[Amps] Grounded Grid(Screen) 4CX250B power gain

[Colin Lamb]

"Anyone have a comment about the theory behind connecting the grid through a resistor to the cathode in a "grounded grid" amp"

>From the 1930s on, people have been using tubes as they were not intended to be used.  On occasion, tetrodes are connected as triodes.  But, there is no universal way to connect them.  Sometimes, the screen is connected to the plate.  This could only be done if the voltage limit of the screen is not exceeded by the plate.  Sometimes, the screen and control grid are connected together.  

We know that in a 4CX250 that the control grid will fail if much power is applied, so it will not operate (for long) in the same grounded grid configuration that we run 4-400 and 4-1000 tubes.  Collins operate the 4CX1000 in grounded grid (screen) in the 30S-1, but do so in an unusual method by floating all the voltages.  In that case, all the voltages are normal.  

Mentioned recently are a couple of different methods of operating the 4CX250 in grounded grid.  One is to connect the control grid to the cathode and ground the screen grid.  Another method is to ground the screen grid directly and ground the control grid through a resistor.  I have never seen any data regading control and screen grid currents in the respective grids, and it would be useful to get such data, including gain of the tube.  Tube manufacturers, of course, do not have such data.    

One of the Radio Handbooks of the late 1940s shows a modulator using push-pull 813s which are connected as triodes.  The control grid is connected to the screen grid through a series resistor.  This would, in effect, be one of the same connections as described in the 4CX250 amps.  I would assume it works.  The 813, however, does not have a control grid that is as fragile as that of the 4CX250.  

As to connecting the control grid to the cathode, it should reduce power dissipation of the control grid, and would certainly reduce gain.  I would sure like to see results of dc tests of both configurations.  It would be easy to set up on a workbench, and almost any tetrode could be used to make the tests.  

73,  Colin  K7FM