John’s quite right about the bulb. Filament life goes as some absurdly high negative power of temperature, so dropping it by a factor of two could easily extend its life by thousands of times.
One drawback is that European bulbs are more fragile due to their thinner filaments.
One could think of using a top side depletion NFET with a voltage divider on the gate, plus a smaller-value power resistor from S to ground. And probably a fuse in series in case the FET shorts.
Cheers
Phil Hobbs
And you think you designed them.
And if you understood circuit design you'd have solved the problem another way. You actually copied the circuit from a Hewlett Packard original, and - when pressed - can spell out what's wrong with it.
To the fond eye of the developer. If you could design circuits you could probably have worked out what an emitter-coupled monostable does, and why it got invented some time before it was described in Millman and Taub back in 1956. When presented with an LTSpice simulation of one you fell flat on your face.
There's an infinite range of fools available.
No surprise there.
A thyristor only stays on a long as it is carrying it's holding current.
Capacitors may keep on leaking some current for an appreciable time, but that is eventually going to fall below the holding current.
There's an infinite range of fools available.
No surprise there.
A thyristor only stays on a long as it is carrying it's holding current.
Capacitors may keep on leaking some current for an appreciable time, but that is eventually going to fall below the holding current.
My proposed circuit is much like that:
The drain resistor would be a fusible part. I don't expect a ceramic wirewould resistor to fail shorted.
I'd use 4 or 5 of these on a little discharge board. With LEDs everywhere inside the box.
Thanks.
John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics
The electric chair is better than being hanged/drawn/quartered, or burned alive, as was popular in England.
Dry nitrogen would be a painless death.
John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics
Ah - so getting rid of the heat is not a problem. Add a timeoff delay heat warning indicator to Liz's circuit to allow the heat sink to cool sufficiently after discharge before the warning turns off, if needed.
I doubt I'm telling you something you didn't already think about, but I figured it was worth mentioning as it might be helpful to someone else.
Ed
It's elementary circuit theory, which you should have been taught and I had to read about.
If you discharge a capacitor through a resistor, the voltage decays exponentially. If you put an inductor in series with the resistor the voltage decay is a more complicated function of time. If you chose the resistance and the inductance to create a critically damped circuit, the voltage across the capacitor will eventually decay more rapidly than you'd see with just the resistor.
Try reading about Laplace transforms. That's a pretty hilarious suggestion to direct at you, but you are the butt of the joke.
Oh, RLC circuits are no mystery.
What's funny about the suggestion is the size of the inductor it would need.
Unless the SCR triggers while the power supply is still on.
Envision flames and smoke.
Not copied from anyone. Oscillator gate input to first edge of the 50 MHz output is under 3 ns.
Did HP ever have an instant-start LC oscillator? In what instrument?
They did use delay-line oscillators in some instruments, like the 5370 counter. They kept them running continuously between uses, to tune the tempcos. At trigger time, they would quench them for 75 ns and restart, which wasted a lot of time.
The Spiced version was pretty bad.
Just to add, there's a reason why mechanical relays are still the default choice for a lot of industrial applications.
Your solid state circuit may be quite clever, but total overengineering, suspect reliability and mtbf, for what is potentially a safety critical application. Less is more, etc.
I'd use four or five discharge circuits in parallel. Each with an LED to show that they are working.
Some logic has to drive the relay coils. If the contacts are in series or in parallel, there are hazards either way.
You only need an incandescent bulb, around 100 or 150 W, placed across the capacitor when you want to discharge it.
It's cheap, low tech, high-power tolerant, has non-linear E/I curve in the right direction, and accomplishes the warning light part of the mission.
120vac bulbs might require two of them in series for 200+ volt dc use.
That's not what you posted here.
So what.
Scarcely wasted.
But you couldn't see what it was supposed to do, which did give you a negative impression of the circuit. That generation of two- and three-transistor circuits were certainly crude, but they worked well enough.
Which you haven't worked out. I spent a few minutes last night trying to work out what I could buy off the shelf from element-14 (the Australian branch of Newark) but their web-site has turned cranky in recent months.
The value of the inductance isn't fixed - that and the resistor can be be chosen to get a critically damped LCR, and I figured that I'd start playing with an inductor I could buy.
You might end up with a non-progressively wound air-cored toroid. With a
0.2F capacitor the parallel capacitance of the inductor isn't going to be an issue, and the winding resistance could be your damping resistor. 8kJ is a fair bit of energy, but you can get copper quite hot before it explodes. It wouldn't stay hot for long.
With you as the circuit designer, that might be a risk. Actively clamping the thyristor gate to ground while the power supply was on would eliminate it.
Envision a slow blow fuse. No flames, no smoke.
I haven’t calculated what ballpark inductance might be required but is air core even feasible for that?
Excellent. You have added yet another failure mode to discharging the caps.
The energy we'll be dealing with, a couple of KJ, is roughly the bang from 16 firecrackers, or five .45-caliber bullets.
The fuse would be fun, dumping a couple kilojoules.
John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics
Here in the Netherlands we do not have the death penalty. The advantage is that when proven innocent later you can be released. There are many cases worldwide, some proven innocent after decennia in jail,
Have something to add? Share your thoughts — no account required.
Ask the community — no account required