2W screen R's for EL34

[mhuss]

HIWATT amps used typical 1k/5w screen resistors on the 50 watt amps, but an odd arrangement for the 100 watt ones. There was a shared 470/10w screen resistor for all four screens, and each screen had an individual 100 ohm 1/2w resistor. I'm sure the 100s were in part to even out the screen current from tube to tube, but they also acted as fuses. I've often seen a bad tube and a single burned resistor to match.

[Chuck H]

I built an EL34 amp using high-ish voltages and even blew up a couple of tubes refining the circuit and finding a tube that could stand up to what might be expected of a REAL EL34. After running numbers I ended up with 3W for the 1k individual screen resistors. They held up fine in the acid test (crank it and play the $h!t out of it for an hour) and are still in the amp and being used by the owner.

470V on the plates and probably 460 ish on the screens. I don't have the amp with me to check actual numbers. But I can tell you that the 3W resistors held up fine when one of a pair of JJ's AND one of a pair of Sovteks blew their screens.

[g1]

For screen resistor dissipation calculation one needs to determine the total RMS voltage across it.
The total RMS voltage is the vector sum of average DCV and RMS ACV.
It calculates as square root of (Vdc² +Vac²).
Unfortunately most so-called True RMS meters are AC coupled and only read the AC part, so it takes 2 measurements (AC and DC).

I was able to take some measurements with the 470R screen resistors but did not have a chance to try with 1K's. Circuit as shown below, voltages at A and B were about 10V less with new PT. Idle current was about 60%, 30mA at 505V. JJ E34L. In this circuit, suppressor grids are tied to bias. Amp put out about 60W clean using 4 ohm load. Signal was 400Hz.

Idle: 2.4VDC
13VAC output (42W): 8.2VDC, 10.5VAC
18VAC output (80W): 22VDC, 30VAC

[Helmholtz]

When you're not sure if your meter takes into account DC in true RMS mode, just connect a battery.
If the meter reads zero, it's AC coupled and ignores the DC component.
My Fluke 87V reads AC rms only, while my Brymen 859s correctly reads AC +DC rms (one of the reasons why I bought it).

[Chuck H]

When you're not sure if your meter takes into account DC in true RMS mode, just connect a battery.
If the meter reads zero, it's AC coupled and ignores the DC component.
My Fluke 87V reads AC rms only, while my Brymen 859s correctly reads AC +DC rms (one of the reasons why I bought it).

Good info to know since my main meter is an 87. I have a few other cheapies for sacrificial purposes or when I want to see needle movement but no other high end meters. So it's good to know what my meter WON'T do.

[g1]

When you're not sure if your meter takes into account DC in true RMS mode, just connect a battery.
If the meter reads zero, it's AC coupled and ignores the DC component.
My Fluke 87V reads AC rms only, while my Brymen 859s correctly reads AC +DC rms (one of the reasons why I bought it).

So this sounds like you are wanting the DC and AC with the Brymen?
I know that my meter does not give me the DC reading when I measure AC ripple sitting on B+.
But I also know that using your formula, the readings I got in post #18 would give impossible levels of screen current.
(edit: last sentence is wrong, mis-calculated)

[Helmholtz]

So this sounds like you are wanting the DC and AC with the Brymen?

Oh sure. I wanted a meter that can read full or total RMS including DC. Always thought the term True RMS is misleading.
To be clear, the Brymen has 2 RMS modes. Default is with AC coupling, so standard "true" RMS.

But I also know that using your formula, the readings I got in post #18 would give impossible levels of screen current.

The formula gives the total RMS voltage, from which the resistor dissipation is calculated.

The screen current specified in tube data is (averaged) DC, RMS current.is not needed here.

[g1]

The formula gives the total RMS voltage, from which the resistor dissipation is calcultated.

The screen current specified in tube data is (averaged) DC, RMS current.is not needed here.

Sorry, I had mis-calculated in post #21 and somehow got impossible numbers.
The correct numbers for 470R screen resistors would be: approx. 0.377W at about 42W output (for this 60W amp) and about 2.9W when heavily clipped at 80W output.

[Helmholtz]

The correct numbers for 470R screen resistors would be: approx. 0.377W at about 42W output (for this 60W amp) and about 2.9W when heavily clipped at 80W output.

Yes, so 5W resistors are appropriate here.
Even though I think EL34s should have larger value screen resistors than 470R to avoid redscreening.
22VDC across a 470R means a DC screen current of 47mA. Multiplying with a corresponding screen voltage of maybe 430V (?) results in a screen dissipation of around 20W while the limit is 8W.