Power Supply Sequencing Considerations

[W7SOE]

I am thinking ahead to the wiring between decks of my 813 rig.

This will be basically the same as the K1JJ rig with a Millen 90800 exciter.

The PTT scheme will be the same as Tom's, a pair of 50K resistors switched in series with the final and modulator cathodes, by a relay, to put the 813s into cutoffishness.

I would like to keep things as simple as possible, ideally a filament and HV switch on the power supply. 

What about the screen supply on the final?  Should it be switched on with the PTT?  It is my understanding that screen voltage without plate voltage will smoke the screen grids.

I am assuming that I can leave the fixed grid bias on all the time as it is a current flow inhibitor, correct?  Is the loss of exciter signal to the grids of since I am using a fixed bias, not a dropping resistor?

I have a nice power transformer for the Millen 90800 that provides 400V for the plates and also the filament voltage.  Since they are in one unit I cannot, without adding another transformer, fire up the filaments separately.  Am I right in thinking that this is necessary, that just firing up the plate voltage and filaments at the same time is a no no?  Or perhaps it is ok if the exciter is un-keyed?  I am hoping that is the case.  I plan on having a separate "key" switch to enable the output of the Millen as I plan to be able to turn it off and use an external exciter.  (ricebox)

Again, I want to keep things as simple as possible. 

Thanks for the help.

Rich

[WA1GFZ]

Yes fixed bias on all the time. When you place 50K in CT leads everything goes into cutoff. Negative fixed bias just makes it happen quicker. I would bias off the modulators before the fianls so you dissipate any energy on the modulator. Then last open the ant relay. Going into TX I would close the ant realy first then the final and modulator come on. This will avoid sparking relays.

[W7SOE]

Yes fixed bias on all the time. When you place 50K in CT leads everything goes into cutoff. Negative fixed bias just makes it happen quicker. I would bias off the modulators before the fianls so you dissipate any energy on the modulator. Then last open the ant relay. Going into TX I would close the ant realy first then the final and modulator come on. This will avoid sparking relays.

I suppose the same applies to the Millen 90800, I think I can apply the HV and filaments at the same time, at power up, as long as it is un-keyed.

Thanks for the advice on the PTT sequencing.  I could use multiple relays one triggering another to make a sequence.  I would like to avoid the complication of delay circuitry or fancy delayed relays.  (Though I am using one on the final B+ as a soft-start.

Rich

[K1JJ]

Rich,

Good luck getting the new 813's class C rig going, OM. You'll love it.

About the risk of wiping out the screens...

As Frank said, with that cathode keying circuit, it doesn't matter what voltages are on or off - no current flows anywhere until that 50K keying resistor is shorted.  

I have had NO screen protection on my rigs since day one. Never blew out a screen.  The key is to turn on all the voltages and carefully check - THEN key the rig. Even with no plate voltage and screen voltage on, you need some RF drive to put the screens in danger.  Also, put a sensitive fuse (1/8 A or whatever) in series with the screen leads as well as a fuse in the AC primary of the screen supply.

I have occasionally driven the class C rigs with RF and no plate voltage by mistake and popped the screen fuses with no screen damage, so I know it works.

BTW, I use a cool sequencer made by Jay/W1VD. It goes for under $50 and has four relay outputs with a variable delay. This made a big difference in calming down mod iron gap arcing and ant relay arcing, etc. If you have problems with TR key arcing in the rig, consider this sequencer and key the correct items in sequence as Frank said.

T

[KC2ZFA]

BTW, I use a cool sequencer made by Jay/W1VD. It goes for under $50 and has four relay outputs with a variable delay.

I found this on the web (it's $60 now). Nice lead, thanks !

One question: is the rating of the four reed relays it comes with adequate for the coils
of, say, http://amfone.net/Amforum/index.php?action=dlattach;topic=25817.0;attach=22671 relays ?

Peter

[K1JJ]

Peter,

Yes, I would think the contacts will handle those relays. I believe I did a direct interface here with relays about that size.

Does Jay's specs say anything about the contact rating?  You could always send him a msg to be sure. He's listed on this board.

T

[KC2ZFA]

the specs listed are "The reed relay contacts are rated for 10VA at 0.5 amp
max./200V max." but there must be a typo somewhere, no ?

[K1JJ]

It may mean that the contacts at 10VA may be the equivalent of 20V at 1/2 A    or 200 volts at 50ma. That's not very much but may be just enuff to drive those relays you linked above.   

Hmmm... I'd have to look, but if that's the case, maybe I used a second interface relay before going to the big relays here...  I use 24VDC coils, but never measured the current. 


T

[W7SOE]

As Frank said, with that cathode keying circuit, it doesn't matter what voltages are on or off - no current flows anywhere until that 50K keying resistor is shorted.  

I have had NO screen protection on my rigs since day one. Never blew out a screen.  The key is to turn on all the voltages and carefully check - THEN key the rig. Even with no plate voltage and screen voltage on, you need some RF drive to put the screens in danger.  Also, put a sensitive fuse (1/8 A or whatever) in series with the screen leads as well as a fuse in the AC primary of the screen supply.

I have occasionally driven the class C rigs with RF and no plate voltage by mistake and popped the screen fuses with no screen damage, so I know it works.

Thank you Tom.

I will make sure to install a fuse on the screen supply.  So it sounds like I can power on the screen and grid bias voltages at the same time as the plate voltage, with the PTT relays off of course, so the mod and final tubes are cutoff.

How does applying RF excitation without plate voltage damage the tubes (assuming the grid bias is on)?  I trust you, I just want to know why.  ;-)

Does this mean that exciter has to be keyed with the PTT?  (It cannot just be left on?)

Thanks  Rich

[K1JJ]

Rich,

Yes, the exciter needs to be shut off (unkeyed) during RX cuz it will interfere with your reception.. :-)  It will be disconnected from the final grid input through the relay contacts, however.

If the final is keyed and RF is applied, you will draw big (damaging) screen current if there is no plate voltage.  At least in my case, if the finals are keyed without plate voltage and no RF drive, with screen voltage on, then the screen current goes near the rated value, but not excessive. The RF drive is what kills it.

It is a balancing act between how much fixed grid bias you have and how much grid leak bias is there. Be sure to CAREFULLY put the final through its paces with no plate voltage applied to see how the screen current acts. A screen supply Variac is important to have. This will let you know NOT what to do... :-)


As you already know, in practice, always key the final with plate voltage applied to be sure screen current will likely be at a safe level.  Bear in mind that if you Variac up too much screen voltage and also load the amp too lightly, you will probably see damaging screen current even with plate voltage applied.  Tetrodes need special care compared to triodes.  Protection circuits or fast acting fusing will help prevent problems.



T