Another Tube Question: Heaters/Filaments, Centre Taps and Current Draw

[frequencycentral]

I've been using 12au7 and 6111 (submini) tubes.

As you will know, the 12au7 requires 12.6v at 150ma across pins 4 and 5, or 6.3v at 300ma across pin 5 and pin 9 (the centre tap).

The 6111 has no centre tap, and requires 6.3v across pins 3 and 6.

So here's the question:

On the breadboard I'm working on a circuit using two 6111s. I've powered it like this:

12v  >  6111 (#1) pin 3
6111 (#1) pin 6  >  6111 (#2) pin 3
6111 (#2) pin 6  >  ground

So i figure that what I've done is effectively created a 12v/300ma heater with an (unused) 6v/600ma centre tap [6111 (#1) pin 6  >  6111 (#2) pin 3].

So.................am I right in assuming that this arrangement will draw 300ma?

[DougH]

It will work. You basically have wired the tube heaters in series rather than parallel.

I've seen some old tube schematics that use 35v heater tubes where there are weird series arrangements to power the heaters. Here's a more contemporary example: http://www.dogstar.dantimax.dk/tubestuf/dv8howit.htm

For breadboarding you are fine. I don't know if I would build it like this- I might just get a 6.3v transformer instead. The EE in me feels jittery about series wiring (as opposed to parallel) but I'm not sure why. If one of your heaters goes bad, all the tube heaters will shut off. And I can't remember what happens to a tube if you have high voltage applied with no heat, but I don't think it's a good thing. So I sense that from a reliability standpoint it may be better to wire them in parallel. Fred's series/parallel arrangement in the above circuit is supposed to account for this but I have not looked at that closely. But for breadboarding/experimenting I see no problem with what you're doing at all.

[frequencycentral]

Thanks Doug, confirms what I thought. However, for me, if it works on the breadboard I'll use it for the finished article too.

Next question:

The centre tap [6111 (#1) pin 6  >  6111 (#2) pin 3] - could I use this as a half voltage source/vref/floating ground for the I.C. part of my circuit?

[slacker]

Like Doug said you've wired the 2 heaters in series, so each one is dropping 6 volts, so the current draw is the same as for one heater running at 6 volts.
The heaters are basically low value resistors, so if you think of them like that it will probably make sense, that's what I do anyway.

If 6 volts/heater resistance = 300ma then 12volts/2*heater resistance = 300ma

[DougH]

If your heater voltage supply is DC it should work for a reference point. However, I'm not sure how much tube heater resistances will vary, so if you swap tubes it may change a little, how much I don't know. If it were me I'd probably use a pair of fixed resistors for a divider and be done with it. Resistors are what, 6 cents a piece? Either way you probably want a filter cap on the Vr rail to stablize it.

[frequencycentral]

Thanks DougH and Slacker. I woke up this morning with this question buzzing through my mind. It's good to get such fast clarification.

[DougH]

You know, now that I think about it, I wired the 2 pentode heaters in my pentode driver in series and never had a problem with it. DOH!  I need more coffee...

[frequencycentral]

Yeah, it opens up a whole new avenue for me. I was shy of using two 6111s in a pedal because of the 600ma draw. Now I've got the series thing figured, there really is no (ma) disadvantage to using 6111s compared to 12au7s - and a big space saving advantage. I'm not one for having my valves on show, sticking out of the pedal. I prefer them tucked safely inside.



When I've finished my current work (tube stereo vibrato/panner/Leslie sim), I'm going to work on a tube phaser based on mac's idea for a FET phaser. After that I'm going to build the tube VCF circuit I found and posted yesterday, I'll make it my variation on a FSH/Dr Q.

Then I will make a Tube Town Pepper Shredder using two 6111s. In a tiny box.

It's going to be a great summer for tubes! I just love working in education - no real work for me until September!!

[petemoore]

  It's not the 'defensive' approach, but as long as the heaters don't go on offense, you're fine.
  I see it as about the same...problematic-wise as parallel heaters...especially if you need all them tubes for basic function.
  Now I'm trying to remember...I think I've seen may 1 or 2 in say 100 or 200 tubes, that the heater failed.
  And it's been a while too...put good tubes in you should be fine...when the one heater goes you're into fix-up territory anyway, and it would be non-catastrophic situation anyway [except for the one already bad heater filament].
  ...A drawback I think of before time expires !
  ~Slight...it's harder to find the failed heater filament when there's a bunch of them and they all go dim at once 
   Much easier it'd seem to debug a paralleld heater design.

[DougH]

put good tubes in you should be fine...when the one heater goes you're into fix-up territory anyway, and it would be non-catastrophic situation anyway [except for the one already bad heater filament].

The only issue I can see is if the heater failed in a way where it dead-shorted. Then it could potentially put the full 12 volts across the other heater and potentially damage it too. So at worst you are out two tubes instead of one - not what I would really consider "catastrophic" I guess. But I would think the likelihood of this happening is slim to none. Most of the time if a heater fails it will open up. Given that the resistance is an intrinsic property of the heater material itself I don't really see how it could short. Then as Pete said, 1 or 2 failures per 100 is not exactly something you need to worry about anyway IMO.

Another thing to keep in mind, in my experience anyway, is tubes are pretty forgiving and failures occur at a relatively slow rate. Apply an out-of-spec voltage to a transistor and pffft! Before you even think about it, it's gone. I don't know how long it would take for 12v to burn up a 6v heater, but it seems to me the time it takes to turn it off when the first tube malfunctioned would be quick enough. I've accidentally run an output stage unloaded for 15-20 minutes twice and never had a problem (forgot to plug in the speaker cable). I wouldn't recommend trying this, but it does seem to indicate a certain robustness you get for free with vacuum tube circuits.

Okay, I think I flogged this one to death at this point...