Indirectly heated pentode SRS551/RS1002

[N4LTA]

I am building a linear amp for 75 meters for my FSQRP AM transceiver using a pair of the above tubes. The tube is a modern indirectly heated pentode transmitting tube made for VHF portable transmitters.

I suspect that I can turn on power - heater/screen/plate at the same time - as long as the screen comes on with the plate. I assume that the tubes will come on softly as the heaters come on since the tube is indirectly heated. Is this correct  or will I have to turn on the heaters ahead of time? I assume that this is only necessary with a directly heated cathode to stop cathode stripping?

Thanks in advance

Pat
N4LTA

[KK4YY]

Well, I'll take a stab at this one, Pat.

I think you're probably okay to apply all voltages simultaneously with those tubes. Probably.

If you decide to try it, put a high value (maybe 50k?) current limiting resistor in the HV lead for the first test. Make sure proper bias is applied too. If you can monitor grid current, that would be a plus.

If all looks good, and you wire it up to run, be sure not to apply drive to the amp before the filaments are fully warmed-up.

Note: I'd really hate to be wrong and have you loose those tubes, but I can't find a reason that it shouldn't work this way.


Don

[DMOD]

Hi Pat,

It takes about 3 to 5 X the max tube voltage to strip any oxide coating off the cathode and that is only "after" the cathode gets heated by the filament. The reason is the space charge doesn't happen until the cathode starts emitting electrons.

I usually use two switches, one for the filament circuit and the other for B+ activation since there is no tube amplification until "after" the cathode is up to operating temperature.

[N4LTA]

Thanks for the information from both of you guys. The filaments are fed by two separate windings on the main transformer., one for each tube  That makes it difficult to turn the filaments first without switching the high voltage 1200 volt winding. If I had a separate  filament transformer it would  not be difficult.

Pat
N4LTA

[DMOD]

Couldn't one use a single 15A DPST switch to interrrupt the filament current between the transformer and the filaments?


Phil

[w8khk]

Couldn't one use a single 15A DPST switch to interrrupt the filament current between the transformer and the filaments?

Phil

I am not sure I understand how switching the filament current would allow the filaments to be turned on before the plate voltage?

Perhaps a switch is needed that could handle the high voltage transformer secondary.  If a bridge, just one contact.  But if a center-tapped full-wave, or a full-wave bridge with the economy circuit for two voltages, then both ends of the secondary would have to be switched. Am I missing the intent of switching the filament voltage?  Are these instant-heating tubes?  That would be unusual for indirectly-heated cathodes.

[N4LTA]

Problem  is that in order to energize the filament windings, I have to energize the plate and screen winding.. To energize the filaments first - I have to open the high voltage circuit which is not easy. It would be easy to power the high voltage first an then the filaments - not easy to do the opposite.

Pat
N4LTA

[KK4YY]

If there's a mechanism that can cause damage to these types of tubes during the transient period of cathode heating, I don't know what it is. Looking into this, I've found only myths and gut-feelings that damage may occur. Untold millions of consumer devices have been produced over the years that apply all voltages simulations without apparent damage.

It seems that a common failure at turn-on is burning out a filament due to the repeated thermal cycling damage caused each time a cold filament is hit with voltage. A soft-start/step-start may reduce that damage and could be used in this application, if desired. A series resistor in the transformer primary would accomplish this, ramping up all voltages slowly. A time delay relay, or even a simple switch, could be used to bypass that resistor.

There are solutions... but is there really a problem?

[DMOD]

I am building a linear amp for 75 meters for my FSQRP AM transceiver using a pair of the above tubes. The tube is a modern indirectly heated pentode transmitting tube made for VHF portable transmitters.

...I assume that the tubes will come on softly as the heaters come on since the tube is indirectly heated. Is this correct  or will I have to turn on the heaters ahead of time? I assume that this is only necessary with a directly heated cathode to stop cathode stripping?

Thanks in advance

Pat
N4LTA

Hi Pat, I think YY's suggestion of a soft-start makes sense in terms of filament and PS turn-on current surges since you don't have a separate filament transformer.

Phil 

[kb2vxa]

Better late than never since I see danger ahead. Receivers go on all at once because the plate voltage isn't all that high without load, but with a 1200V plate voltage under load that transmitter threw a red flag. Since it's much higher under no load conditions while the cathode is warming up taxing filter caps and HV components I'd have separate filament and plate switches. Taking my best guess at the HV secondary having CT to ground as in a full wave rectifier, a switch or relay in that lead can be the plate switch. Just a thought.......

[KD6VXI]

A 20A 3 phase contactor in series with the hv output would do it.

Wire each phase in series and if you use a 480 contactor you now have over 1200v AC interrupt capability.

Even a 2 pole 480 contactor would do it.

Or a vac relay.  I've bought several glass vac relays from Poland off ebay lately.  All good, and about 35 bucks a piece in any decent qty.  Last time I bought 5 (and broke 2......).

Something to ponder.

Also, some of the GG Cb amps with the 6lf6 would smack the plate with up to 1200 volts at turn in, but the cathode was open in these cases.  Not good engineering practice, but the did it.

--Shane
KD6VXI

[KK4YY]

A 20A 3 phase contactor in series with the hv output would do it.

Wire each phase in series and if you use a 480 contactor you now have over 1200v AC interrupt capability.

Even a 2 pole 480 contactor would do it.

Or a vac relay.  I've bought several glass vac relays from Poland off ebay lately.  All good, and about 35 bucks a piece in any decent qty.  Last time I bought 5 (and broke 2......).

Something to ponder.

Also, some of the GG Cb amps with the 6lf6 would smack the plate with up to 1200 volts at turn in, but the cathode was open in these cases.  Not good engineering practice, but the did it.

--Shane
KD6VXI

Breaking a contact (switch, fuse, etc.) with DC is different than with AC. AC passes through zero volts twice per cycle and won't sustain an arc like DC will. I'd be careful about using AC rated circuit breakers, fuses, and switches with DC. Just sayin'.

[Opcom]

If the plate and filament of a transmitting or high power tube are energized together, then there should be some step to prevent conduction until the cathode is hot. Just killing the screen may still allow some current. It is well worth a cheapest possible time delay relay to delay the plate voltage for the time specified in the data sheet, or if not, then some prudent time.

The OP's tube data sheets don't seem to include a warning or warm up time. Perhaps it is assumed.

I believe "life-shortening damage" can happen in higher-power and higher current tubes and that it does not happen right away when cold, but risky time starts as the cathode starts to emit, and ends when it is close to fully hot.

I didn't find a scholarly reference but only warnings in data sheets and mention of the problem in patents. I don't know if 'cathode stripping' is the correct term for it but that is what people use. Damage to a cathode can also occur when ions bombard it. Ions can be produced by electrons hitting gas molecules or internal parts of the tube. If the ion is positive it will strike the cathode causing incremental damage. It is massive compared to the electron. The increment of damage seems small on the face of it, no doubt the thickness and chemical makeup of the cathode is important.

Damage being caused might be seen as sparking when a 5U4 is energized cold with a high current load such as a resistor already applied -this is not a normal situation.
In normal situations the 5U4 heats up before the load's receving-tube heaters warm up their indirectly heated cathodes.

Perhaps none of this is worth worrying about but some tube data sheets give warnings:

Tung-Sol believes there is risk of life-shortening unless the tube is prevented from conducting current until it is warmed up for the specified time.
see page 7 of the attachment.

RCA TT5, page 71 -filament heating time. yes it is for Transmitting rectifiers, including indirectly heated, but what evidence is there to dismiss the concern in other transmitting tubes without finding a real paper on the topic?

STC manual for 4X150 and 4X250 gives a 30 second heating time when used as high current regulator on page 58.

The presumed damage may be more likely whenever the current exceeds the cathode's ability to easily deliver electrons, more or less regardless of the cathode's temperature.
https://patentimages.storage.googleapis.com/73/c3/80/59a0f3a33094ad/US5130606.pdf

https://patentimages.storage.googleapis.com/df/90/30/fe2c4d76519a1d/US1897556.pdf

https://books.google.com/books?id=HWKBAwAAQBAJ&pg=PA17#v=onepage&q&f=false   page 17

[KK4YY]

A mosfet with a sufficient voltage rating could be used in the HV lead. It can act as a voltage regulator and the circuit could be designed to very slowly charge up to the regulated voltage level. That provides the time delay for the filaments to warm and, as a bonus, gives you regulated HV. This is probably less expensive than buying a separate filament transformer and smaller too.

[N4LTA]

It should work. I have something similar in my PWM power supply. Would probably have to put two MOSFETS in series. My screen regulator on the SR551 amp has an input of 600 volts with a single MOSFET. Shouldn't need a large heat sink if the transition happens fairly fast - a second or two.

This may be a  job for Phil?

Pat

[KK4YY]

This is what I had in mind. But, it might be more expensive than a filament transformer! And, it might not even work!!!


Don

[N4LTA]

The data sheets say that the SiC MOSFETS don't work well in linear mode but this application might work. I had thought about using them as a linear amp.

Pat
N4LTA

[KK4YY]

This Si device STW9N150, is more expensive than the SiC device C2M1000170D, but has built-in gate protection. There are others to choose from. Pay your money, take your choice.

I can't find a way around having a lot of C in this circuit, so a separate filament transformer may be the more economical (and easier) choice for your application.

[N4LTA]

Yep - being a Hammond dealer allows me to get filament transformers at a real good price.


Pat
N4LTA