current limiter thingie

Sure. Jfets enhance a bit as the gate goes in the conducting direction, positive in an n-fet. It's not huge. Phemts, with higher transconductance, can enhance well over 2:1.

Its really the Vpinhoff/Vtho bit that does the damage. There are two main parameters, a W/L (gm) factor, and the pinchoff. Pinchoff varies all over the place. gm factor not so much.

-- Kevin Aylward

Can't say I'm overally keen on that technique. For low voltages, its a floating gate. Not wonderful for noise pickup. HF Feedthrough upsetting it not that great either.

As I know you are well aware, its electrically better to switch the gate to the relevant terminal, dependant on the input polarity. A tad more circuitry though...

-- Kevin Aylward

Indeed.

-- Kevin Aylward

We need to make those jfets great again.

It was a demo for current-limiting.

Sure >:-}

Of course. The demo was for the paisan-level... they'd gag and throw-up if I showed a full-fledged protection scheme for an unpowered bus driver (bus signal voltages larger than local VDD).

...Jim Thompson

It's a welcome change. ...Jim Thompson

Odd, that you'd call high capacitance 'terrible'. High capacitance makes them good for a parametric amplifier input. Horses for courses.

Actually the typical protection case would switch the _body_ (P-channel MOS device)... the gate would a separate control mechanism... tri-state, whatever. ...Jim Thompson

In my business, 1 pF is a lot of capacitance.

I'd think the high series resistance of a jfet diode would be bad news for RF. Have you used a jfet as a parametric amp? Why not use a real varicap?

Ah, define 'real'. Capacitance versus voltage IS real for Zener diodes, JFETs, varicaps, photodiodes, etc. Low-input-current parametric amps aren't really the target for varicap duals, the varicap specs for leakage and capacitance near zero-voltage bias are ... sparse to empty.

Not sure about series resistance of "a jfet diode", either. Balancing inputs, how does that enter into figure-of-merit?