Self powered High voltage indicator

What do you want the indicator to do?

How much current can you steal?

AC or DC high voltage? What is the operating voltage range?

Can it have a battery inside? A lithium cell might last 20 years.

Indicate that a Voltage > 50V is present ... assuming a 3KV is the rated voltage is the equipement

As low as possible, let say 1 or 2mA

DC. range 35V .. 3KV rated ... 5K surge (20ms)

No battery. self powered or otherwise the guys are laughing on me ... !

Please John take a look here after LTSpice netlist (6W on each MOSFET because of excessive drawn current ... but this was my first idea) :

Version 4 SHEET 1 1188 852 WIRE 320 -528 96 -528 WIRE 448 -528 320 -528 WIRE 320 -512 320 -528 WIRE 96 -496 96 -528 WIRE 448 -480 448 -528 WIRE 320 -400 320 -432 WIRE 400 -400 320 -400 WIRE 96 -368 96 -416 WIRE 320 -336 320 -400 WIRE 448 -288 448 -384 WIRE 448 -288 384 -288 WIRE 448 -272 448 -288 WIRE 320 -176 320 -240 WIRE 448 -176 448 -192 WIRE 448 -176 320 -176 WIRE 320 -128 320 -176 WIRE 448 -112 448 -176 WIRE 320 -32 320 -48 WIRE 400 -32 320 -32 WIRE 320 32 320 -32 WIRE 448 80 448 -16 WIRE 448 80 384 80 WIRE 448 96 448 80 WIRE 320 192 320 128 WIRE 448 192 448 176 WIRE 448 192 320 192 WIRE 320 208 320 192 WIRE 448 240 448 192 WIRE 320 320 320 288 WIRE 400 320 320 320 WIRE 320 384 320 320 WIRE 448 432 448 336 WIRE 448 432 384 432 WIRE 448 448 448 432 WIRE 320 544 320 480 WIRE 384 544 320 544 WIRE 448 544 448 528 WIRE 448 544 384 544 WIRE 384 560 384 544 WIRE 384 560 304 560 WIRE 528 560 384 560 WIRE 704 560 528 560 WIRE 528 592 528 560 WIRE 704 592 704 560 WIRE 304 640 304 560 WIRE 704 704 704 672 WIRE 304 816 304 704 WIRE 384 816 304 816 WIRE 528 816 528 656 WIRE 528 816 384 816 WIRE 704 816 704 784 WIRE 384 832 384 816 FLAG 96 -368 0 FLAG 384 832 0 FLAG 704 816 0 SYMBOL nmos 400 -112 R0 SYMATTR InstName M1 SYMATTR Value STW11NM80 SYMBOL res 432 80 R0 SYMATTR InstName R1 SYMATTR Value 100 SYMBOL npn 384 32 M0 SYMATTR InstName Q1 SYMATTR Value 2N3904 SYMBOL nmos 400 240 R0 SYMATTR InstName M2 SYMATTR Value STW11NM80 SYMBOL res 432 432 R0 SYMATTR InstName R2 SYMATTR Value 100 SYMBOL npn 384 384 M0 SYMATTR InstName Q2 SYMATTR Value 2N3904 SYMBOL res 304 192 R0 SYMATTR InstName R3 SYMATTR Value 5meg SYMBOL res 304 -144 R0 SYMATTR InstName R4 SYMATTR Value 5meg SYMBOL nmos 400 -480 R0 SYMATTR InstName M3 SYMATTR Value STW11NM80 SYMBOL res 432 -288 R0 SYMATTR InstName R5 SYMATTR Value 100 SYMBOL npn 384 -336 M0 SYMATTR InstName Q3 SYMATTR Value 2N3904 SYMBOL res 304 -528 R0 SYMATTR InstName R6 SYMATTR Value 5meg SYMBOL voltage 96 -512 R0 WINDOW 123 0 0 Left 2 WINDOW 39 0 0 Left 2 SYMATTR InstName V1 SYMATTR Value 3000 SYMBOL cap 288 640 R0 SYMATTR InstName C1

SYMBOL res 688 576 R0 SYMATTR InstName R8 SYMATTR Value 1k SYMBOL zener 544 656 R180 WINDOW 0 24 64 Left 2 WINDOW 3 24 0 Left 2 SYMATTR InstName D1 SYMATTR Value BZX84C15L SYMATTR Description Diode SYMATTR Type diode SYMBOL voltage 704 688 R0 WINDOW 123 0 0 Left 2 WINDOW 39 0 0 Left 2 SYMATTR InstName V2 SYMATTR Value 2.2 TEXT 8 640 Left 2 !.dc V1 0 3000 1 TEXT 744 744 Left 2 ;LED

Is its only function to light the LED?

Yes ... and no. I did not mention firstly that few other things are mandatory but as far i know it low power electronics (opamps ...etc.)

Habib.

Something a little simpler, like this, ought to work....

It could maybe use a little more protections, like some source-gate zeners, depending on how radically the HV might change. Two more of the 15 volt guys maybe.

Use more stages if you can't find a suitable 1KV mosfet.

Good point indeed. I saw somewhere in the spec that Vin could change very rapidly (something like dv/dt = 500V/0.1ms) i think it's a bit problemeatic for MOSFETs.

Best regards, Habib.

OK, this is a little better:

The series LEDs improve efficiency and get the mosfet power dissipation down.

The three (or more!) top resistors are sacrificial, in case the fets short. The cap reduces dv/dt into the fet stack. The zeners are 8 volts roughly; 8 or so volt zeners work well at low currents.

Of course, you'd have to find a way to use multiple LEDs in the display, or find an illumination-type multi-chip LED, 12 volts maybe.

Depletion fets, like LND150s, would be interesting, especially at lower input voltages. The gate protection would have to change, but the lower zener can go.

Or a hybrid, with just the bottom fet being depletion. That's cute.

Interesting problem.

But will it work if you pot it?

Just kidding, just kidding. Regards

Does your circuit light the LED's for a voltage just greater than 50V? Which was how the OP spec 'd it. ...Jim Thompson

The range he specified was 35V to 5KV.

Does it light up the LED's at 50V?

From earlier in the thread...

"> What do you want the indicator to do? Indicate that a Voltage > 50V is present ... assuming a 3KV is the rated voltage is the equipement" ...Jim Thompson

How about raising your lower detect limit and just use a neon bulb and resistor?

regards, al

Do a search for an electrostatice 'leaf' indicator, or such. Can't remember the name now. We used to make our own using gold leaf

It's self powered as the leaf tilts up proportional to the voltage, and with calibration can geet fairly 'accurate'

You didn't say, whether this is 0.01%, or 1%, or 10%, or ???

leaf electroscope.

What is wrong with a neon bulb, AKA NE-2; hold it near the HV wire and the bulb will light up. The illumination will indicate AC(both elements light), DC positive or DC negative (only one element light - which one is polarity sensitive).

Hell, if the voltage was high enough, connecting one lead of the NE-2 to the HV wire, and letting the other lead dangle for corona, will do the job (and show if AC or DC polarity).

Jim,

Yes it might light the LED for Vin >= 50, although the schematic here in is a bit simplistic, the main problem is the thermic and dv/dt on stacked MOSFETs,

BTW you are understanding the topic what would you do in such a case ?

Best regards

Thanks for the name, could NOT remember it. I remember accidently applying way too much voltage only to see the leaf be 'blown' away. Those things are NOT easy to repair.

yeah, don't even know what you're referring to re 35V [guess there's some kind of 'new' constraint, in addition to lighting an indicator!

I responded to the original request, which made no mention of 35V, nor accuracy, nor lighting something, nor .... To me, a semi-mechanical solution was an interesting suggestion, but evidently not.

AFTER responding, I find there's a trail of 'extra' details. As, in oh by the way, I need to also...

Although self powered is interesting, I'll bet after everyone here solves the problem, there'll be another constraint, like how much impedance appears to the HV that one is measuring. 100MEG? 1G?

IIRC you cited 500V/0.1ms as "fast"? That's only 5V/us! There are _many_ OpAmps faster than that.

As for a solution... a >50V detector buffered to withstand a 3kV source?

"1 or 2mA" isn't going to light the LED very brightly... unless you had it blink... like in that "Miser" chip from National (Bob Hirschfeld, RIP).

Back when I had to measure 36kV (when I was a student at MIT), I used a sinusoidally-variable capacitor... bottom plate moving, and top plate connected to the 36kV...

I = V*dC/dt ;-)

I'll think on it. ...Jim Thompson