1kW MOSFET amplifier

[jgwinner]

Folks:

Maybe I should change my name to "Kilovolt". My secret master plan to rule the world (well, maybe to just watch a good movie) is to build an Electro Static Loudspeaker with a direct drive solid state amp. Or four. Thus, I've been looking at the output designs for the larger amps, not so much for the current / wattage but for the higher drive voltage.

I actually sort of like the idea of no transformer (I had a treadmill that worked that way), but would want an isolation transformer probably.

I actually think getting some MOSFET's to drive the ESL's might be too hard at the kilovolts needed, but I figure if I can double or triple the output voltage, I can have a 25:1 transformer instead of a 50:1 transformer, with hopefully higher bandwidth.

I'm probably going to use Sloan's Opti-MOS design as is - I've read both of his books and looked at a few other designs but based on 20 year old EE knowledge, I like his design.

Anyone know any Kilovolt capable MOSFET's ? If Kilowatt paralleled them, would it give him a Megawatt amp?

[Kilowatt]

2SK2258-01 is rated 1000V, 4A, 100W. 2SK2654-01 is rated 900W, 8A, 150W. They are all N-channel and I bet there's no P-channels that high, so you'd either have to go single ended or quasi-complementary. I'm thinking any bandwidth gained by using a simpler transformer will be lost in the amp with insufficient slew rate. I don't know how you're going to get high enough slew rate, keep it stable, etc, but it's a really neat idea if you can do it. That would be quite an accomplishment.

Djss, I see you're into the kW too. What are you using for a speaker system for your monster?

[jgwinner]

Kilowatt:

Good point on the slew rate. What parts catalog are you looking these MOSFET's up in? What's the manufac? I have a lot of digital databooks but not so good on the discrete components.

I'm building my first few amps with the regular Opti-MOS design, with 100w parts, and would eventually use these for bi-amping the subwoofer. This would give me some time to build 2 transformer based panels and then evetually do a high voltage version and compare to the transformer.

[Kilowatt]

Allied Electronics. Made by Fuji Semiconductor.
http://www.alliedelec.com

[SteveG]

jgwinner,
Why solid state? Many people have driven electrostatics direct with tubes (seems like a natural match up, as far as impedance is concerned). Even if you had to use a transformer, it could be something like a 1:2 or 1:4, instead of that 1:50. I'm probably not telling you anything you didn't know, but I thought it was worth pointing out.
Steve

[jgwinner]

SteveG:

I can't answer that without starting a huge thread but yes, I'm aware of that.

[Kilowatt]

Jgwinner, maybe you should start a new thread. It's an interesting subject. There was Bill F's thread about a super-low-impedance solid state amp, my thread(s) about a fairly high impedance, really high power amp, so you should start one about a super-high-impedance SS amp. Some of the things I'm doing here might be useful to you and vice-versa.

As far as my project (it's my thread), I have come up with some arbitrary component values (actually, I calculated to the best of my limited knowledge, and I still have no CAD software) that would probably work, but there might be room for improvement. I have also altered the protection circuit and removed the sense resistors from all but one set of outputs, as they should all have about the same load on them and I feel there's no need for more resistors. Also, with that protection circuit, I'm not quite sure how to calculate the values of R35 & R44, I want it to start operating at about 6 ohms on the output, so I may need some help there. I would like to achieve a slew rate of about 90 or even 100V/uS, and I don't think it can quite do that now, I might need a little more tail current, but It needs to be stable too. There might also be some question about the way my OPS and VA tie together, it's a little wierd. What do you all think?

I'm still deciding on transistor values, but MPSA42/MPSA92 looked good for small signal. Outputs are IRF240/IRF9240. Any transistor recommendations are welcome. The most common ones can't handle my 160-170V rails.

I'm sure many of you can offer some great insite into the design of my amp.

P.S. I have no idea why the text turned out purple, it's black on my computer. ????????

[Kilowatt]

That's awfully hard to see that, maybe this'll help.

[jgwinner]

Clear as mud!

Did you attach a JPG or a GIF? If it's a JPG, the compression type will smear like that. If you email the original (big one) to me I''ll try and correct it / sharpen it up and size it down. I know some tricks.

There's a small freeware CAD program a couple of others on the forum have suggested, but I don't have it on the laptop I'm on now, or I'd see if I can clean it up.

My intent wasn't to hijack your thread but to note that except for number of output pairs, we may be interested in similar amps. I'll keep an eye on this thread, and you're right, I should start a new one.

[Kilowatt]

It's a GIF, but I fixed it.

I wish I'd posted this one to begin with, I hate to clutter up the thread , but it's too late now. At least there's a clear one up for you all to see. Any word on it? (the design, I mean)