectively common mode and cancel out?
If so, it seems a bit dodgy to me; if signal and screen conductors were fed into either side of differential inputs then I can see it would work, but as it's fed into a single ended pre-amp, then I'm a bit sceptical that it would be improve hum performance

With magnetic interference, when you have any loop, the voltage induced into the loop is proportional to the rate of change of flux linking through that loop. So to minimize that you need to keep the area enclosed by the loop as small as possible. If the area can go down to zero - there would be no magnetically induced voltage in that loop.

Inputs to grid and cathode are 'differential', i.e. +1V onto the grid and +1V simultaneously onto the cathode has zero effect on the output of the tube.

[Malcolm Irving]

I almost finished before a vein popped in my head.

Sorry Chuck. Send me a bill for a Paracetamol.

[big_teee]

It's too early here, I caved after second paragraph.

[Leo_Gnardo]

It's too early here, I caved after second paragraph.

It's a humbucking coil right at the grid input, piece o' cake. One more coffee & we'll both be up to speed.

[catalin gramada]

Hi. some consideration thinking about...
First-if you have magnetic issues interferences you need a magnetic shield means iron more than copper instead. I doubt there are iron braided signal wires, but who knows...second- a iron wire exposed to power magnetic field tend to develop some called eddie parasitic currents which can interfere with signal, third- this sort of screen should be tied directly to star ground and not into decoupling point that can complicate the wiring plan
..still puzzling
magnetic issues usualy are solved more by orientation than screening. shielding becomes effective using some dedicated feromagnetic materials like permaloys. That.s simple task if have to solve some components issues( trannies ,coils) into chassis, but if you have a gig near a district distribution station transformer I doubt can help

[Malcolm Irving]

A good reference in this area is ‘Noise Reduction Techniques in Electronic Systems’ by Henry W. Ott.
On page 40 of the 1976 edition, for example:
“If a non-magnetic shield placed around the conductor causes the current to return over a path that encloses a smaller area, then some protection against magnetic fields will have been provided by the shield. This protection, however, is due to the reduced loop area and not to any magnetic shielding properties of the shield.”

[Enzo]

Dose makes the poison. I have not seen reference to amounts here. Electronics is full of phenomena, but many have such minuscule presence in our circuits as to be ignored. Skin effect, for example. So just how much magnetic signal is induced into the wire in question, that shield with one end grounded and maybe 6 inches long tops? Heater wires may have a tiny magnetic field around them, but from several inches away? Good construction practice is to twist heater wire pairs, and also wherever possible, and to the extent possible, have them cross signal lines at a 90 degree angle.

SO are we jumping through hoops to shave off a couple pico volts of hum?

[catalin gramada]

I doubt a magnetic field can perturbate input signal, except maybe you have a input transformer coupled...

[Mike Sulzer]

Sorry Chuck. Send me a bill for a Paracetamol.

So using dc on the filaments can get rid of hum and eliminate a headache.

[Malcolm Irving]

I doubt a magnetic field can perturbate input signal, except maybe you have a input transformer coupled...

Yes, that's the bit I'm not sure about. I think it's difficult to assess that either experimentally or by computer simulation. There are fairly heavy currents in the heater wires, but twisting them minimises the magnetic interference they produce.

[Malcolm Irving]

Dose makes the poison. I have not seen reference to amounts here. Electronics is full of phenomena, but many have such minuscule presence in our circuits as to be ignored. Skin effect, for example. So just how much magnetic signal is induced into the wire in question, that shield with one end grounded and maybe 6 inches long tops? Heater wires may have a tiny magnetic field around them, but from several inches away? Good construction practice is to twist heater wire pairs, and also wherever possible, and to the extent possible, have them cross signal lines at a 90 degree angle.

SO are we jumping through hoops to shave off a couple pico volts of hum?

Yes, I can't quantify the 'problem', so perhaps its a 'solution to a non-problem'. I'll see if I can think of a way to quantify it.

[catalin gramada]

a quick A-B-A test with a dc source can clarify. but ussualy ac heaters don.t put problems. The last traces of noise comes from couplings between transformers or into transformers windings. That suposed the amp circuit have not bugs inside