Fault Protection

3-500Z under typical fault conditions

The primary fault path is from anode to the grid. In dozens of amplifiers I've seen with 3-500Z's, the fault is primarily from anode to grid support cone. Since the area is widely distributed there often are no marks. You can see the fault is concentrated at the edges of the anode with a primary target at the edges of the grid support cone. The area under the plate is actually plasma free, as evidenced by the dark color.

This is NOT generally caused by a parasitic.  A good tube isn't any more likely to arc or fault at VHF than at very low frequencies. This particular tube is into plasma at 7MHz at 4500 volts peak RF voltage.

This is a simplified circuit showing where hard faults commonly occur in an HF PA:

This is a typical fault path. D2 is removed:

The grid meter reads backwards with 167 amperes. The plate meter reads forwards with 167 amperes flowing through the meter.

Adding D2, we have:

In this case we have 161 amperes flowing through D2 and 5.7A through the plate and grid meter shunts. 167 amperes flows from the grid through F1.

If F1 opens before supply voltage dumps, we have:

The GK diode turns fully on. If cathode saturation current is exceeded the grid-cathode voltage will rise to a value that causes the grid to arc to the cathode. Either way, all the HV dumps through the grid to the cathode instead of F1.

The danger of this is the exciter, if connected, receives a transient that may be hundreds of volts. The very sharp rise time of the arc means we essentially are running a spark transmitter backwards into the radio. D2 is no longer in the arc path of the circuit, and so the plate meter is exposed to the full fault current.

Instead of losing F2 or D2, we now greatly increase the risk of damaging:

1.) The exciter

2.) D1

3.) The input system

4.) The plate meter

The tube still has just as much surge current at the arc strike, but now has a sustained arc at lower current until some other device having enough hold-off voltage to quench the arc opens.

This is why grids should NEVER be fused.

Return to fuses and grids