Grid Bias Voltage

[SurreyNick]

I have a few questions regarding choosing and applying the correct grid bias voltage to the three-valve battery powered TRF I am building. The receiver is based upon a theoretical circuit by FJ Camm. I have read and re-read his book and unless I am missing something obvious he doesn’t explain what I need to know. I am hoping someone here can point me in the right direction.

Attached are two pictures. One is a partial snapshot of the circuit showing the GB arrangement. The other is the valve characteristics of the Triotron SP2 which is the valve I am using in the two amplifying stages. Maximum anode voltage is 150v and it says GB should be -12v at 100v HT (by the way should it read -28v at 150v? I think there’s a misprint)

Q.1. How do I use the graph and specs to calculate the GB at 120v?

Q.2. Having determined the ideal GB how do I then apply that voltage to the circuit? This is the bit I don’t understand. Most GB batteries have taps at -1.5v increments. The circuit shows GB-1 and GB-2. I take it that doesn’t refer to voltages, just the numerical denomination. GB-1 connects to the secondary of the LF Transformer, so assuming the valve GB should be -12v do I select that tap on the GB battery? But what if I run the valve at 150v and the GB is then -28v, where do I get -28v from? If GB-2 also needs -12v, do I connect GB-1 and GB-2 to the same tap on the GB battery? This is the bit that’s confusing me.

Q.3. The circuit also employs automatic bias. I am taking it that as long as I calculate the correct value of resistance being R=(Vg x 1000)/Ia (where Vg is desired bias voltage and Ia is the anode current in mA) that I don’t need to otherwise concern myself about the actual GB voltage applied to the valves?

Lots of questions I know, but I am keen to understand the correct methodology to apply, especially as I may elect to vary the HT voltage to see the effect it has on performance.

Thanks

Nick

This does depend on the valve specs. at a guess I would say -3 for V2 and -6 for V3, best practice in the day was to use the maximum negative bias with minimum distortion at listening level (no point using less bias, it costs money). That way the HT current was the least saving on a (very) expensive HT battery. Only adjust the bias if using a multi tap battery when the set is off, with a variable supply it will be fine left on and is probably the best way to get a feel for it.

[BobGreen]

I recall that Grid Bias Batteries were used in those days with sockets at every 1.5v increment up to 9 volts. The idea was to chose the bias levels which gave the best result. Well before my time!

[Herald1360]

Looking at the SP2 characteristics, it's clear that it's an output valve- at 100V Va and -12V Vg the anode current is around 22.5mA which is a lot for any battery powered valve. It might do for the final stage in the set but it would be unsuitable for the AF voltage amplifier stage.

-12V is high for battery GB and -28V even more so. There's no physical reason why you couldn't use a 20 cell 30V battery with taps but it would be rather unwieldy. It's far more likely that the SP2 was intended for a mains powered set with a separate bias supply, though its filament would still use a 2V accumulator cell for supply.

[SurreyNick]

OK, thanks. I think this answers Q.2. of the three above. Can anyone help answer Q.1. and Q.3.?
Thanks
Nick

[Herald1360]

Q1. Just interpolate between the 100V and 150V curves on the spec sheet to get a reasonable idea of what happens at 120V. About -19V ish? Given the likely tolerances on such an old valve, sorting it by trial and error to give about 20mA Ia would probably be best.

Q3. I can't see any signs of auto bias in the circuit posted. It may be a reference to how the detector stage biasses itself rather than anything to do with the audio stages which clearly use fixed bias.

[SurreyNick]

Thanks Chris. That way of estimating the GB is quick and easy. Q.1. answered

Now just to Q.3.

You're right, my circuit doesn't currently include auto bias. At the moment my receiver design employs battery bias. However, if I understand FJ Camm, by employing automatic bias (which he claims can be done with this circuit) I can dispense with the GB battery altogether. The modification he proposes to the circuit is shown in the attached picture.

My original Q. 3. was very badly phrased and I apologise for that. I should have been more careful proof reading what I had drafted. I should have started the question as follows: "Q.3. The circuit can also employ automatic bias..." Had I written it that way then the whole question would have made sense. i.e. as long as I calculate the value of the resistances correctly (22 and 23) according to the desired GB for valve two and valve three and also the respective anode currents in mA then I don't need to concern myself with anything else?

Nick

then I don't need to concern myself with anything else?

Not if you get it right, it's more of an automatic global bias, not for individual valves (like cathode bias), with only three it's a fair compromise. May be easiest to start with a couple of pots and a meter in each anode as they will be interdependent and you have old valves.

[SurreyNick]

Oh. I read it that R22 was for one valve and R23 for the other! Camm says in his book..."Fig.23 shows the arrangement of the two resistors for the two values of bias required."

The idea of the pots I like. Further on Camm does mention something about using a pot along with a fixed resistance for the output valve because a certain amount of preliminary adjustment of bias can be needed and this arrangement then avoids the risk of the valve being run entirely without bias.

Nick