A question on Reflex Klystrons, how to pulse?

[AdrianPH]

I have a Varian Reflex Klystron type VRX 7213 dated September 1979, it is new in the box, the spec sheet says it will go from 8.5 to 10.5 GHz which is nice as it will cover an amateur band for playing with, only about 100mW.  I appreciate it was probably designed for use on a receiver such as the GHz local oscillator, but I would be interested in trying to pulse it and this is where I get stuck.

It must have the reflector voltage present before the beam voltage is present or it suggests it could be damaged,   So that to me suggests that that has to be a steady supply, can one pulse the Beam/Anode supply and expect it to work.

I know that if using these for production of FM modulation small changes in reflector voltage will cause a shift in carrier frequency, also that the reflector voltage could be driven to provide an AFC loop against another Klystron or oscillator.

So to the point!  Has anyone played with these devices in this way, pulsing the Beam voltage and had any success?  Also looking for a valve circuit to pulse 300 Volts at 30mA?

At present I had in mind a Valve regulator circuit, where I clamp a grid when not wanting output volts, but that leaves me with a rise time on the circuit dependent on the grid capacitance and therefore gives me a CR constant to deal with.

The issues with not knowing!

Adrian

[Diabolical Artificer]

Never played with these, re pulsing one what about using a 555 timer circuit with PWM to drive a hefty OP valve like an 807? Either that or use a circuit out of a scope TB where they convert a sawtooth into sharp pulses by using back to back diodes.not sureif this is possible in the Ghz range though, miller C as you say would play havoc.

To search for old research papers have you tried searching on Gooogle scholar? You might find something there. good luck, Andy.

[ppppenguin]

This is very much not my field but here goes.

I don't think that Adrian is trying to pulse at GHz. At kHz and up to very low MHz, 300V and 30mA isn't a huge amount of power. Well within the range of many valves and power FETs. How much capacitance needs to be driven? If it's a lot, then a cathode/source follower is a good idea. Miller effect can be dealt with by using a cascode stage. Effectively running the main pulse amplifier in grounded grid/gate.

[AdrianPH]

Yes the idea is to pulse at a few KHz, but the pulse width I would be trying to achieve will be approx 1uS (micro-sec). The driver/regulator valve I am thinking of trying is a QV04-7/CV309 just because I have one in the boxes it's max is 50mA Ic, it also was used at 150 MHz.

I do not know if the reflex klystron will like being pulsed which is the big guess.
As one might guess, I am trying to play with making a low power RADAR A scope, a wish I have had for some months and I try to approach small bits of it from time to time and see if I can build a small part of one.

It keeps the brain ticking along.

Adrian

[ppppenguin]

So you'll need a bandwidth of a few MHz to get decent 1MHz pulses. You reall y need to know the capacitance you're driving. Then you can experiment with the pulser driving a dummy load and get it right before touching the klystron.

Using an RF power valve is a good idea, but that's a very old design. If I wanted to get a pulser working quickly I'd do it with a FET. By SMPS standards this is not a demanding role. If I had to use valves I'd probably start with a TV line output valve such as the PL81 or PL36. Plenty of them around and they are rugged.

The other thing that would be very useful is a bench HT PSU.

[AdrianPH]

A lot of it for me is the old technology, and probably the fact I dislike some of the modern tech mainly down to the 40 years in the radio comms industry where basically it went from component fixes to replace the whole radio, I was never a designer just had to fix them, bolt systems together and make things work. But I digress.

I have found a largish pdf on the Vmars archive '2814_AP 1093E_Inter_Services Radar Handbook_Radar Techniques.pdf'. Which will take some reading through it, but it does look promising.
I have a few smaller TV output valves in the boxes of odd's I have, PCL83, 84, 805, but driving the heaters becomes an issue most of my transformers are for 6.3 Volts, the QV04-7 came out of the old aircraft transceiver I got for parts the CV415/4046 already pressed into audio amp service.

The spec sheet shows the input capacitance of the QV is 8 pF. I have not managed to source a HT bench psu, so tend to build as I need it, plans are to get to around 400 Volts unregulated to give a reasonable chance of managing 300 Volts regulated and pulsed, will have to check what transformers I have to hand. I do have an old Unilab Blackburn Klystron power supply 042.871, but think the transformer does not have enough headroom for this job, I could not get any information from Unilab on the unit, just to old.
I certainly think this is a long term project I will dip in and out of.

Adrian

[ppppenguin]

The spec sheet shows the input capacitance of the QV is 8 pF. 

I was referring to the capacitance of the klystron that the valve or FET needs to drive. The input capacitance of the power valve is a secondary problem which will only trouble you when you come to design the low power pulser. I'd still be tempted to use a cascode design, probably with a transistor or FET as the lower device.

[AdrianPH]

On that I have no idea, the spec sheet I have is more a results sheet stating 300 Volts beam Voltage at 30 mA, -158 reflector voltage 10.5 GHz at 103mW.
A true spec sheet I doubt I can ever find, probably as the unit was designed to be on all the time I doubt I will discover that. I had in mind using a resistor load of 10 Kohm at a few watts to see how switching would go before using the Klystron..

I thought about using a thyratron using the discharge pulse to provide the pulse, thinking that a smallish cap and largish resistor in the CR time constant could also mean a very short pulse duration. It could also be used to trigger a scope. I was getting a 81 KHz crystal as part of the project, but looks that things are going awry on that.

Adrian

[BusyBee]

I am a bit late here and have never used one but my first port of call would be the Ham radio VHF / UHF manuals. I presume they would be good for the basics. I have the RSGB VHF-UHF Manual (third edition) which describes circuitry and modulation of a 723 A/B Klyston at 10GHz. I have not looked closely but I think that most of what you need would be there. Let me know if I can help.

It is also likely useful information could be found on the World Radio History site.

Tracy

[AdrianPH]

Hello Tracy;

I have the same manual, the circuits in there, modulate the reflector voltage slightly to give FM modulation. I did some playing with wide band FM in the 80's with the PW exe using a 5mW gun diode and a 60cm dish, think my maximum distance contact was around the 100Km mark from hill top to hill top.

With this project the idea is to keep the reflector voltage constant and pulse the supply, although I need to sit down and figure out how ,as the beam supply is typically grounded as it is connected to the waveguide, so it means a negative cathode and even more negative reflector?

So here is me thinking that if I could get a 99 km contact on the exe could I get a limited 100mW RADAR to work over, say 2 miles still using a 60cm dish, or two one for TX, one for RX. Hence the interest and project.

Adrian