Parallel analog switches.

But it's audio, so methinks your suggestion should be fine (unless there's something fancier than mentioned happening).

We could ask you the same...:-)

Cheers, James Arthur

I assume we're all smart enough to clamp Vgs with a zener or two. A bigger hassle is needing higher-than-rail gate drive but that's trivial for us too, just routine.

In practice, 40v drive through a large resistor with a zener Vgs clamp is all that's needed. That's s.e.b. stuff, right?

It's audio, so drive 'em slow and the feedthru is trivial.

Nah, zero volts opens them. Closing them is the bigger nuisance.

Sacrifice some lithium to the tron-gods daily to avoid their wrath, and Bob's yer Uncle.

Cheers, James Arthur

On a sunny day (Tue, 3 Nov 2015 06:42:38 -0800 (PST)) it happened snipped-for-privacy@yahoo.com wrote in :

My xperiences with zeners in MOSFET inoput include mysterious oscillations. As soon as you zener yo ualso feed into the signal path.

Large resistros wil lmake i tslower, couple of nF input C, dothemath,

Says who?

If it is just audio use 2 amops, as Jim already suggested.

How does it taste?

See (representative only... your output stage may vary :-)...

Do the switching before the output stage... switches are then low current. ...Jim Thompson

You're worried about the coupling through a, say 2K resistor or a back-biased zener, at audio frequencies, into a power amp's output?

Or a zener clamp making a hard-switched FET oscillate?

This is just illustrating the limits of text-based communication. We're not talking about R.F., it's audio for Pete's sake.

That's the goal--a slow ramp won't feedthrough.

But let's do the math: 2k onto a .5-ohm power amp's output can't feedthrough more than .03% for an infinitely fast gate-drive step, and much less for a ramped gating pulse. That's pretty good "thump" suppression.

He'd probably want a much smaller FET than one having 2nF Cgs, so he could use a larger resistor, etc.

Cheers, James Arthur

Nicely done, Jim.

Cheers, James Arthur

Thanks, James!

(I've done it before on IC's, so I just needed to put it to "paper" with discrete components ;-) ...Jim Thompson

On a sunny day (Tue, 3 Nov 2015 08:05:08 -0800 (PST)) it happened snipped-for-privacy@yahoo.com wrote in :

You got me there...

I think you make a mistake applying that math, the thing has to drive a 500 mA load, at 30 V, Mr Ohm says that makes a 60 Ohm load. No idea what he drives it with though. If he drives it from a close to zero ohm amp that is better.

I like that! I'll see if I can't fit something like that in, thanks.

Cheers

You are quite welcome! ...Jim Thompson

It's a bit above audio - 10kHz to 100kHz - but not too bad. Making extra rails is something I'd avoid if possible as the whole think has to operate at high temperature so simplicity is best, and in a low noise and cramped environment so switching noise is best avoided.

Cheers

Be sneaky and discrete - design it with power logic FETS.