Directly heated cathodes

[skajam66]

Hi,
I'm looking for clarification on how manufacturers specify voltages relative to the cathode when the cathode is directly heated. Since a directly heated cathode will have a potential drop across it, does a voltage relative to the cathode (e.g. grid voltage, Vgk) refer to the plus end, the minus end or halfway between?

[tomchr]

That's a good question, actually... I'm guessing as the directly heated tubes used to be heated by AC voltages, the Vgk refers to the average value of the cathode voltage.

It doesn't always make that much of a difference, though. Take a 300B for example. Common operating points involve Vgk = 75~85 V so the 5 V across the filament don't make that much difference in the operating point.

That's my hand-wavey explanation, anyway...

~Tom

[skajam66]

Thanks for that tomchr - one reason I asked is because I'm looking at low voltage tubes and there the Vgk is much smaller -1V or -2V is not untypical

[skajam66]

tomchr has provided some thoughts on this - does anyone else have any perspectives to share?

[Enzo]

I think your question assumes the heater voltage is with respect to ground, when it may not be. For example in a 5v rectifier tube, the 5vAC for the heater is not referenced to ground. SO while the two ends of the heater may be 5vAC apart, they are not changing their voltage with repect to ground. Think of it this way: If I hold a ball in front of me 2 meters off the ground and move it up and down 10cm each way, then the altitude from the floor will also vary by 10cm. But if I move the ball 10cm sideways, parallel to the earth, then the motion is still there, but the altitude does not change. In other words that 5vAC is not in series with the 500v of DC or whatever is coming out of that cathode.


I know you are not asking about a 5U4, but perhaps this is a useful perspective.

[Rod Coleman]

If you look at many RCA data sheets (eg the 45) you can see that they assume you ac heat, and that you connect the centre-tap of the filament secondary as the reference point for the specified grid voltage.

Marconi data sheets give the application circuits exactly like this, and the assumed grid voltages alongside (see, eg, July 1956 PX4 data).

I believe that this is the assumption for most large DHTs intended for operation on mains power. It's only a specification reference. This is understandable - the idea that DHTs would sound better in the 21st century when heated by very carefully designed dc current-driven supplies would not have occurred to the original designers!

Meanwhile, the Western Electric 300B is more explicit [see January 1950 data], and specifies ac heat as the assumption, and a footnote instructs the designer working with dc filaments to subtract 2.5V from the value measured at the negative filament, to equate the characteristics with ac-heat.

OTOH, With 1.2V and 2.0-2.4V (and some 4V) filaments the design intention was clearly battery-powered filaments, with one terminal connected to ground. Some battery valves give characteristics for both the +ve and -ve sides at ground (see Philips 1969 DM70 data sheet, page 3, notes 1 & 2.).

If your battery-filamentary triode's data sheet is not specific about whether to connect + or - filament to ground-reference then simply do a quick test with FIL - at ground, and a little more than the typical grid voltage at the grid. You can use alkaline batteries for the neg-bias (see "26 preamp" thread) and these work very well for audio. This assumes you use dc for the filaments - you will certainly not achieve the full potential for sound using ac. Search the forum for DHT Heater to see the various solutions for DHT heating, and see what others think of the sound of them.

[artosalo]

With 1.2 V filament tubes, such as soviet subminiatyres I have tested, there is big difference in anode current when you change the polarity of the filament battery.

[skajam66]

enzo, rod, artosalo, thank you very much for your posts. Let me go away and do some reading and testing and I'll report back...

regards..

[Junm]

Does anybody tried using 5Vdc on filaments....what are the pros and cons...?