bias: cathode current vs plate current

[Gizzer]

from the Marshall forum
Cathode current will always read higher than actual plate current because again cathode current includes BOTH the screen grid current AS WELL AS actual plate current !
So when for bias calcul, an EL34 tube (25Watts) we have to add 8Watts of screen dissipation for a total of 33 Watts (measuring by cathode current)

Exemple:
If the plate voltage is 400Volts

- biasing with plate current : (25W/400) X .7 (70%) = 43.7mA (maximum bias reading)

- biasing with cathode current : (33W/400) X .7 (70%) = 57.75mV into a 1ohm resistor = 57.75mA (V=RI) maximum

Does it make sense ? ? ? ?

Re-biasing this way (cathode current) made my amp sounding like it did never sounds that good before !

[martin manning]

I would not do it that way. You can’t assume that the screen is running at its limiting power. From the data sheet (Mullard, class AB at 425Va and Vg2) screen current is about 15% of plate current, so when biasing using cathode current, where screen voltage is equal to plate voltage, consider the max dissipation to be 25 x 1.15 = 28.75. Then, if you want 70% Pa, bias at 0.7 x 28.75 or 20.1 W.

[tubeswell]

You need to accurately work out the expected screen current for the screen voltage you are using, and for the load line presented by your OT and speaker load. Otherwise you can easily over-dissipate the screen.

For this reason, if running a single ended EL34 at 100%, the screen voltage should be around 50-60V lower than the plate voltage (to keep the load line through the knee of the Vg0 grid curve), and you should have a 1k-2k Rg2 (to eat up excess screen current under big signal swings).

[martin manning]

The load impedance doesn’t matter in compensating for screen current at the idle bias point, and screen current at that point will more or less go with plate current as the screen voltage is varied. If you want to know exactly what the screen current is (and the fraction of the plate current it represents) at the idle point, just measure the voltage across the screen resistor and calculate it.

[tubeswell]

I wasn't talking about screen idle conditions, I was talking about heavy signal conditions when the screen current rises. Load impedance matters at that point. If the loadline is too shallow (so that it passes well below the knee of the Vg0 curve, in the region where plate current drops off dramatically), there is a risk of dramatic rise in screen current. If the load is kept in the region of the knee, this reduces that risk. On the chart below - if the screen voltage were around 300 and the load remained at 10k, the load-line would arguably be too shallow.

[martin manning]

Yes, but the OP is only asking about adjusting his biasing technique to set 70% plate dissipation when the measurement is cathode current.

[Stevem]

Over the years and in terms of commonly used guitar amp output tubes I have found that play wise and fell wise there is a 5 to 6 ma cathode current range where the output tubes produce the Balls you would expect and once in that range is really where the hard clipping starts also !
Soft power supplies will clip and fart out much sonner then stiff ones!

Listen real close while you play with the Balls range of your cathode current measurements!

[pdf64]

...Exemple:
If the plate voltage is 400Volts

- biasing with plate current : (25W/400) X .7 (70%) = 43.7mA (maximum bias reading)

- biasing with cathode current : (33W/400) X .7 (70%) = 57.75mV into a 1ohm resistor = 57.75mA (V=RI) maximum

Does it make sense ? ? ? ? ...

See bottom of p2 http://www.mif.pg.gda.pl/homepages/fran ... e/EL34.pdf
Note that Mullard suggest much lower idle currents at 400V (30mA plate, 34.5mA total cathode current).

[Gizzer]

Ain't that low Wattage dissipation ?
400V X 34,5mA (cathode bias) = 13,8Watts
400V X 30,0mA (plate current bias) = 12 Watts

You'd get a clean tone for sure though !

"Tonaly" speaking, I will try to get in the 17Watts zone; a little less maybe, like biasing at 40mA "plate current"

[nickb]

To find the plate current by measuring the cathode current you need to make one additional measurement.

Measure the voltage across the screen grid resistor and divide by it's resistance to find the screen current. Now subtract that from your cathode current measurement to find the actual plate current.

[pdf64]

...Re-biasing this way (cathode current) made my amp sounding like it did never sounds that good before !

The nature of human perception is such that an increase in sound signal level, even at a barely detectable level, is almost always perceived as 'better'.
If the amp's bias control is twiddled (over a safe range) whilst it is passing signal, you'll notice that as plate/cathode current increases, so does signal level.
So yes, it's understandable that your amp will sound better as the bias is adjusted to increase plate / cathode current, but that may well be just because its gain is higher.
But it doesn't mean that it's working better or is in a safe operating area.
Mullard appear not be have used the '70%' guideline in their suggested operating point; consider the possibility that the rationale behind the 70% guideline is far too generic for it to sensibly supersede such venerable manufacturer's info.

[Gizzer]

Anyone here owns a Bias King or Bias King Pro ? (cathode current) discontinued since two years !
Check out the "formula" on the following picture :

[martin manning]

That's rubbish. Here's some measured data backing up the correction I proposed above:

[M Fowler]

Martin when I click on your image it brings up a black screen with white cells and I can't read any numbers?

Mark

[martin manning]

Hmm... Seems to work for me. Anyone else get that?