The AM Forum

In the process of finalizing an amplifier layout and am hung up on how to handle the standby transmit relay configuration. In the past for grounded grid it is common place to use a single relay as shown in figure (A). This is documented and works well with no stability issues at 1 kW and I suspect higher. Example, pair of 3-500's. However, what about grounded cathode? The beam power tubes in use have plate to grid 1 feedback C of 0.2pF if properly shielded or mounted. I suspect that if I wrap the output line back to relay one in (A), that I am going to be in big trouble fighting instability.

I have seen in several cases, folks use TWO SEPARATE relays as in (B) to handle this issue. Now the I/O are physically isolated.

Has anyone been down this path and got zonked using a single relay to handle standby/TX switching in grounded cathode?

No issue, just would like to get it right the first time and supporting two relay units is straightforward.

Tnx, Alan

I took an isolation measure of the relay. If I assume the total feedback C is 1 pF then worse case with a 50 ohm termination, were talking 55 dB cross talk plate to input at 7 MHz. I would like the relay to be better than that.

I made an insertion probe that would allow me to probe the relay contacts with the relay open. The probe alone had isolation from 1 MHz to 30 MHz of -88 dB to -58 dB.

Then inserted the relay. I measure -62 db to -33 dB! At 7 MHz I see -45 dB!

Yikes, I think I may go with 2- relays.
Comments?
Tnx, Alan

I unfortunately can't offer any suggestions or help on this subject, but I think its very interesting.
Keep us posted what you learn, and maybe someone with the knowledge/ experience will chime-in .
good topic.
AG5UM

Hi Alan,

I have always used two relays for that circuit.  

The first reason, as you have discovered, is in/out isolation and potental positive feedback.   It's not a big issue when in GG, but every bit helps when it comes to in/out isolation. Layout, short leads, etc., is what it's about.

But another big reason is due to the potential for a double pole relay to arc across its coil or arms. When running an AM KW, I once had the poles arc together and put blistering RF thru my reciever. It was worse than a lightening strike and totaled a nice Kenwood TS-930.  The Kenwood repair depot avoided it like the Corona once I told them what had happened.

Since then I use vacuum relays for my antenna RF switching.  I use golf-ball-sized Russian 15KV DPDT (or seperate vac relays as discussed)   They cost $100 or more, but I've never had one fail yet. Vac relays usually have excellent internal isolation too.

Also important is to make sure your keyup and unkey sequencing is smooth and avoids arcing. Watch the RX S-meter and listen for audio pops and noise for clues. A sequencer is usually needed for plate modulation, class E, etc. For linears, you can usually get away with a simpler solution.

T

Thank you for the replies.

Two relay's is the path.

Tom, the heads up on reduction of arc over is a good one to note.