Is 6.8V DC Too High For Heaters?

[CraigGa]

My Benson Monarch build uses DC for the preamp heaters and I've just powered it up and I'm getting 6.8V on the heaters with 0.45V AC ripple, Is that a bit high?
Most discussions I've seen seem to quote either +7% or +10% as a maximum and that drops right in between those.

Or should I just put a diode in and drop it by 0.6V

The output tubes are at 6.6V AC so I guess my transformer is a little bit high.

Thanks in advance

Craig

[R.G.]

Short answer: yes, that's too high in my opinion. Knock off a silicon diode drop (0.6-0.8V) or so by putting one in series or by tinkering the DC supply to lower it a bit. It's better to run non-power tubes a bit lower than nominal than a bit too high. The tube heater lifetime will be shorter if you run them high. You may or may not care if they wear out a bit faster, but that's up to you.

Long answer: Yes, it's too high, and will shorten the life of the tubes. Heated filaments wear out on average inversely proportional to the fourth power of the applied voltage. So if you increase the voltage by 10% over nominal, the lifetime goes down by a factor of 1.1*1.1*1.1*1.1 = 1.461, or about a 32% shorter life (if something else doesn't kill them first!)
Electron emission goes down as the temperature goes down. However, you only need "enough" emission to keep the space charge around the cathode in place, and this is relatively easy for low current signal tubes with oxide-coated cathodes - that is, preamp tubes. You can get even longer life by running them a bit lower than nominal, as the same math applies for voltage on the heaters: tubes run at -10% of heater voltage will have heater lifetimes about 52% longer than if they're run at nominal voltage. So it's better to run signal tubes a little lower on heater voltage rather than a little high if you want best life.

The power tube heater voltages are OK. Power tubes need all the electron emission they can get without going higher than +10%.

Is this a "vintage transformer"? Perhaps designed for a 110Vac electric line, not 125Vac, which is common today?

[CraigGa]

Thanks for the comprehensive answer R.G.

I'll drop a diode in there and get it to a more reasonable value.

No the transformer was custom wound with a 240V primary, the wall voltage was 243 at the time.

Craig

[CraigGa]

I just found some 1N5400s in my stash, 50V 3A diodes.
I'll put one in tomorrow.

Craig

[katopan]

Wanted to add an experience post to back up the theory.

A lot of datasheets say heaters are +/-10% which means up to 7V. Years ago I used to live in a house where the wall voltage was right at or sometimes above the allowable supply spec. House was full of dimmers and back then the only dimmable high efficiency light globes were halogens. They used to fail with way shorter lives than normal. Also my 18 Watt amp was the only build I had at that time and was the only amp I really played through, always VVR'd right down. Just under 7V at the heaters burnt out the EL84 plates way quicker than normal and using VVR for 90% of their life meant there was no way plate dissipation was a factor in that.

[Reeltarded]

and off topic but on subject

kato's experience is very similar to mine while variacing everything.. because a tube is basically a light bulb. i found reduced voltage not only to sound just great and hardly different but also to increase tube life to near infinity? i have a couple heads that are 50 years old with the original tube sets. that is insane.

[CraigGa]

Thanks guys, the light bulb analogy makes perfect sense.

I've put the diode in and rather confusingly the voltage on the heater pins has dropped to 5.9V that should be ok shouldn't it?

If not I can easily move the PI to AC heaters and take some load off the DC supply.

Craig

[martin manning]

If I understand correctly, what you have now is a single diode in series. If so, you’re getting half-wave rectified DC. To drop approx 0.7V you need to put two diodes anti-parallel. Then to center the heater voltage around 0V you need to use an artificial center tap, 2x 100 ohm resistors connected after the diodes.

[CraigGa]

No Martin,
It's already bridge rectified, I just added a diode in the +ve leg to drop the voltage a bit.

Craig

[martin manning]

Ah, should have read more carefully. 5.9V is fine. What size is the filter cap?

[CraigGa]

4700uF, I tried paralleling another across it and it didn't make much difference.

Craig

[martin manning]

If anything more capacitance would reduce the ripple and increase the average DCV. A smaller cap would lower it. In any case you have a solution that works.

[CraigGa]

Thanks Martin,

I just have to speaker connections and B+ to do now before it gets a proper power up.

Craig