Reducing heater voltage with resistors calculation

[frank57]

Right now I'm using 2 sets of diode to reduce 7.7 to 6.3 or so.
They tend to add some buzz.
I was thinking to try 2 resistors.

Are these calculations accurate?

So 4 12ax7 *300ma=1.2 amps.(phase splitter is 12at7)
2*.76(el84)=1.52
Total:2.72amps.

7.7-6.3=1.4volts
1.4/2.72=.515 ohms?
1.4*.515=.721 watts?
Use 2 x 5 watt .51 ohm?

[tubeswell]

What sort of diodes are you using? I would use a pair of nose-to-tail 6A diodes in series with one or both sides (depending on whether a 0.6V drop or a 1.2V drop is needed) of the winding.

[sluckey]

Are these calculations accurate?

So 4 12ax7 *300ma=1.2 amps.(phase splitter is 12at7) OK
2*.76(el84)=1.52 OK
Total:2.72amps. OK

7.7-6.3=1.4volts OK
1.4/2.72=.515 ohms? OK
1.4*.515=.721 watts? WRONG!

P = E*I or I²*R, or E²/R, but not E*R, so P = 1.4*2.72 = 3.8W

If you want to maintain a balanced filament string, then split the .515Ω resistor in two, placing a .25Ω/5W in each filament line.

[frank57]

Well I guess I blew that calculation.
Is 5 watts enough?


I am using 2 sets of diodes now, but I find they add  some buzz when you turn up.
Reducing the voltage has reduced the hum on this amp quite a bit.
With one set of diodes you end up with an unbalanced voltage.
Thought I'd try resistors.
What's the downside of using resistors?

[kagliostro]

What's the downside of using resistors?

The involved math

---

Here you can find some documentation

Kagliostro

[tubeswell]

If your heater winding is centre-tapped, another option could be a 1.5V reverse-biased zener in series with the centre tap of the heater winding and the ground return (banded/end cathode pointing to ground).  You'd want a 10W zener.

[stingray_65]

I'm failing to understand how adding a Diode in line with your heaters could case a buzz?

Not saying the buzz isn't there or that the diodes are causing it.

I've used diodes a few times to drop my heater voltages when I use vintage PT's salvaged from old equipment. Have yet to experience any noise.

Could some one explain.

Ray

[frank57]

What Paul Ruby told me:
That's an easy way to drop a half volt of filament voltage but it adds a lot of high frequency noise to the filament lines (buzz....). A power resistor to drop some voltage (if needed) would be better

In my case it's a poorly designed amp to boot which is not helping.
The diodes seem to be adding some extra noise when turned up.
You don't notice it low.
There is a big buzz problem to begin with though when turning up the gain which might be caused by this:
high-frequency diode commutation from the B+ rectification is "jumping" across the B+ winding to the filament winding inside of the power transformer.
Anybody know anything about diode commutation?
 

[tubeswell]

Well if you took the diodes out (or bypassed them) and the buzz goes away, it would probably be them that's causing it. If not, then its something else.

[frank57]

Oh those diodes on the filaments are adding extra buzz on top of the original buzz.
They didn't twist the ac wires going to the mains switch very well either.
What gauge wire can I use there?They cut them too short to twist properly.

[frank57]

I gave the resistors a try.So they are generating less bzz if any compared to diodes.
The resistors are bringing it down from 3.79 to 3.017 and 3.021.
On 9 to 4/5 I get 6.07-6.09.(from 7.7 or so)
But the 2 5 watt resistors are getting hot to the touch.
Should 10 watt be used?
Maybe try .22 which would lower it a bit less.
Is it 3.8 watts dissipation on both resistors right now?

Can you get away with 5 watt ones or better to try 10?

[sluckey]

It's OK for power resistors to get hot. But,,,

1. Measure the voltage across one resistor.
2. Use this formula to determine the actual power dissipated by the resistor.    P = E²/R
3. Double the resultant number in step 2 for a safety factor.
4. Use a standard wattage resistor ≥ the resultant number in step 3.

[frank57]

Okay both resistors are pretty bang on.
I get 3.84 going in and 3.08 3.079 or so going out.
so a drop of .76-.77 or so.
.76*.76=.5776 / .25=2.31
2.31*2=4.62
Is it too close to 5 watts?

If it reduces a bit more say .78
then .78*.78=.6084/.25=2.43
4.86.

There is some fluctuation on one resistor, but not too much.
I don't hear any bzz or hiss coming from them and they are reducing hum on the master volume.

They are pretty hot though.

[sluckey]

Is it too close to 5 watts?

no

[frank57]

Why are the resistors so hot?
They are awfully hot to the touch.
Is it normal?

[sluckey]

Why are the EL84s so hot?

[frank57]

The plate voltage is 382 or so and the bias is -16.78.
The tubes are all super close.
I think it's 3/8 of an inch space between them, including the two el84's.
The yellow heater wires do run a bit close to the el84's on the outside of the amp.