Just playing w/ a Roller

[Naysayer]

Experimenting w/ shorting unused section of roller inductor before I replace it with some other device.

1st try yesterday:
5.6uH molded choke (1w sized-body) series connected to a 250w 50 ohm RF termination resistor w/ heatsink (TO-220 size). Seemed ok on 10 & 11 and 20 into DL with 5w AM drive, using only 1 plate xfrmr (1800vac, 400mA, FWB). Saw 600-ish watts for few moments on each band 10, 11 and 20 before I tried 40. Heatsink on 250w RF R not even warm. I was feeling optimistic, yeah?

40m always shows higher PO and saw the expected 750-ish but after a 4-5 seconds the molded choke flashed brightly. A bad smell & bit of smoke but choke body still intact. Disassembled picture shows darkened choke. It was a brighter green before experiment. A small bubble on 1 side from heat but still show conductivity on VOM.

2nd try today:
This time I did not bother with 10,11 & 20. Amp was still set from yesterday. I replaced the molded choke with a homemade choke: A 5w 200R mox with 1 layer of 24awg magnet wire the full length. Not much room in front of Roller by the open/unused terminals so to prevent arcing, I wrapped homemade choke in 1 layer of HV glass tape plus 1 layer Kapton. Resistor got hot very fast (5-10 secs) but saw the same 750-800w PO (although this time it was CW not AM). This time the Term R got hot too. Heatsink too hot to touch, fast. Picture shows the wrapped wire discolored from heat.

Next try
I admit that I expected PO to go way down with the unused ends of coil shorted. Maybe a simple wire closing the ends is not the defeat I expected it to be. There will be more on this. The rest of the amp running fine after the rebuild last few months. The Input working great (red LEDs to show what band its set for, hooray!) The Blower motor noticeably smoother with the series connected 70uH power inductor after the triac speed control.
n2eye

Edit: In fact next try will be down to 50R the RF Term is 250w so I'll try larger heatsink.

 

[Naysayer]

The 50R 250w RF resistor shorting the unused turns of roller gets very hot on 40 but much not so on 10-20. So far amp working fine roller stopped arcing. A long key-down on 40 would be too much for the heat. Larger resistor? Resister on a heat sink that is 1 x 1.25" (no fan) mounted high on roller's end ceramic support so solid wire to coil about 1" long. Other end of resistor mounts to roller-wheel bar. I really expected that shorting the unused turns would decrease PO but I'm not seeing that at all.

 

[brandon7861]

Not sure I have anything to offer here since I'm just a low-voltage rookie, but is there a schematic for that amp available?

If the arcing is from the self-resonance of the unused section, I would think a higher value resistor would spoil the Q enough to prevent the arcing. 50Ω may be overkill for that.

If there was a schematic available and one knew the coil dimensions and tap point, maybe some theories could be developed. The network likely acts as a LPF as well, so maybe on 40m, there is some harmonic content that your Q-spoiling resistor is seeing. IDK.

 

[Naysayer]

50 was working rather well so I'll play with that for awhile. Waiting on two 100ohm I'll put in para for 50 at 500w also got a small heatsink w/fan. I want the roller to work it's my last-ditch effort before I replace it with something else.

BTW anyone have a suggestion on how to mount these kilovac switch/relays?

 

[Cable Guy]

Cool relay, only thing I can imagine for mounting is perf board on a bracket.

 

[nomadradio]

That relay is normally mounted with a saddle that slides down over the ceramic end and clamps the flange to the rim of the hole in the chassis deck. The saddle has a screwhole on either side, like the mounting plate for a twist-lock filter cap.

Seems to me the last vacuum relay I used that looked like that said it was for cold switching only, not for switching a live circuit. Used it to place added capacitance in parallel with the plate tune in a SB220. Was part of a conversion to add the 160 meter band to it. I'll never undertake that errand again.

73