LED driver

The obvious way is to charge a supercap off the primary supply, which runs a blinking LED @ >=1Hz that is gated by a divider off the HV. The supercap would be sized to greatly exceed the HV bleed-off time. Blinking the LED makes it a LOT more attention-getting, and extends run-time.

That's better than a lithium cell--lithium cells eventually die.

The more *fun* way might be to make a HV relaxation oscillator into, say, a 4kV flash trigger coil in reverse, making a HV dc-dc switcher.

You could avalanche a BJT as the HV switch, and charge it with an r-c. It's still dreadfully inefficient, but many times more efficient than just a straight r-c.

It's been a while since I tried, but several ms' flash of even a few mA might be a lot better than you think, indoors. Blinking gets you a lot too, maybe a 10x visibility improvement compared to constant illumination.

Wanting to blink all the way down to 48v is a big limitation--is there really that much stored energy? After all, from 5kV, that's 10,000:1.

Another problem is that a relaxation blinker is a VCO -- the blink rate will plummet to 1/oo as the HV discharges.

An alternate might be separating the HV switcher and using *that* to power the LED blinker, avoiding the blink-rate variation.

There's a solution in there somewhere.

Supercap & blinker is probably best, though. Tough, & simple.

Cheers, James Arthur

diodes. two chains hv diodes the diff drives leds

Certainly true of the direct-line models, but I've also seen the ones with internal converters (besides the industrial grade high-ripple electrolytics, some massive inductors -- the kind made with literal gallons of potting... cost, what's that?), same story.

Yup. TVs are known to be nasty until shunted. And with the dielectric absorption of glass, it's not enough to spark just once. I don't think I've seen case studies of TV-related deaths, but the warnings are always provided. The present case should be somewhere around there in power level, it would seem. So I wonder if the regulations include different procedure for that sort of equipment. But still, the point is to, you know, actually look at the regs..

Tim

--
Seven Transistor Labs 
Electrical Engineering Consultation 
Website: http://seventransistorlabs.com

Aha!

I have a marvelous solution, the description of which is too elegant to fit into these margins. If only I could afford wider messages... :^)

Tim

-- Seven Transistor Labs Electrical Engineering Consultation Website:

"John Larkin" wrote in message news: snipped-for-privacy@4ax.com...

How long does it take to charge the supercap?

Where have I hear that before?

I have no NDA. I can say whatever I want.

The power level of the load is irrevelent. I want a source-powered LED indicator that is visible from 5KV o 48 volts. It's a circuit design challenge, whether you approve of my wanting it or not.

What would the number of line phases have to do with the LED circuit? And as I've said, prime power is 24DC.

No. If we do the job, we'll manufacture the gadget in some decent volume.

I never sell engineering just once.

and hence must know how

The problem has been clearly explained.

--

John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    

Precision electronic instrumentation

That won't work down to 48 volts if the bleeder power dissipation is reasonable. Or it would, if there were an aux power supply to run the detector/LED circuit.

--

John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    

Precision electronic instrumentation

Lots of times.

--

John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    

Precision electronic instrumentation

Assume we'll do everything else right [1]

I want the LED circuit. This is an electronic design group, so give it a try.

[1] The customer has an electrical safety group, and they do high voltage things for a living, and they will be involved in the design and testing from beginning to end. But I still want my LED.

--

John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    

Precision electronic instrumentation

That all sounds right.

20 years maybe. It will only run the LED during the brief power-down discharge intervals. But a supercap makes more sense.

I did an LED flash experiment in the similar HV LED indicator thread a while back. I suppose I should look up my own research.

The customer wants 48, because that's officially considered to be safe.

If I can run 2500 volts max, and use a 24 volt "lighting" LED, it almost sort of works with a resistor, or at least a simple flasher circuit.

--

John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    

Precision electronic instrumentation

Looks like my electronics might need maybe 80 watts. 1 watt seems reasonable to light an LED, and I might have two. There are of course thermal and packaging issues.

--

John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    

Precision electronic instrumentation

How about a leakage detector from a mosfet or jfet which is protected from esd damage? The gate has a short "antenna," catching the leakage electrons spraying from the 5kv lines?? Another antenna harvests more electrons to light the led. (snicker - beating you to it. :) jb

Could you see a 190mA pulse around 2uS wide at 1 Hz rep rate?

The pulse has a steep rise and slow tail. Do you think it would be visible with high efficiency LEDs?

At least in Europe, the warning that a layman should not open the back cover of a TV, was well justified, not because the 16-25 kV anode voltage, but because of the AC/DC power supply construction, in which any metal parts, such as the chassis, could carry the full line 220 Vac (or 240 Vac) depending on the polarity of the mains plug.

In TV repair shops, it was required to use isolation 220:220 V transformers, but of course, hobbyists just checked the plug polarity and made dot markings on both the plug and the socket for "correct" polarity.

diodes, hv same path two sets with the difference lighting the leds. it's inline so demonstrative of hv.

Use a piezo-electric actuator or perhaps an electrostatic voltmeter to move a mirror or vane. Instead of an LED, use some phosphorescent material 'charged' by a UV LED run from the 24V supply so that it retains light output after power failure. Zero power taken from the HV, more or less.

Go on, I dare you.

Cheers

--
Syd

Trivial.

It's charged from the primary supply, and low e.s.r. means you can fill it as quickly as you want.

Cheers, James Arthur

Sorry I don't.

If you can bear all the parts of my offering just dropping the cc chain to 200uA will hit your 1W bleeder loss target, the period between led flashes will be longer but still attention grabbing.

piglet