Why reduce input valve heater voltage?

[Martin_Calva]

I have an Advance VM77A valve voltmeter. The first amplifier valve, V1, is an EF95. There is a 2.7 ohm resistor in series with each of its heater connections. I calculate this reduces its heater voltage by about 1 volt from the 6.3 volts fed to the heaters of the other valves.

The manual states "V1 is best selected for minimum hum and noise" although it does not say how this should be done. It does not mention the two resistors in series with V1's heater nor explain why there are there.

What is the purpose of reducing the heater voltage of the input valve? I don't recall ever having seen this done elsewhere.

[winston_1]

I’ve come across this technique before on valve voltmeters. In my case the valve was an EF37A and the heater supply was fed from a higher voltage via a 280mA barretta. It is done to give greater stability.

[Chris55000]

This was commonly done in instrument circuits as lowering the temperature of the cathode reduces the tendency of the valve to pass "contact current", produced by electrons in the stream within the valve landing on the control grid wires.

When two electrodes of a valve are in contact by an electron stream from the cathode, with a very high external impedance between them, a negative "contact potential" appears between them which, although suitable for grid–bias in low–level a.f. stages, is liable to cause inaccuracies in measuring circuits!

Therefore, under–running the heater reduces the inaccuracy from "contact potential" and effectively substantially increases the input impedance of the valve.

There are special valves called "electrometer" type valves that are specially manufactured to give virtually zero "contact potential" negative grid current, therefore providing an exceptionally high input impedance, but reducing heater current by about 15–20% of nominal achieves almost as good a result with a standard commercially made valve!

M.G. Scroggie explained this on his January 1952 article in "Wireless World" for the "High Accuracy Valve Voltmeter", from which the VM77 and the Mullard E7555 circuits are derived!

Chris Williams

[Chris55000]

PS!

Regarding the recommendation in your VM77A manual to select the input valve for minimum hum and noise, the simplest method of achieving this is to switch the instrument to the lowest d.c. voltage range, leaving the input terminals open circuit, there will be a small deflection of the meter due to hum and noise picked up on the input stage grid circuit, self–induced hum and noise, etc. – then all you need to do is try different valves in V1 position in turn, selecting the one that gives the lowest overall meter readings!

Chris Williams

[Leon Crampin]

Grundig used this technique together with a humdinger adjuster in their 50s tape recorders. The EF86 first stage had a dedicated 6V heater supply.

Leon.

[G6Tanuki]

Running the heaters at a lower voltage is also used in the noise-limiter circuits of some communications-receivers; I guess it's something to do with the 'contact potential' here too.

[TonyDuell]

I've heard that the ME1400 electrometer valve, used in high-impedance voltmeters, etc, is essentially a selected EF37A but specified to run at 4.5V on the heater.

[Martin_Calva]

My thanks to all for the interesting and helpful information.

As the VM77 measures ac voltages only, I am not sure why contact potential in V1 should be a concern. Maybe the fact that the grid resistor has a value of 10M has something to do with it.

Martin

[John10b]

Interesting read, I’d forgotten about this, thanks.
John

[Robert Gribnau]

See page 5 of this datasheet for the E80F. As an electrometer pentode it runs with 4.5 V on the filament instead of the normal 6.3 V.

https://frank.pocnet.net/sheets/009/e/E80F.pdf

[turretslug]

Another reason I'd heard quoted was that there are often (as can be seen by inspection) loops of heater filament protruding beyond the end of the cathode cylinder and these can emit if running above a certain temperature. Obviously, the cathode is intentionally coated with a low work function material, so there's a point where the exposed heater emission is sufficiently reduced but the cathode emission remains adequate as the heater voltage (i.e consequently power) is reduced. One well-known case in point is with later production AR88s having a 5 ohm resistor in series with the 6H6 noise limiter's heater to cut AF hum with it switched in. I found by experimentation with full heater volts applied that 6H6 hum susceptibility varied a lot between samples, but that glass 6H6GTs were invariably noticeably better than steel can 6H6s.

[Maarten]

Lowering the heater voltage might cause less inherent noise, though I forgot the exact mechanism.

[tritone]

That is an interesting question Martin and one that I have been wondering about myself for a long time concerning a piece of audio equipment. Why the lower heater voltage? Maybe this thread will also answer my curiosity as well. Thank you.

So, from my electronic ventures, I have serviced and repaired these high gain guitar preamps called an ADA MP-1.

The voltage to the 12AX7 valve heaters in these preamps is 8/9 Vdc as opposed to the normal 12.6 Vdc.
I have never enquired to why this has been designed this way and like yourself Martin ‘What is the purpose of reducing the heater voltage of the valves in this situation? I don't recall ever having seen this done before either.

http://adadepot.com/mods/ada_mp1a/ADAMP1tub.jpg


Thanks for your input forum.

[daviddeakin]

Originally Posted by Maarten

Lowering the heater voltage might cause less inherent noise, though I forgot the exact mechanism.

At very small anode currents it can reduce noise, as well as buying improved lifetime and lower grid current. Quite a few advantages under the right circumstances.

[Maarten]

Very interesting indeed!