Charging a capacitor to 6kV

how much voltage can you get out of a piezoelectric spark lighter?

--
Return address is VALID!

They are typically 15kV, judging from the spark-gap. I'm not sure they produce enough charge to actually charge a (even small) capacitor though.

used to be the Model T spark coil was the experimenter's tool of choice for such things. where did they all go? anyway, i tried to make HV using a (modern) car ignition coil with a doorbell buzzer in series to provide the breaker points action; no luck. don't know why. it buzzed nice.

On a sunny day (Tue, 1 Apr 2008 08:27:07 -0700 (PDT)) it happened z wrote in :

I had the old spark coil on the output of a tube audio amplifier, about 4 watts (EL84), and then did some feedback, so it oscillated at some kHz. That gave me enough HV for using a TV tube as scope. You need a diode too. Else an old transistor TV horizontal output and some 555 timer. I have something like that for a Helium Neon laser supply too, transformer, transistor, and 555 I think. Maybe you can find one of those (used to be in old laser video players).

You'd need a diode on the coil output in order to charge a cap with one of these. If you just connect the cap to the coil, it will discharge back through the winding following the HV pulse.

Putting a cap (or any load other than a spark gap) on the output of such a coil might load it down and reduce its output voltage.

--
Paul Hovnanian	paul@hovnanian.com
-----------------------------------------------------------------------
Procrastinators: The leaders for tomorrow.

Two possibilities: I have an "electronic fly swatter" from Harbor Freight that cost all of about $3 on sale (and sometimes they're even cheaper), that charges a somewhat higher capacitance to 4kV. It uses a simple blocking oscillator, and it should be easy to voltage- multiply that up to 6kV or higher. It runs on, as I recall, a couple D cells. It would be really simple to remove the little board, add multiplication as needed, and repackage into a small box.

Or, use the transformer out of an LCD display panel backlight. Biggest problem with that is that it seems to usually be the transformer that dies in those supplies, so finding a dead supply to scrap a good transformer from may be frustrating.

You can indeed use an audio transformer to do the job, along with a multiplier. Especially if you can find one designed for use between a small vacuum tube and a speaker, the turns ratio could be reasonable. I'd be cautious letting the secondary go beyond perhaps 500 volts though--it's hard to know just how carefully the winding was done with respect to high voltage insulation.

Cheers, Tom

I actually happen to have one of these from Harbor Freight. It is quoted as going to 1.5kV, and I have measured it to be about that (with a multimeter - not the best test). If I could in fact use an existing circuit (like the fly swatter) I'd be happy. I dissected it, but it seems to be quite complicated.

The primary of the transformer is connected to the positive terminals of the battery, and has a center-tap which I think is being pulsed using a transistor-capacitor combination. The negative end is sent straight to the large capacitor, and the high-voltage (400V) pulses from the transformer through a slightly strange-looking rectifying (and presumably voltage multiplying circuit). I guess I could try cutting the rectifier out the circuit, and putting the output from the transformer through a Villard cascade?

I have several small audio transformers already. I usually use the relative impedances to calculate the step-up. One is 12KOhms to

32Ohms (step up of about 19). Would this be suitable?

Thanks!

Like this?

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Do you need to implement control loops in software?
"Applied Control Theory for Embedded Systems" gives you just what it says.
See details at http://www.wescottdesign.com/actfes/actfes.html

Do you really need a capacitor charged to that value, or do you just need the spark?

How about an ignition coil from a car, motorcycle or small engine? One of those should certainly be good for the spark, and may be good for the capacitor charge if you can rectify the output.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Do you need to implement control loops in software?
"Applied Control Theory for Embedded Systems" gives you just what it says.
See details at http://www.wescottdesign.com/actfes/actfes.html

I didn't trace out the whole circuit of my "fly swatter," but it looked like a standard blocking oscillator to me. I was amazed at how high a voltage mine went to, since the output capacitor is rated at

630 volts. I'm sure it was either 3kV or 4kV; I don't remember for sure (though a search of the group archive should tell you, since I measured it in response to a similar question about a year ago). I measured through a high-value precision resistor, probably 10^9 or 10^10 ohms. That puts much less load on the circuit than I'd expect a multimeter to do. A 20kohm/volt meter would be only 40 megohms at 2kV full scale.

I suppose the inductance of the 32 ohm side would be about 0.1H to

0.2H. How fast do you want to charge the capacitor? You need to put (C*V^2)/2 joules into it, or 18 millijoules to get 6kV across 1nF. Note: a typical audio transformer design isn't optimum for flyback service. Wild guess: a peak current of 30mA in the secondary is OK. Then on each cycle you could store about 0.1 millijoules in the transformer, and the current will ramp up at about 10-20mA/millisecond with about 2.5-3V applied, so the cycle time might be 2-3 milliseconds. Expect to take on the order of a second to charge the cap--unless I missed a factor of 10^3 in my mental arithmetic, or am way off-base with my assumptions about the transformer. I'd measure some transformer parameters, and then SPICE a circuit to see what things looked like. Or--I'd just adapt that fly swatter and save some effort.

Cheers, Tom

You know the "Wheeee" sound the fly swatter makes when you turn it on... what actually causes that? Stress in the coils? It sounds very flyback-like.

Here are the transformers I have:

Wild guess: a peak current of 30mA in the secondary is OK.

I need to learn how to use SPICE.

Am I correct in saying that I can simply add a Villard cascade (1kV diodes and 400V, 0.01uF capacitors) across the oscillating output from the transformer?

I have experience to design Arc Generation. It usuall needs two steps to get to that voltage, which mean you would need two transformers.

I am not sure what capacitor you are using, but the size of the capacitor could be as big as a plate to endure that voltage. and pay attention not to touch it. and don't do it alone, it could be dangerous.

oh, man, you are cool guy.

Need a 600V capacitor across the points; forgot the value (steal one from the car)...

Naturally, the diode needs to be rated higher than the peak-to-peak output transformer voltage...

have:

cycle you could store about 0.1 millijoules in the

I suppose it's the core of the transformer that whines. That's usually the case. "Blocking" oscillator is another term for flyback. A flyback can be driven by a separate oscillator, or it can be self- oscillating.

The transformer data sheet doesn't list the inductances of primary and secondary, but since it's rated 300-3kHz, it may be a little lower inductance than I guessed; my guess was based on about 100Hz low end. Not a big deal. But to make it self-oscillating, it's convenient to have a winding about half the number of turns as compared with the winding driven by the collector. In this case, since you're looking for a high flyback voltage, even lower turns ratio would be good: you don't want to drive the base negative by more than about 5V. If the primary (driven winding) flys back to 25V, say, you'd want a 5:1 ratio to limit the secondary driving the base to 5V.

Yes, you should be able to use a standard multiplier cascade to get to your 6kV. You have to worry about things like diode leakage and diode capacitance when you get to high multiplications, but I don't think

6kV will be difficult. A secondary fly-back voltage of two or three hundred volts shouldn't be too difficult with your little audio transformer, and you may be able to get somewhat more than that. Note that the cascade of 0.01uF caps will represent quite a bit of net capacitance at the 6kV output node--maybe you don't need the separate output cap at all.

SPICE: download LTSpice for free. There are quite a few example circuits, and on Yahoo Groups, there's an LTSpice group that's very helpful. Beyond drawing a schematic and making sure there are values assigned to all the parts, there's not a whole lot you need to know to use it.

Cheers, Tom

have:

cycle you could store about 0.1 millijoules in the

I notice that not a soul in this thread bothered to tell this person that whatever cap he wants to charge, it has to be able to handle the voltage he wants to charge it up to.

So for 6kV charge point, you would want at least a 10kV capacitor, minimum. 12kV or 15kV is better.

Also, you do not have to worry as to whether the source will do the job or not. If you add voltage to these caps, they accumulate a charge. Time is the only factor.

That's why you measure it.

Sometimes you get better help if you disclose the end objective. You can get a "spark" with a chunk of flint.

The solution may be critically dependent on whether you want to build one that fits in a suitcase or a million that fit in a thimble and have to cost less than 3-cents. Solutions for a "high school science project" are less complicated than a science experiment you're sending to Mars. Your use of the "plural" suggests more than one, but that's not much help. The "cap" suggests you need DC, but you can get plenty of "spark" with AC.

Vague questions get vague answers based on what the other guy GUESSES you might be trying to accomplish.

More than once, I've seen people suggest alternate solutions to the original objective that far exceed the performance of the original concept.

My suggestion of the spark lighter was motivation to "think outside the box".

Depending on your input wave shape, you can get something akin to rectification using a spark gap.

Concise, complete expression of the objective is the hardest and most productive part of a project.

--
Return address is VALID!

have:

cycle you could store about 0.1 millijoules in the

I'm using capacitors in series. I have three 2kV poly caps.

I've still got to give them a charge at a high enough voltage.

Not necessarily. If you charge it by sparks, you should be able to get roughly half the peak-to-peak output voltage, since the peaks are so asymmetrical.

Tim

-- Deep Fryer: A very philosophical monk. Website @