heater/cathode query

[kianbach]

I've built a cathode follower preamp with 1/2 an e88cc direct-coupled to a 6j5.the cathode of the 6j5 is at 150v.I've used a divider from B+ to raise the filament potential.Just wondering if it would be easier to take the centre of my 2X100 ohm resistors straight to the cathode of the 6j5 instead.This would put the centrepoint of the filament at 150v,same as the cathode or would this do more damage?

PS: I AM A NEWBIE>only built for a year or so.....

 

[Johan Potgieter]

Careful here, Kianbach.

If you use one heater winding for both tubes, the maximum heater-cathode voltage for the ECC88 (cathode negative) is only 100V. The value for the 6J5 is unavailable in my manual; I would believe some 100V. So you have a small window only if you have only one 6V winding.

Anyway, I would not attach heater windings unnecessarily to signal points, even if low impedance as in a cathode follower. There is a way out if absolutely necessary; connect the heater tap through a resistor some 10X your load resistance and decouple on the heater side with an electrolytic plus polyester capacitor. That way you keep rubbish coming through the power transformer out of the signal. Otherwise stay with a dc tap from a voltage divider.

. . . . and being a newby is fine. That way you are still unencumbered by some of the ridiculous biases often encountered!

 

[PRR]

There was a time we didn't do this. OK, there was a time when all cathodes (mostly filaments) had to be common. But for many decades after that, you would look long and hard to find a cathode floated more than 50V above ground, unless it had its own heater winding. I think this fad to float cathodes all up and down the power supply voltage comes from transistor thinking....

> I've used a divider from B+ to raise the filament potential.

Bias the heaters to about +90V. That is within 100V of either cathode and so should not strain the insulation. It is positive to the gain-stage cathode and sucks-off any heater leakage. It is negative to the CF cathode, but cathode followers don't hum unless something is terrible wrong.

As Johan says: use a fairly high DC impedance bias source with a bypass cap. That way if a heater does leak a little, the bias supply drifts to minimize the leakage.

 

[SY]

use a fairly high DC impedance bias source with a bypass cap

Johan and PRR: Interesting! I've seen emitter followers recommended to provide the bias for the heaters- what do you see as the flaw in that? (FWIW, I've always done it your way)

 

[PRR]

> I've seen emitter followers recommended to provide the bias for the heaters

You didn't see it from me.

Flaws? Rather, WHY???? Maybe too many transistors cluttering the bench? I sure don't see the point. Heater-cathode leakage should be, and often is, so close to zero that no special precaution is needed. Positive-biasing the heaters is really more about Production than operation or sound. If 2% of tubes have noticable leakage, it is cheaper to add a couple resistors to bias that away than to do tube-rolling on the assembly line. In DIY it is reasonable to toss-aside tubes with excess heater leakage. Use them in a high-level stage. Or with so many types still selling new/tested for under a buck (in 1959 money), just toss the leakers.

I don't know which way this emitter is pointing, but if a tube REALLY wants to leak, it is going to find a way to pop the transistor. And the amp may then work very fine.

 

[Johan Potgieter]

Yes, I have the same question as PRR. Or do I understand it wrongly; is this just a dc bias or was the cathode (signal) put on the base of the emitter? Then there is some sense in actually letting the heater follow the output cathode with a sort of "low impedance following" but without taxing the signal circuit (yes, this is perhaps what is meant). I would then support it, but as PRR said, somewhere the transistor is in danger of being popped (protection diode in BE circuit, etc.)

I have also read that it is good practice to have the heater some 20V positive wrt the cathode. That way cathode-heater cannot act as a conducting diode capable of inserting hum into the signal. The writer was of the opinion (or found) that even with perfectly good tubes some heater-cathode interaction was unavoidable.