dumb!

John Larkin brought next idea :

Ribbon cable to board layout. And the pot has no knob.

--- news://freenews.netfront.net/ - complaints: snipped-for-privacy@netfront.net ---

negative rail? vs Gnd.

Cheers

Isn't it your job to figure out what's wrong?

What's the value of R2? or what does switch SA-1 do?

You let all the smoke out by grounding J12?

We can only guess from your description. Does it have something to do with J12 shield un-GND'ed? And why isn't X7 GND?

Are the resistors wire-wound?

Yes. They're fine.

We use * to indicate an optional, not-loaded part.

SA-1?

You push S1 to discharge the caps in the DUT.

Do the instruments to be used at J6 and J4 have DC-coupled 50R inputs?

If so, that looks like a bit of an instrumentation problem.

We use a floating-input scope with this fixture, a TPS2024, 1M Zin. The "hot" connectors have a warning screened on the test board next to them, the triangle-exclamation thing, just as a reminder.

Nothing to disconnect the source power when discharge is engaged?

People can do that! No harm is done if the discharge button is pushed while the DUT is powered up.

Shouldn't X7 be grounded? Or is that done externally>

Not sure that's kosher with the FETs, seems that will overvoltage that gates unless you provide a current path (like R2) on Q2 that keeps Vgs from rising to 48V, and the abs max spec is probably 10 or 20V. They may not die at 48V, but..

Best regards, Spehro Pefhany

X10 is +48, from a capacitor bank on the DUT, charged by the notorious boost-doubler supply. X7 goes to the drain of a mosfet on the DUT, a biggish pulsed closed-loop current sink. The two power resistors simulate a bar laser load.

Yeah, that's the problem. D-S leakage can pull up the source and blow the gates.

Most mosfets can tolerate numbers like 75 volts on the gate, but these guys can't. I can add an R2, 50K or something, and kluge a resistor across S1, to save the fets. That messes up my standby current measurement a little, but I can allow for the small extra current in my test procedure.

I should have put the switches in the drains. The depletion fets are cool here, though. Cap discharge is linear, not exponential, and the LED stays on almost to ground.

This has implications for certain cascodes. I've got to remember this.

I'm not following how those FETs are wired, or what the pushbutton switch does, or the whole LED with another switch.

The pushbutton discharges the capacitors in the laser driver board. You do that before you remove it from the test fixture. The switch lets you turn off the HV warning LED, so that we can measure the standby power consumption of the DUT. Maybe that should have been a pushbutton too.

The fets are both depletion types, normally on at zero gate voltage.

I'm not seeing what switches the FET on or off. The gate seem to be wired to ground.

Even if the Q2 is on or off, why does it matter if the only real switching is from the pushbotton switch anyways?

it is not you "typical" fet

-Lasse

Why not just place a zener diode from source to drain? That way you can prevent the gate going more negative that 10V or whatever versus source and since there is no need to enhance in this scenario you don't even need a regular diode in series with the zener.