[Amps] 8877 at 50 MHz

[TexasRF@aol.com]

Hi Dick, I guess we all have our pet designs and as usual, there is more than 
one answer.

I have seen enough KAPOWS! over the years that I tend to design towards 
minimizing them. If the cathode is connected to one side of the heater then there 
is no chance of a cathode to filament breakdown in case of a fault condition. 
If the cathode is connected to the heater and the heater is connected to ground 
then the possibilty of excess fault current blowing the filament open exists. 
Therefore: I would connect cathode to heater, bypass the heater for rf and 
float the heater transformer above ground. To make sure all of the drive power 
is applied to the cathode, I would use a bifilar heater choke with appropriate 
bypasses to ground at the transformer end.

A common failure mode in case of a plate to ground short (including plate to 
grid flashover), is the B- can rise to near the B+ potential without proper 
surge protection circuitry. If this happens, you can have a cathode to grid 
surge, possibly damaging the tube. You can also subject the plate and grid current 
meters to destructive currents. In addition to the usual 25 to 50 ohm B+ 
resistor, a single hefty diode connected from B- to ground can shunt the fault 
currents to ground. If the cathode and heater are not connected together, the 
same fault condition can damage the tube cathode and heater.

There are definitely some "gotchas" here!

Hope this make sense.

73,
Gerald K5GW