Infinite VSWR tolerant power amplifier for 0.1-1MHz

[tesseract]

1 mhz power amplifier

Hey all. A physics professor friend of mine asked me to build an amplifier that could deliver 300Vpp across a purely capacitive load (literally, alternately polarized plates in a vacuum) with a 100kHz to 1MHz sine wave. This didn't seem like a very challenging problem to solve at first but now I think I may have dismissed it a bit too readily!!!

I realize that I could resonate out the load capacitance with a series inductance (detuned to a moderate Q with a series resistance) and greatly relax the amplifier requirements, but the people that will be operating it are a bunch of physics grad students who, you might guess, can't be counted on to properly tune the inductance each time they change frequency, so I'm trying to avoid this as much as is possible or practical. That said, I'm thinking I'll either have to go with a single ended Class A design, or else a high speed (i.e. - current feedback) op-amp driving a voltage gain then a current gain stage (probably with switching MOSFETs for the power devices in each), because these designs don't care what the VSWR of the load is. Am I on the right track here, or is there a practical way to employ a much simpler push-pull amplifier style that is frequently used in RF???

Thanks for any comments!

 

[G4BCH]

understanding vswr

First thought s would be to look at the deflection amplifier of an oscilloscope.
They are designed to drive voltage across two plates in a vacuum, maybe not as large a plates as you need or as many volts, but worth a look.
Peter

 

[tesseract]

ap400b

Thanks for the reply, G4BCH. Unfortunately, the capacitance of this load is a lot higher than found in a CRT 'scope so I need to deliver "real" (er, apparent) power here. Right now they are using an ENI AP400B 400W power amplifier to drive the device* at 250kHz, 440Vpp and the amplifier reads 40W output. This implies a capacitance of approx. 180pF., though that doesn't even remotely match what I measured with my LCR meter (more like 5800pF) so I'm assuming the wattmeter on the AP400B is measuring actual power, rather than apparent... At any rate, I am planning on a power level of around 100W. Obviously, if I have to go with a single-ended Class A amplifier with choke loaded drain that means I'll need to design (and power) the thing for 400W or so. If I can get away with a push-pull design, even if it is biased more Class A than B I can at least save a few watts in the power supply (understanding that when the VSWR=∞, the reflected and forward voltages add together, and in a choke loaded single ended Class A that, I believe, results in the drain seeing 4x the supply voltage... I may be wrong, though, and, frankly, I would love to be).


* - an ion "concentrator" or funnel; it uses RF to focus a stream of ions onto a target substrate.