Improve company productivity with a Business Account.Sign Up

  • Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 2373
  • Last Modified:

Building a simple High Voltage Power Supply out of disposable camera capacitors

I need to know if I can build and use a bank of capacitors to create my own High Voltage Power Supply (I need to produce 10 to 20 kV).  Pulse is ok..I don't need to have it give off a sustained current nor do I expect to have the Voltage setting be variable.  I have a son who is working on a science project for school; it's an ion engine (a "lifter" in point of fact).  My wife helped him build the triangular lifter using wire, tin-foil and balsa sticks.  Now it is time to hook it up to an energy source and I didn't realize that we needed a high-voltage power supply (yes, very short-sighted on our part...not involved early enough).  So, I do have a box full of used disposable cameras (each has a relatively powerful capacitor, which I've pulled out and used on other projects).  The capacitors are each rated at 330wv 80microFarad.  If I can just get the 10 to 20 kV out of them for a half-second on each charge, I be happy.  That's enough to display what he wants to show.  Anyone know how to do this?
David L. Hansen
David L. Hansen
  • 9
  • 9
  • 2
  • +3
4 Solutions
I've attached a simple circuit diagram that I found on the internet ages ago and used myself to make a simple hv psu

Quite dimple operation, basically input voltage is stepped up by a transformer then multiplied using the high voltage multiplier circuit.

This is some info on the multiplier circuit
David L. HansenProgrammer AnalystAuthor Commented:
Thanks.  I'll use the diagram (not sure what the OBT is or VT1 (the switch and fuse I do understand).  Honestly, I'm not even sure what a transformer looks like.

So, I can construct the multiplier using the camera capacitors??  If so, how?  I see diodes being used to connect the capacitors, but what kind of diode?  What kind of fuse (does it matter)?  Mind spelling things out for me a bit more??  Appreciate it.
What Kind of Coding Program is Right for You?

There are many ways to learn to code these days. From coding bootcamps like Flatiron School to online courses to totally free beginner resources. The best way to learn to code depends on many factors, but the most important one is you. See what course is best for you.

I take it that you need VERY little current.
If that is correct you should consider a tesla coil
Below is info found on e-bay

US $9.99
Buy It Now
or Best Offer
look on e-bay yourself for what may be suitable. Also google "tesla coil power supply" and follow the many links you will get.
Dave BaldwinFixer of ProblemsCommented:
Frankly, it all matters when you're dealing with voltages that high because they are lethal.  The "HVM1" (High Voltage Module) is usually a potted insulated assembly.  I believe this is the original article with instructions on building the module:
An alternative to the Tesla coil would be to take the globe off a plasma ball (which after all is a Tesla coil). Might not be high enough voltage to drive the lifter though, they're only about 2-5KV.

As per Dave Baldwin, these things can be lethal. The current's very low but the high voltage can stop your heart beating.

Another even more dangerous option would be to take the back off an old TV and connect to the anode cap. No worries there because the school won't let him do it.
>>  The capacitors are each rated at 330wv 80microFarad.

To get to 10kV you would need to put 30 capacitors in series.
That would give you an absolutely lethal capacitor bank -- 2.7 uF  at  10 kV.

Do you have a link to your lifter design so we can get a better idea of the electrical requirements?

The current you need is very small, probably on the order of microamperes.  So you don't
need or want large capacitors.

The best way to go would be to get a small HV power supply.  The smaller the better.  The Emco Q101 at 50 uA might work.  I just saw a Wikipedia that suggested you can lift 1 gram per watt.

There are also some workable units on available on eBay for $40.

The simplest way to get from 3 kV to 9 or 12 kV would be to make a 3- or 4-stage Marx Generator.  This example shows the sort of capacitors you would want to use, small HV ceramic or mica disk caps.
For better control, you can replace the spark gaps with switches (preferably relay contacts) so you don't have to wait for the spark gaps to break down.
David L. HansenProgrammer AnalystAuthor Commented:
Sure, here is the link:

...and this is what I'm following for the HVPS:

And thanks for the warnings.  I'm aware of the dangers and will not allow any children near the experiment when it is live.  The HVPS will just be off limits, period.  I've worked with high voltage power supplies before, but never built one from scratch.

Since I don't have 30 capacitors to run in series, I'm planning on purchasing one today (or ripping one out of a microwave).

If I space the components farther apart can I avoid using the sunflower seed oil (I saw several designs that don't require any oil buffer)?

By the way, I'm using a riding lawn-mower battery to get the 12Volts to power the HVPS.
Looking at the Wikipedia Ionocraft article and doing some rough calculations:

They claim 15 gF of thrust at 54 kV for one example device.
They also claim efficiencies on the order of 1 gF per watt.

(15 watts)/(54 kV)   gives a nominal current of 278 microamperes.

(54 kV)/(278 uA) gives a nominal load of 192 Megohms.

An RC time constant of 10 seconds would give you several seconds of operation.  The total capacitance required would be
(10 s)/(192 Meg)  =  50 nF  =  0.05 uF

So you might want to make a 20 stage Marx Generator with a 3 kV input.
You would need twenty  1 uF 3 kV capacitors.
Something like this would work, but is probably not appealing at $400.

These smaller, cheaper capacitors would give you one or two seconds of operation for $120

If you can find cheap surplus caps with the right specs to go with the $40 power supply I mentioned earlier, you might get away for less than $100.  But I just took a quick look at eBay and didn't find anything suitable.

If you do proceed, take another look at the Marx Generator

One simple way to control this sort of generator is to build it as shown, but bend the spark gap wires on the individual capacitors so that they touch.  Then you can separate all the gaps with one long strip of mylar or similar material.  
You would charge all the capacitors in parallel (for 50 to 100 time constants).  Then you would disconnect the power supply, leaving the bank charged.  Finally you trigger it by pulling out the mylar strip (with a long string).

Good luck.
My last post does not address anything in your last post.  Let me catch up.

I do understand HV electronics, but I don't have any feel for the electrical requirements for the lifter.  Somewhere between 10 and 50 kV I guess.  No mention of current required anywhere.  And just a single mention of power:  the 1 gf/watt from Wikipedia.
The numbers in my last post could certainly be off by a factor of 10.

The lifter link doesn't give any estimate of the mass or weight of the device.  You should weigh your device for the sake of science.

The HV power supply design you posted mentions 30 kV but doesn't give a current spec.
This design does require an AC input.  It will not run properly from a DC battery.  In a car, the engine turns the distributor and the electronic ignition system turns the DC to the coil on and off rapidly to generate the HV sparks.

Once it is running properly on AC, the output of the supply would also be AC and would require rectification.  This is where the string of diodes from the first link comes in.  You might need 25 1kV diodes, like a 1N4008.  These are cheap.  You can spread them out and avoid the oil.  One good way to handle them is to solder them end to end, and put the whole assembly inside a piece of Tygon tubing.
Dave BaldwinFixer of ProblemsCommented:
I have to recommend against using 25 diodes in series.  While forward voltage is quite predictable, reverse withstanding voltage is not.  Power supplies that use multiple diodes in series typically put small capacitors across each one to keep the reverse voltage approximately equal.  You are better off with high-voltage diodes made for the purpose.
Could a Wimshurst machine plus Leydon jars be enough to make the ion lifter take off?
From the Wikipedia article on the Wimhurst Machine:
     A typical Wimshurst machine can produce sparks that are about a third of the disc's
     diameter in length and several tens of microamperes.

The voltage is probay too high and the current too low.
The Instructables design for the HV power supply leaves a lot of room for improvement.

  1.  Automobile ignition coils are designed to run with 12 VDC, not 120 VAC.

  2.  The coils are designed to operate with short submillisecond pulses.  The light dimmer
       can energize the load for up to one half power line cycle.

  3.  The capacitor is critical.  Too large a value could allow the coil to overheat.  Too small
       will limit the output power.  No idea on the proper value.

  4.  A typical operating frequency for the coil in automobile engine would be
            ( 300 rev per min) * (6 sparks per rev) / (60 sec/min)  =  300 Hz

  5.  This design is limited to the line frequency.  You are not getting any where near
       max output power at 60 Hz.

The simplest standard design for this sort of thing is Flyback Converter.  You can use a potentiometer and a 555 timer to generate the gate control pulses for the MOSFET.
David L. HansenProgrammer AnalystAuthor Commented:
Thanks d-glitch.  I appreciate your efforts here :)  

So, I've purchased the ignition coil and an old microwave (for the capacitor).  I even got some vegetable oil.  I was a little worried about not using the same oil as the author though; I did however, find this bit of information which put my mind at ease:  (see figure on page 4 under dielectric breakdown).

I think I'm ready to assemble the parts.  At this point I'm just worried that you and others have expressed concerns about this particular homemade power supply.  Honestly, I just want my son to be ready for tomorrow's show.  He has the model ready and I don't want to disappoint him.  At the same time, I don't want to embrace a faulty design just because I'm running out of time.  I am ready to assemble it tonight (not sure if I should use the battery or not though).
You can't use a battery unless you have some way to turn the current to the coil on and off very fast.  60 Hz might be good.  300 Hz would be better.

The Dimmer Switch and Capacitor design from Instructables can work:
   1.  Set the Dimmer somewhere in the middle of its range.
   2.  Set up the coil first to just make sparks.  Leave off the diodes and he output capacitor.
   3.  Run it for a second or two at a time, and make sure the coil isn't getting too hot.
        If the coil gets hot, or you trip the AC breaker, then the value of the 600V capacitor  
        may be too big.
   4.  If you don't get any sparks, or only small sparks, then the value of the capacitor
        may be too small.
   5.  Adjust the dimmer to get the longest spark you can.  Measure it.  
        Cold dry air breaks down around 30 kV per inch.
   6. Then connect the diodes and the HV output capacitor and give it a try.


If you get desperate, one way to make a fast mechanical switch is with a screwdriver and saw blade.  This would be for the battery.
   1.  Connect one side of the battery to the coil.
   2.  Clip the other side of the battery to the saw blade.  I'm thinking about a 20 inch rip
        saw with fairly big teeth.
   3.  Connect the other side of the coil to the blade of a screwdriver.
   4.  Run the screwdriver lightly back and forth over the saw blade.  You want it to make  
        contact at each tooth and disconnect between teeth.  
   5.  Do it for a second or two and check that the coil doesn't get too hot.
   6.  Also check that the battery doesn't get too hot.  This sort of connection can discharge
        it pretty quickly.
   7.  You should see sparks between the screwdriver and the say blade.  This is high voltage
        due to interrupting the inductance of the coil.  You can get small shocks even on the
        12V side with this kind of switch.  Don't hold the saw and the screwdriver.
   8.  It it works, remember to recharge the battery and save it for the big show.
Good luck.  And please let us know how things turn out.
David L. HansenProgrammer AnalystAuthor Commented:
You are really going the extra mile here...thanks!  
If I just use the regular AC from an outlet here in the U.S.A (remember the author of that article was British) how do I do that?  Simple question I know, but he never describes how he does this.  Also, he describes "2 core main" wire.  I bought some insulated 16 gauge wire but it doesn't have 2 cores.
I think two core main wire is just a two wire extension cord with the plug on one end and bare wires on the other.  

The connections are made so the current flows through the dimmer, the 600V AC capacitor, and the ignition coil.

The schematic is on the second slide.  The circle on the left side with the squiggle (sine wave) is an AC source, the plug half of the extension cord in this case.

You do seem brave enough, but please be careful.  
     Work on a clean, dry, uncluttered wooden or plastic table.  
     Don't let anyone touch the equipment when you are plugging it in.  
     Don't work alone.  
     Make sure your partner knows how to pull the plug.  
     Know when to quit.
Something like this would work for the power cord.  
You would use the black and white wires, but not the green.
David L. HansenProgrammer AnalystAuthor Commented:
Well, not much to report on...I connected just the ignition coil and the dimmer switch and got spark, but only when the wires came within like a centimeter of each other (There was one decent spark but only because the wires actually touched and even that was pretty unimpressive).  Before this I was concerned that the capacitor would be too much since it is rated at 0.75uF and 2100V (over 3 times what we were shooting for).  The coil was only the slightest bit warm at this point.  So, I was thinking perhaps the 2100V rating would make up for the lack luster performance of the coil.  Not so; after completing the circuit with the capacitor, the results were about the same (maybe even worse).  I actually touched the wires together several times and only got a very slight crackle and just a little visible spark.  I switched poles on the capacitor hoping that would help; however, since it is a non-polarized capacitor it of course made no difference.

I still plan on finishing this but I guess my son will just be showing off his model and explaining ion propulsion at the science fair today..darn it.  His display looks amazing and he can describe the principles quite well, so I think he'll still get a good grade.
David L. HansenProgrammer AnalystAuthor Commented:
Also, I'd really like to award the points now and continue in a follow up question.
Actually, 1 cm spark is pretty good.  Probably about 10 kV.

The reactance of your 0.75 capacitor at 60 Hz is
     1/(2*Pi*60*(0.75x10^-6))  =  3500 ohms

If the capacitor is the dominant part of the load (and I think it is), this limits the current in the ignition coil to
    (120 V)/(3500 ohms)  =  34 milliamps

This is probably limiting the performance of the supply.  You could try increasing the capacitance by a factor of 10 to 20.  Some thing in the range of 6.8 to 15 uF with a voltage rating of at least 300 VAC.
The 2100 V rating on your capacitor is adequate, but a larger voltage rating does not contribute to better performance.
You can't use electrolytic DC capacitors on the AC side of the ignition coil.  So your camera flash capacitors are no good here.

You could try pulsing the ignition coil with the battery directly.
     Connect two wires to the coil.
     Connect one wire to the battery solidly.
     Tap the other wire to the battery by hand.
     That should give you one very good spark per tap on the output of the coil.
     You should see little sparks at the battery terminal as well.
     Be careful not to let the wire weld itself to the battery.

If the battery gives you good sparks by hand, you might want to reconsider the saw and screwdriver again.  This design could benefit from some series resistance.  
     It would help if you could measure the resistance on the input side of the ignition coil.
     There might be a resistor already built in.  I don't know for sure.
     You would like to limit the current in the coil to something like 500 mA.
     Try a 2W resistor of 25 ohms or less (depending on the measurement) to start.
David L. HansenProgrammer AnalystAuthor Commented:
Haven't had a chance to get back to this till now.  Thank you all!

I'm starting a new EE post to continue this project (I'd love to get the lifter off the ground).  I just felt like the points have more than been earned here.

If you'd like to help, please continue at:
David L. HansenProgrammer AnalystAuthor Commented:
David L. HansenProgrammer AnalystAuthor Commented:

Want to help me with the next post? So far I've only had one comment on it, and it was more of a passing thought than an attempt at a solution.  

Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

Join & Write a Comment

Featured Post

A proven path to a career in data science

At Springboard, we know how to get you a job in data science. With Springboard’s Data Science Career Track, you’ll master data science  with a curriculum built by industry experts. You’ll work on real projects, and get 1-on-1 mentorship from a data scientist.

  • 9
  • 9
  • 2
  • +3
Tackle projects and never again get stuck behind a technical roadblock.
Join Now