Self Powered Indicator (end of the story ...)

One neat nuclear device is a beta emitter battery. It generates nanoamps or microamps at hundreds of KV. The problem is, what sort of dc/dc converter accepts a few hundred KV input?

Stacks of avalanche transistors self-synchronize pretty well, so the overvoltage problem becomes a feature instead of a bug.

I'm way too busy to do anything on it this week, unfortunately--it could be pretty good fun, and of course it doesn't have to be very efficient to be a lot better than the dropping resistor method. ;)

Cheers

Phil Hobbs

An electrostatic motor-generator ?

Cheers

Phil Hobbs

Or, a thermoelectric generator! It's just resistor, with a few dozen thermocouples in series, and a LOT of electrical insulation. Gee, even just a resistor is hard to make work at a few hundred KV!

Something like that has been done, actually.

What about using DIACS to fire the capacitor string inverse marx? Very cheap and trigger voltage in 20-30V range, not too low a hold current either? Just an idea.

That's a possibility too, but you'd need a great many more of them.

Cheers

Phil Hobbs

Yeah, I doodled up one such design, it should work. You have to put a little inductive kick (maybe just a ferrite bead) into the discharge-through-an-LED path, so the diac gets turned OFF reliably. The result was a pulsing light that was slow at 30V, faster at 1 kV; it used a resistor/capacitor though, and multi KV resistors are dangerous and odd items.

His spec is 50-3000 volts, which makes the problem interesting. Ignoring the striking voltage problem, the range of NE2 current would be extreme.

Neons don't last at decent brightness anyhow.

I ask again, if the voltage is greater than about 68V,what is wrong with using something a complicated and exotic as a NE-2 bulb?

Well, if the voltage is less that about 68V,the use of a NE-2 is contra-indicated - but one can use the other complex,exotic system: the LED plus resistor.

Here's a somewhat non-portable one:

They transmit +533 kV on one wire, and -533 kV on another wire, something like 0.5km away running parallel.

If one line fails, they can use ground return at reduced capacity.

Best regards, Spehro Pefhany

Fine, if it can stand 5KV.

OK, how about a high voltage PTC resistor?

How about a simple negative resistance NPN circuit to charge a cap enough to pulse the LED for a good duration?

At least it will work at a much lower level.. Of course, Spice does not model that behavior..

Jamie

Around 2KV and up, all you need is to connect one lead of the NE-2 to the supply, and let the other dangle; for lower voltage indication, use an end-fed-zed (suitably insulated for higher voltages. That gives a low-brightness indication as the guesstimate current is

The other ultra-low drain system is even more exotic, and was mentioned in this thread and someone took a shotgun to the idea. It was even mentioned that it could be calibrated; what was not mentioned was that there are voltmeters based on this ultra-modern technology. And yes,it does have the capability* to indicate voltages down to the

• Under mechanically refined conditions.

Excuse me! No salty comments allowed.

ly on John Larkin and Jim Thompson ideas)

t.

supply

right?

Never heard about Cree. You're right that LED luminance isn't very high wit h 2mA/100ms but it could be sufficent in that case (Indicator embedded in a power train railways).

Thanks, Habib.

That's way too practical for SED.

Cheers

Phil Hobbs

Maynard should have noted how I did a current bypass to keep charging current from exceeding "UJT" holding current. The same scheme could be applied to an NE-2.

And Jamie. What can you say about Jamie ?:-} You can model just about anything in Spice if you know your math... however, we all know that "math and simulation are both crutches" >:-}

I believe that, in the far past, I created a good negative-resistance Spice model for the NE-2... I'll look for it. ...Jim Thompson