Mystery of the frying blocking cap

[ka1tdq]

I built a variation of the 6146 transmitter that was featured in the ARRL handbook in the 60's and I'm having a problem with it.  Every time I cause the final to conduct I fry the blocking capacitor going to the output network. 

The oscillator works fine, plate choke is ok and there are no shorts in the output pi network.  What would cause this?  B+ is around 375 volts and the only value I guessed on was the screen dropping resistor of 10k. 

I added a modified Heising modulator to the rig.  I also calculated a neutralizing network of 11pf off the plate going to a .001 off the grid.

All help will be greatly appreciated!

[Rick K5IAR]

What is the value of the blocking capacitor and at what voltage is it rated?  You're gonna get some serious voltage going through it.

Rick/K5IAR

[ka1tdq]

The largest voltage rating that fried was 2kv at 500pf.  Needless to say the 1kv's went too... they were .oo1.

[ka1tdq]

Could it also be a problem of grid bias not set correctly?  I have a tank circuit going into the grid resonating at 3885, but below it just a 27k resistor going to ground.  Do I need to replace the resistor with a choke and supply a fixed negative voltage, say -75 volts?

[N3DRB The Derb]

you fried a 2 kv cap with 375v B+ ?

how'd you do that? that's mad.

[w3jn]

Bigassed parasitic.

[AB2EZ]

I agree with JN.

The tuning capacitor presents a low impedance to ground for harmonics of the fundamental frequency, and also for high frequency parasitics.

If the 6146 is oscillating at a high frequency (a high order harmonic of the signal applied to its grid, or a parasitic) the current through the tuning capacitor and the DC blocking capacitor (and the tube) will be very high.

I suggest that you try removing the neutralizing capacitor ... Given the high value, and, in particular, the way it is installed (plate to grid via a .001?), it may be providing positive feedback instead of neutralization.

Best regards
Stu

[The Slab Bacon]

A 6146 with only 375v on the plate prolly isnt swamped down hard enough for good stable operation and could be very likely to go parasitic. Be sure that you have it biased down far enough and driven hard enough to be well into class C operation.

Also grab a scope and see that is comng out of it!! You may well find out there is a lot more coming out that thing than you want. (I.E. a "bigassed parasitic" ) 10Pf of capacitance fed back into the grid is more than enough to induce an oscillation if it is not properly phazed. Depending on frequency, and circuit constants it is very easy for "negative feedback" to become "positive feedback" and start oscillating.
So try disconecting your "neutralization" circuit as well. You dont want to directly couple the plate and grid together, you want to couple the plate back to the "cold" end of the input coil. But first try just disconnecting the neutralizing circuit.

                                                         The Slab Bacon

[N3DRB The Derb]

yeah, whut they all said. 

[ka1tdq]

Thank you for the advice.  I will do exactly that!  Jon

[W3RSW]

Interestingly, in the p. 182, 1962 handbook 6146 'all purpose' amplifier, I see the plate block cap at 500pf 20kv.  Now I wonder?  That was at B+ of 750 though. Was that hv cap all that was on hand or a real need? Hmmm,  Surely the ARRL labs had some smaller, cheaper caps, say a measly 5kv N750? heh, heh.

I see they also biased it at -75 at quiescent with a 0A3, & grid biased at op. conditions through a voltage divider.

Now in their cool p. 478, 1962  mobile transmitter, complete train using a 6883 at 600 volts in the final they use a 500 pf/ 6kv cap.

Did you include the parasitic coil/res. on plate cap?

I'd look at the input to the 6146 with the scope too.  But fwiw, I've even put parasitic suppressors on 6146's in modulator service.  Quite a hot tube at times, huh?

[WA1HZK]

Sounds like a power oscillator to communicate with the mother ship. Just hang a clip lead near the tube and look at it with a scope. That should point out where the 1 gigawatt pulses are coming from.
 

[n2bc]

500pF @ 20KV were a dime a dozen... Every junker TV had one in it, standard stuff in all homebrew transmitters of the era.