If it were about the size of a truck maybe.
You can estimate the inductance in your head. To discharge 0.2F in, say 100 seconds, you need 500 ohms. That would of course be the tau of an exponential decay. Adding an inductor would crisp that up.
R*C = 100 seconds so we want L/R to be in that ballpark.
So L is around 50,000 H.
Check Digikey for that.
John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics
You want to discharge it a lot faster than that. 100 seconds is a long time to have a potentially dangerous voltage hanging around.
Probably a lot less.
Probably not the right place to look. Broad line distributors don't like stocking bulky components.
By eliminating one?
But not available in a way that would let it do the damage that either can do, unless you exercise your ingenuity even more perversely than usual.
Both need something to burn. You might volatilise a bit of wire and aluminium can burn in air, but aluminium wire is hard to solder.
It's more electrical than electronic engineering.
Critical damping happens when the expression for the impedance of the series RLC has two identical roots. This happens when L = R^2C/4, near enough, so L should be 12.5 kH.
With the initial voltage 200V, it will need to briefly store just short of 600 J. That doesn't look like a practical solution.
I like the incandescent lamp suggestion. I should spice that.
Jeroen Belleman
Wrong.
<snip>5H looks more like a sensible value. I dug out Grover, and that looks like an 11 cm OD air-cored toroid with a 2cm OD winding and 63 turns of wire. That's a single layer of 0.4mm OD wire.
If you did a two layer non-progressive winding you could use heavier wire but I don't think you'd need to for a 1 second current pulse. The damping resistor for critical damping is 10R. It would be hard to get that much resistance in the inductor. It's about 50cm of wire, and
0.4 mm OD copper wire has a resistance of about 0.1R per metre.I could work out the volume of wire and from that it's heat capacity, but it seems scarcely worth the effort for one of John Larkin's brain-farts.
5 henries? That's crazy. Those numbers would make microHenries.
You aren't having much luck finding work as an engineer. It's obvious why.
You'd have a promising career as an insult comedian, if you had a sense of humor.
John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics
100 seconds is a lot too long from a safety point of view
If you start at the wrong end.
You can vary both L and R. A much shorter time constant means a much smaller inductor
I dug out my copy of Grover and thought about a 5H air-cored inductor.
5H - a 1 sec time constant - would be practicable - but big. You'd need a great deal of copper wire to make it work.It might be worth thinking about an iron-cored inductor. We are looking at a fairly slow event so the current induced in the iron would be just one more dissipation mode.
0.5H might work. The time constant of 0.32 sec means that your 4kJ looks like 13kW while it is dissipating, but it would be being dissipated in what could be a fairly substantial resistor which wouldn't warm up much and would have time to cool off.Why not? Air-cored coils can store a lot of energy. If you put a lot of current through the turns the mechanical forces eventually rip them apart, but that's a very different regime.
Correct. It took me a few minutes to for the penny to drop and I deleted the post, but not fast enough.
Grover's formulas generate microHenries, which I knew, but managed to forget for a few moments
If you make it to 83, you may run into the same problem.
Your judgement in such matters isn't great.
In fact, judging from my own experience of winding small transformers, and air cored coils, by hand, for radio and mother work, such an aircored coil would look more like microhenries.
An inductance slug to handle that sort of power, would probably weigh
50lbs, perhaps much more.
It did when I rechecked the calculation. Grover's formulas generate inductances in microHenries, which I sort of knew, but it slipped my mind for a minute or two. I deleted the post when I woke up to my mistake, but I didn't do it fast enough
It's not the power that the problem, it's just the volume of space that you have to enclose.
As I've posted since, one might be able to get 0.5H in an air-cored toroid, but it would have to be bulky and contain a lot of wire to keep the series resistance low enough to do the job.
Since it only has to handle a slow transient pulse, I'm wondering whether a copper-wound iron-cored toroid might do the job. If you discharged the capacitor in about a third of second the eddy current induced in an iron core would be just another dissipation mechanism.
<snip>
Fortunately, I can't be fired.
We were thinking of doing the opposite: start with a power resistor and kick in a fet near the end of the discharge, to kill the exponential tail.
We'd use several and leave them on all the time, which avoids a bunch of logic with its own failure modes.
Because an air-core 12.5 kH inductor is *big*.
Jeroen Belleman
Maybe a boost converter that regulates to a constant current until it runs out of steam? It would have to run it down to safe levels.
The load could be an old-school 500W halogen construction lamp for the European market (230V) which wouldn't get to full brightness but should still be able to light up half the road for a while Griswold-style.
But your firm can be sued for incompetence, or just go bust.
As I managed to work out, after an unfortunate slip of the mind.
12.5KH is much bigger inductance than you would want or need. 100sec to discharge a capacitor is much too long. 5H and and 1sec makes much more sense but the air-cored inductor would still be impractically large.A carbonyl iron core might work
A bigger core could accommodate more turns of thicker wire, but that supplier doesn't do one.
An iron tape core would offer more nH per root turn. They are commercially available and in larger sizes
<xxxx>
It's interesting. 4kJ would run my ordinary 3kW kettle for about 1.3 seconds and you'd not notice the water temperature rise, so therefore
4kJ is nothing much.But 4kJ would throw a 5kg bowling ball 160m (45 degrees, flat ground, no air, natch) which is a lot.
Under-running a halogen lamp considerably shortens its life.
Depends how much it's under-run, obviously. I've used 12V halogens as interstitial heaters in a string of NaNiCl cells glowing up to red hot and never seen a failure.
In any case, John's application is, I'm guessing, pretty low duty cycle.
The reduced lifetime with under-running is only going to be a problem in the intermediate range where the filament is hot enough for tungsten to evaporate but not hot enough for the tungsten halide to be decomposed and redeposited on the filament. Very low duty cycles or very low filament temperatures should not be a problem. John
Have something to add? Share your thoughts — no account required.
Ask the community — no account required