Rf Power supply @ 4kV / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Rf Power supply @ 4kV

previous 1 2 3 4 next

Move Thread

LAN_403

Ash Small

Tue Mar 12 2013, 07:50PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Noah Hoppis wrote ...

it doesn't have to be a linear amp; to be frank I know very little about amplifiers, only that triodes could be used. all I know about amps came from the feedback circiut in my recently constructed 572b VTTC. Really I just need help finding a simple HV amp.

From what I know, if you want to amplify a signal at 4kV, you basically have two choices, a vacuum tube/valve, or you have to series transistors, presumably IGBT's, which are unlikely to operate at 20MHz. (I did some research into this last year, for a magnetron control circuit for my electrostatic linac.)

I've been playing around with a schematic using a microwave oven power supply (4kV) and a single valve/tube. (and an LC tank circuit). I think it could be done, but I'm not quite there yet.

There are plenty of single valve oscillator circuits out there (Hartley, Colpitts, etc). I think using a single valve/tube does make a lot of sense here.

EDIT: I decided to post my 'unfinished schematic' anyway. I'm sure there are problems with it, but maybe someone can point out some mistakes. The capacitor labelled 'C' is supposed to get charged to 4kV by what is basically a microwave oven transformer and doubler circuit, the valve/tube gets switched at 20MHz, and the tank starts to oscillate/resonate. What I've not put in is a feedback circuit to limit voltage in the tank to 4kV (and possibly to synchronize the valve to the tank). Please point out any other problems, as this is a learning experience for me as well. (resonant frequency of the LC tank is obviously 20MHz).

1363119288 3414 FT1630 4kv Oscillator Circuit

Does the tank circuit require a spark gap, for example, like an SGTC circuit?

The charging circuit from an inverter microwave oven could have advantages here, when maintaining the capacitor 'C' at 4kV, but would require modification. There is a thread here somewhere regarding the modifications necessary, from a couple of years ago.

The 'D's' in the cyclotron would form part of the capacitance in the tank circuit.

EDIT: I'd imagine a Hartly or Colpitts oscilator would have advantages, I've just adapted another circuit I'm working on to power a tank circuit, but I'm sure there are problems with the circuit I posted. It can't really be this simple, can it?

Noah Hoppis

Tue Mar 12 2013, 11:49PM

Registered Member #10072 Joined: Thu Feb 14 2013, 05:12AM
Location: seattle wa
Posts: 21

Ok, I added a matching transformer and variable cap (not for synchro operation). I think is should be fine but I'm not sure, the only thing i see though is that the high end of the wave from the RF source could direct the flow to the grid instead of the cathode and cause damage to the tube.

Ash Small

Wed Mar 13 2013, 01:00AM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

I don't have enough information to follow that circuit, Noah.

It possibly might work, I'll wait for others to comment.

The circuit I posted includes the capacitance of the 'D's' as part of the tank capacitance.

I know there are other ways to achieve your goal here, but I'd suggest they involve either a Hartley or Colpitts oscillator.

Noah Hoppis

Wed Mar 13 2013, 03:14AM

Registered Member #10072 Joined: Thu Feb 14 2013, 05:12AM
Location: seattle wa
Posts: 21

Its true that a single tube would be nice, but originally I was designing an amp and though I have no problem with the idea of building in the RF source, I though it would simply be more difficult. Another problem I have been struggling with is that one tube would only supply full wave if it was a class A amp and 15% of an 811a's output is pathetic! so I imagined that I would need a class B in push pull to get both sides of the sine. I still really have no idea what I am going to do so any input would be wonderful.

Sulaiman

Wed Mar 13 2013, 06:14AM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

I imagine the two D parts as a mutual capacitance
and capacitance from each to 'earth'

with no ions being accelerated, an air-core inductor between the two D parts
would form a high-Q (>100) resonant circuit

power can be inserted into this resonant circuit by either
- inductive coupling using a smaller 'primary' coil
- capacitive coupling via two low value high voltage capacitors.

The real power will be required to accelerate the beam
e.g. 1 MeV @ 1 mA = 1 kW (is this correct?)
IF you can work with a small current then
rather than design/build a P.A. (which may become expensive)
could you use 27 MHz instead?
This would allow you to use cheap C.B. amplifiers
the main areas to work on then would only be
. signal source to drive the P.A.
. primary and secondary coils for voltage ratio and resonant frequency.
. many larger CB PAs protect themselves from load problems
some don't but are cheaply replaceable.
I would consider capacitive positive feedback from one of the D parts
to form an oscillator at the natural resonant frequency.

If you go the high power valve route
I suggest a zvs (cfpr) inverter for it's simplicity and efficiency
or any other self-oscillating design.
with a separate signal source drive there is the possibility of driving
an out-of-tune load and bad things would happen.

Steve Conner

Wed Mar 13 2013, 08:33AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

27MHz is in the ballpark. You can sometimes find 13.56MHz amps on EBay too. The process is very inefficient, so if you want 1kW of beam power, you'll need to throw a lot more than 1kW of RF at it.

Fred Niell has built a cyclotron, so get over to his site and read all his papers, to see if he published any details of his RF amp and matching network. Basically Sulaiman is right, you could use a standard ham or CB type circuit, and couple it to a tuned circuit that uses the dees as its capacitor plates.

Ash Small

Wed Mar 13 2013, 10:17AM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Generally the 'D's' do form 'part' of the capacitance of a tuned circuit.

I'll have to do some more reading, but 1mA does sound a lot for beam current for one of these cyclotrons. (there is a wealth of information here:

)

What Noah asked for was a 'simple circuit that oscillates at 20MHz, at 4kV, using the 'D's as part of the capacitance'.

The matching circuits for the RF supplies I've seen all have expensive variable vacuum capacitors, and they are genarally controlled via stepper motors, with suitable drivers and feedback control circuits. I'm guessing here that the resonant frequency will change with the load, and that the load won't be constant. I'm assuming that the load will vary with gas supply to the ion source.

I have a 600W 13.6MHz RF supply in the shed, which I was going to use for the ion source for my accelerator, but decided to use a magnetron based ion source instead, because it was impossible to design a matching circuit without knowing the load (see my first thread on this forum)

I see what you're getting at, use a 1kW (or whatever) standard amp @~20MHz, with a 4kV output transformer.

Maybe with an air core transformer? with the secondary resonant with the capacitance of the 'D's' plus tuning capacitor, and feedback to control voltage on the secondary resonant circuit (Tesla coil style)

Is there a way to just drive a resonant circuit at 20MHz and at 4kV, without using a complicated matching circuit to match an existing RF amp to the the resonant circuit?

I'm just trying to suggest simpler alternatives here. Maybe RF supply and matching circuit is the only way to achieve this.

Shrad

Wed Mar 13 2013, 11:17AM

Registered Member #3215 Joined: Sun Sept 19 2010, 08:42PM
Location:
Posts: 780

maybe a grounded plate you would move back and forth your secondary would act as a variable capacitor and make the output tunable in a usable amount at these frequencies?

Ash Small

Wed Mar 13 2013, 04:21PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

I think I've just realised something. You will have to use a tuned length of, say 75 ohm coax. to connect from the RF supply to the 'D's? there's no other way to do it is there? This means you need a matching circuit? As the load varies, you have to compensate for it too?

This is due to the 20MHz frequency? there's no other way to do it? The layout of the circuit will also be critical due to the frequency?

Am I correct on all the above points?

EDIT: I doubt if this circuit will work either

Can anyone tell me why?

Noah Hoppis

Wed Mar 13 2013, 11:34PM

Registered Member #10072 Joined: Thu Feb 14 2013, 05:12AM
Location: seattle wa
Posts: 21

Oh Yes! it might be good to mention the 20mHZ is not the exact frequency, it was the top end to give an idea of the most RF needed in reality it may only be required to give 1mHZ but the give room is for matching the frequency to different gases. I have also seen fred's site and even talked to him about my design. Still stumbling about the RF though...

previous 1 2 3 4 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob