• You can now help support WorldwideDX when you shop on Amazon at no additional cost to you! Simply follow this Shop on Amazon link first and a portion of any purchase is sent to WorldwideDX to help with site costs.
  • Click here to find out how to win free radios from Retevis!

GI46b linear, persistent AC hum on modulation.

nomadradio

Analog Retentive
Apr 3, 2005
7,536
12,397
698
Louisville, KY
www.nomadradio.com
Here's a puzzling design fault we see over and over. Well, once or twice a year anyway. Seems every linear builder using big tubes understands the need for a center-tapped filament supply for tubes that have a simple cathode, directly heated. To keep the AC voltage that lights it up from modulating the amplifier circuit, a tap wire is installed at the center of the transformer winding. The net AC voltage here is zero and it becomes the DC-circuit connection point leading to the negative side of the high-voltage power supply.

Just one problem. The GI46b russky military triode is not built that way. I won't say who built this amplifier. I'm allergic to pi$$ing contests. After all, every amplifier that's built with a design flaw is job security for me. Bad-mouthing someone who builds them is not my job. Too much like biting the hand that feeds me
.
This tube has an indirectly-heated cathode not unlike a 6KD6 or 6LQ6. Those tubes are not ceramic transmitting types. This tube has one side of the heater connected to the indirectly-heated cathode. This makes life easier on the insulation that separates the heater wires from each other and from the cathode.

This amplifier had a persistent "AC hum" on the carrier it amplified. Turns out that using the filament transformer's center tap wire for the tube's DC circuit return modulates about 6 Volts AC onto the tube's grid circuit. Caused a 60 Hz modulation "hum" added to the radio's carrier.

Here's the relay that activates the tube when you key, along with transferring the coax sockets to the input and output circuits. The single black wire leads back to the tube's cathode circuit.

ZVx8k7.jpg


The black wire disappears under the filament transformer, but trust me. It's the center tap wire.

nZkzNU.jpg


The cure is to move this black wire from the transformer to one side of the cathode choke. The side
connected to the tube's cathode and heater both.

QjwcCC.jpg


Have to take one side of the choke loose to identify which of the two tie-strip lugs is the right one.

IyEsnp.jpg


The black wire now goes to the correct side of the cathode choke. The fat yellow bypass cap got replaced by the smaller blue one when one lead wire broke off it. Stuff happens.

kZ1JRC.jpg


mxkxq7.jpg


All in all this is not a badly-built amplifier. Anyone who has ever built a linear will probably have an opinion about how someone else does it.

FLMa9I.jpg


But if you run into a persistent "AC hum modulation" coming from an amplifier with a russky ceramic triode, have a look to see if the builder fell into the center-tap trap.

73
 

Nomad Always great info!(y)

On my own note: What was done to a quality tube socket and how it was installed is just darn right appalling!!! (botched and disgusting) IMHO

Looks like chassis might have been repurposed from a previous build,,maybe??
 
It didn't take a wizard to build that thing... By the way, you can make those Russian tubes hum without a center tap filament transformer. The first time I used them, I assumed the center pin was the cathode and applied bias there. For some reason the foreign data sheet had me entirely focused on the numbers and I never even looked at the picture of the pinout. I'm embarrassed to admit how long it took me to figure out, that picture was worth a thousand words. If you get one that hums, make sure someone didn't do what I did and apply bias on the center pin. The middle ring is your cathode connection on these tubes.
 
  • Like
Reactions: Crusher
ChrisNomad, about how much bias did that 46B need. I’ll have to hook up a temporary plate current meter on this one so I can get 50-100mA of idle current per tube when amp is keyed, but no drive. The one I have here had about 15v of bias, but they used 1A diodes. I’m currently waiting on new tubes. But 15v seems kind of low. Tube arc’ed, took out 20Kv @2A rectifier ( doubler circuit) I made new with 10A diodes. No glitch ( it has one now). Bias not fused, ( it is now) 15A fuses, I changed to 10. One was blown. I had to check and make sure I still had continuity on windings. Plate choke was resonant ( nice burn mark a third of the way up). Amp was ugly. New chimneys with 3/8” holes. They used a countersink to do 1/4” holes.
 
Depending on the plate voltage, you could need close to double the bias voltage to get it where you want. The real trouble is finding one that won't arc. Once it arcs without any glitch resistor, you'll find the dome grid wires have been burned through in at least one place and those sharp ends will now cause the arc to happen at much lower voltages than it took the first time.

When you run into this problem with those Russian surplus tubes, don't waste your time trying to get the amp repaired and working with the same tube. It will just come back. Reject the defective tube for the junk it is and hope you'll get a decent one on your next try.
 
  • Like
Reactions: Crusher
Shockwave, it didn’t help that the shaft coupler for Tune/ plate cap spun without turning shaft on variable capacitor. Also the load cap was a air variable that only went up to 100pf with a 100pf doorknob in parallel with it. The Tune cap by the way only went up to 38pf. Granted you have some anode capacitance to add to that. But in order for it to work right, the circuit had to have a Q of 10. Assuming the plate voltage was 2200vdc. Which I haven’t been able to check yet. Needless to say this amp was Fubar’ed.
 

dxChat
Help Users
  • No one is chatting at the moment.
  • @ Mark Malcomb:
    Hello BJ. Been a long time since I've been on. You doing well? Mark Malcomb
  • @ Naysayer:
    I’m
  • @ kingmudduck:
    Hello to all I have a cobra 138xlr, Looking for the number display for it. try a 4233 and it did not work