"High" Voltage Mosfet Amp

Why bother with a 500V supply if you only need 20 volts of output swing?

Cheers

Phil Hobbs

Audiophool syndrome ?:-) ...Jim Thompson

he

I'm confused, the supply is 500V, and the input is 1V with a gain of

George H.

Sure I would like more than a gain of 20. 100 Ideally. If I use drain feedback bias there is no way to get past 30 since this would easily put me over the max V_GS of 30V.

"Jeff Johnson"

..... Phil

Yeah, I need to see a circuit.

ed

me

As a said to Phil, can you post a circuit?

Where's the load? What's the load? Why 500V? (that's dangerous)

George H.

he

You're not expect linearity with DMOS I hope. You might want to see who is left making fets with lateral current flow.

Are you trying to reinvent the Fetron, a high voltage FET with the same pinout as some well known triodes ?

• Nope, the wanker trying to re-invent the wheel.

He wants to make a square version.

BTW:

Fetrons used high voltage j-fets in a metal pack that typically simulated a

..... Phil

...a hopped-up pair of rollerskates...

He's probably trying to drive an electrostatic speaker. Quad Electroacoustics sold one for many years, and it was highly regarded - low frequency response is limited by the physical size of the speakers, and high end freaks filled it in with separate sub-woofers.

One of my ex-colleagues was quality control manager there for a few years, and was deeply impressed by the weird scheme that Quad used to get a minimally conducting film onto the moving membrane - think Gohm's per cm.

Back in the 1960's there were a bunch of articles about building your own electrostatic speakers, and the high voltage amplifiers to drive them. Quad just used step-up transformers, which made a load that many audio-amplifiers didn't like driving.

One of my ex-bosses - Ralph Knowles (now dead) - had a pair of Quad Electrostatics and had enugh money that hi-fi shops would try to sell him hi-end amplifiers to drive them. Most of the amplifiers were amateur-night rubbish.

-- Bill Sloman, Nijmegen

at

any

The Quad ESL63 had rather more than a transformer to drive the electrostatic moving parts

The deformable membrans that moved the air were split into a number of concentric circles, and the voltages that drove each circle came from a passive RLC delay line in which a series of transformers contributed the inductance and some of the capacitance.

It's not surprising that Quad managed to make the input impedance look more of less resistive - once they'd got the delay aspect of the RLC network right, tweaking it to make the input look resistive can't have been all that difficult. There were certainly any number of parameters available for them to adjust.

During my - unsuccessful - job interview at Quad I got to see (but not to hear) the three dimensional analog of the Quad ELS63, which apparently worked, but couldn't be made cheap enough to sell.

-- Bill Sloman, Nijmegen

Different subject, but if you want gigohm conductive glass, you can buy expensive stuff, or use welder's glass.

John

Possibly...

Well, I don't mind a little bit of non-linearity... The datasheet shows it might be pretty bad though ;/

Interesting. I worked at FEI for a while and then afterwards helped develop a product, self-employed, to help extend the average life expectancy of their electron emitters hidden behind the wehnelt.

Jon

Interesting. Is it bulk or surface conduction?

Cheers

Phil Hobbs

Don't know, but I suspect bulk. Somebody told me this, and I thought I'd pass it along as a hint.

John