[Amps] When does an GG Amp draw current

[Chuck Counselman]

Paul Baldock <paul at paulbaldock.com> wrote:
> 
> My 2 x 3-500Z amp starts to draw a few mA of grid current with about 
> 1 watt applied. Bias is about 8V.
> 
> My home brew GS35B amp  starts to draw a few mA of grid current with 
> about 15 watts applied. Bias is about 40V.
> 
> Is my GS35B amp behaving as it should? When does an GG Amp start to 
> draw grid current? Is it when the peak rf voltage applied to the 
> cathode exceeds the bias voltage?


Roughly speaking, yes; it’s when the grid-to-cathode voltage goes positive.  This answer is only approximate because several aspects of the design of the tube matters.  Some tubes are designed deliberately to draw very little plate current when the grid-to-cathode voltage is zero.  The 3-500Z is such a tube and the suffix “Z” in the type number “3-500Z” signifies “zero bias.”  <http://www.umich.edu/~umarc/station/docs/3-500z.pdf>

Whereas, quite a large negative grid-to-cathode voltage may required to reduce the plate current of another tube to a small value; and such a tube may draw quite a high plate current with a small or moderate negative grid-to-cathode voltage.  Such a tube may be designed to operate as a quasi-linear amplifier with its grid never going positive with respect to its cathode, and with its grid never drawing positive current.

For any tube, for any grid-to-cathode voltage, the plate current also depends on the plate-to-cathode voltage.  This is especially true of triodes.  To produce maximum output power, a tube may require very high (positive) plate voltage; and, with this high plate voltage, a large negative grid bias may be required to limit the average plate current and the average plate power dissipation to safe values.  For a tetrode or a pentode, the plate current is relatively independent of the plate-to-cathode voltage, and it depends strongly on the screen-grid voltage.

In a tube having a directly heated cathode (such as a thoriated-tungsten filament), the cathode voltage varies from one end of the cathode (filament) to the other; so the grid-to-cathode voltage varies from one end of the cathode to the other, by the magnitude of the voltage that heats the filament.  In a 3-500Z, the RMS filament voltage is 5 volts, usually 60-Hz AC, so the peak filament voltage is about 7 volts; and 7 volts is not insignificant relative to the DC grid-to-cathode bias voltage.  (The plate current does not have much 60- or 120-Hz modulation because, when one end of the filament is a emitting a greater than average flux of electrons per unit length, the other end of the filament is a emitting a smaller than average flux.)  However, this spread of grid-to-cathode voltage smears-out / softens the cutoff of plate current as grid-to-average-cathode voltage goes negative, in the vicinity of zero average grid-to-cathode voltage.

To understand what’s happening in your amplifiers, you need to look in each tube manufacturer’s data sheets, at the graph of DC plate current vs. instantaneous grid-to-cathode voltage, to see what the tube’s DC plate current should be when no RF drive is applied and the grid-to-average-cathode voltage is just the DC bias voltage derived from a negative-voltage supply, a zener diode in the cathode circuit, or whatever.  Also look at the graph to estimate the average DC effect of the AC filament-heating voltage.  Then look at the graph to estimate the average DC effect of adding RF excitation to the instantaneous grid-to-cathode voltage.  In a grounded-grid amp, this grid is at RF ground and the excitation is applied to the cathode, but this fact does not materially change the analysis.

Thus, you should be able to understand the different behaviors of your amps.


73 de Chuck, W1HIS, who grew up without transistors and feels at one with his g.-g. triode amp.