In this Instructable you will learn how to make a High Voltage High Frequency power supply in 5 minutes and for less than $20.

All you need is a compact fluorescent light (CFL) and a flyback transformer.

Flyback transformers are found in TVs and CRT monitors. They make the high voltage, high frequency current necessary to trace the electron beam across the screen. They are small and compact, and you can take them out from an old computer monitor or TV.

CFLs are very popular high efficiency fluorescent lights. They are similar to their ancestor the fluorescent light tubes but use electronic ballasts instead of the big and heavy ballasts in the old technology.

The electronic ballast works by generating high frequency currents that are fed to a tiny high frequency transformer that boost the voltage and run the fluorescent tube. It is the high frequency that makes the assembly compact.

The electronic ballast generates less than 1000 volts. But by replacing the fluorescent bulb of the CFL with a flyback transformer, spectacular voltages can be achieved.

Please visit my latest instructable, how to build an 2.4Ghz yagi antenna for long range WiFi or surveillance cameras:

Step 1: Some Info on CFLs

Picture of Some Info on CFLs

CFLs can come in a variety of shapes and sizes. Generally the bigger the wattage the larger the voltage output. For this Instructable I got a 65 Watts light bulb.

Most CFLs have a similar circuit topology. All of them have 4 wires coming out of them. The wires are in pairs, and each pair connects to a filament inside the light bulb.

The CFLs I came across have the high voltage on the outer wires. You only need to connect the outer wires to the primary coil of the flyback transformer.

You will find a comprehensive description of CFL circuits on this page

Step 2: Some Info About Flyback Transformers

Picture of Some Info About Flyback Transformers

Flyback transformers come in all different shapes and sizes. Pick a big one.

The challenge with the flyback transformer is to find 3 pins out of 10 to 20 pins. One pin will be the high voltage ground the other two pins will be that of the primary coil that will connect to the CFL's electronic board.

If you can get the schematic of the flyback transformer that will save you time. However you can figure out the pins by following the instructions here.

Danger - if you are going to get the flyback from a TV or CRT you need to discharge it. It can hold a dangerous charge even days after the TV or CRT is turned off (see picture for details).

Step 3: The Finished Setup

Picture of The Finished Setup

This is how the finished high voltage supply looks like.

Remember, this is a DC supply. The output from the thick wire is positive. In TVs and CRTs this high voltage output drives the negative electrons from the filament to the screen.

If you need AC high voltage, you have to remove the built-in diode or find an old flyback transformer that does not have a built-in diode.

Step 4: Troubleshooting

Picture of Troubleshooting

The first time I build the circuit, it worked immediately. I used a 26 watt CFL.

Then I decided to get a bigger CFL and I build it exactly like the first circuit. It didn't work. I was disappointed. I thought that the CFL electronics were shot.

But when I reconnected the fluorescent tube to the four wires, the CFL worked again. I realized that this type of CFL circuit needed to "sense" the filaments in order to operate. Remember, I was only using the outer wires and leaving the two inner wires alone.

So I put a resistor across the outer wire and the inner wire. The circuit worked! But within seconds the resistor was in flames.

So I decided to use a capacitor in place of resistor. The capacitor allows AC currents but blocks DC while a resistor allow both AC and DC currents to flow through it. Also a capacitor does not heat up because it provides a low resistance path for AC currents.

The capacitor worked great! The arcs produced were very big and thick.

So in summary there two things that can go wrong:

1. You wired it wrong, either on the CFL side or the flyback side.
2. The CFL electronics needs to sense the filament and you can use a capacitor as a substitute.

Use a high voltage rated capacitor. Mine was 400V and I got it from another CFL circuit.

While troubleshooting, be very careful, you are dealing with very high voltages and high currents.

When soldering, disconnect the circuit from the power outlet.

Step 5: Disclaimer

Picture of Disclaimer

The circuits in this Instructable use very high voltages and currents.

These currents and voltages are deadly! You can easily hurt yourself, as well. Build this circuit at your own risk.

This type of high frequency high voltage current is used in surgical cauterizers. So if you get shocked you will burn yourself and cut your flesh. There is also a considerable fire hazard from the circuit.

Use the Nikolai Tesla's safety techniques when working with high voltages:

1. Only use one hand (put your other hand on your lap or pocket)
2. Wear insulating shoes
3. Use a dead man stick or insulated pliers when touching or manipulating the circuit.
4. Use a power bar with a thermal fuse rather than sticking the circuit directly in the socket. This will limit the current that will go through your body.
5. When soldering, disconnect the circuit from the power outlet.

Generally, in electricity it is the the current that kills. if the currents are low there is little danger even if the voltages are very high (think of Tesla holding the his Tesla coil).

This circuit has high currents which makes it considerably dangerous.

a 65W CFL can deliver 65mA easily (65W/1000v).

And if you look at the picture below, at greater than 50mA the little guy is dead.


BalanB1 (author)2017-06-18

I used 32w CFL circuit, i mounted Heat sinks to transistors and i connected a ceramic capacitor 1nF 5Kv in parralel with primary coil and I powered up,first i was scared : 30 cm Sparks! awesome, I used my Fluke high voltage probe and a 10 Mohm impedance analog multimmeter and i measured 90 KV.

GergelyN (author)BalanB12017-08-22

30mm you mean, right? (1.5MV sounds a bit too high)... Still impressive.

BalanB1 (author)BalanB12017-06-18

Also i changed the full bridge electrolitic capacitor with a 100uF 400V electrolytic cap from TV for more power!

drugpacan (author)2016-09-15

The mistake was that I connect the transformer to the extreme conclusions from four to CFLs. I measured the resistance of the filaments of CFL lamps. It turned out each strand has a resistance of 3 ohms. Thus, the total resistance of the two strands was 6 th. I have created a semblance of snubber 6 ohm resistance + series capacitor 3.8n 1200 volt and transformer connected. Since the capacitor on the board already exists, between two close-pin soldered resistance 6 ohms, and the transformer took another couple of conclusions. If the CFL 26 watts of power, the arc starts at a distance of 4 centimeters and is very noisy, blue and red and thick. If the CFL power 11 watts, the arc starts at a distance of 2 centimeters. If the distance is increased to 4 centimeters, there is a glow discharge and begins to secrete large amounts of ozone

BalanB1 (author)drugpacan2017-06-18


BalanB1 (author)2017-06-18

Ok i builded it,and i used a resonant capacitor but it consumes a lot of power and full bridge capacior just exploded,i don't recommend to use capacitor from CFL circuit,try to use a lower capacitance capacitor and install a safety spark gap to discharge HV ripple to avoid damaging the oscilator.

aladdin_2005 (author)2017-05-21

thank you good work

siriusfuenmayor (author)2017-05-16

this circuit could be run in reverse? i mean, if you input high voltage in the flyback you get low voltage on the CFL inputs?

TheGreatResistor (author)2016-08-24

Hi thanks for the instructable i tried it was awesome i got 2.3cm sparks
but it also gave off a weird smell i thought the transistor burned off
but it didnt after research i found out it was OZONE. Is the ozone
produced by this circuit dangeros

yes, my friend, ozone is very dangerous, ..but it small quantities is safe ... ..but I don't recommend you to "sniff"it directly from the assembly anyway .. do it in the room with opened window to make sure the air is circulating in the room ....there really IS something as ozone poisoning ..but you're ok , don't worry

Aghast. made it! (author)2016-09-24

I made it. Simple and fun. I have three electronic ballasts. I tried connecting the primaries in parrallel and secondaries in series to try and triple the output voltage. It didnt work though :(. Not sure why

MichaelC00 (author)Aghast.2017-04-24

A ballast only converts ac power to dc. in order to triple the output you would need the transformers

pchowdhary1 (author)2016-11-30

check this out ppl... look how does 3.5 MILLION VOLTS look like .... beautiful

LikhinM (author)2016-09-26

Hi, can I know the value of the capacitor you used in place of the filaments...?

Thanx in advance..

drugpacan made it! (author)2016-09-15

and here is a photo

drugpacan (author)2016-09-14

Greetings! I have nothing works. Ballast CFL - rate transformer scheme. It is unclear where the soldered capacitor - parallel with the primary winding, or sequentially. If sequentially, the same nothing works. I do not know why not start CFL ballast?

SHOE0007 (author)2016-08-14

For a flat aluminum glass capacitor I am using this as a power supply simple and safe to use.

SHOE0007 (author)2016-08-08

From e-bay just for 5-10 dollars you can get pulsed high voltage generators. Mine operates DC from 3.6 volts DC to 20,000 volts 50 mA. I know the Danger. I have worked with voltages like this for lasers used at school.

You can get some that operates up to 1 million volts!!!

aphasic made it! (author)2016-06-23

Wow, this Instructable has a lot of very low quality comments.

Anyway, thanks to Biotele for the neat idea.

I opened a 15W CFL and turned it on. It wasn't a problem to measure it with my AC voltmeter, which is rated for 300V. I measured the frequency, it was about 54kHz. I measured this with a Radio Shack digital voltmeter with a frequency function, using only one probe - the other was connected to nothing, but the readout was steady. I also noticed that when I connect one end of the AC voltmeter to one of the right two pins of the CFL bulb, and the hold the other in my hand (even without touching the probe metal), the needle goes way up - just from the capacitance in my body. I can touch both ends of the CFL bulb while it is powered, without a problem. It starts to burn my finger after a while but no heart attack. I tried first with two fingers on the same hand, just in case. When the bulb isn't fully connected, or is partially connected, then I get a shock, so be careful.

The bulb was connected to four pins as in the pictures Biotele shared. I connected the outer pins to the transformer and got a 1/4" to 1/2" arc, but only after bringing the high voltage wires close together and drawing them slowly apart. The arc melted the ends of the wires a bit. Disconnecting the bulb from the pins, so that only the transformer was connected, did not seem to produce a stronger arc. The transformer primary effectively shorted out the bulb - i.e. the bulb did not light up when the transformer primary was connected across the outer pins. Moreover, the circuit was oscillating at 82kHz with the transformer in parallel like this, so I think the presence of the transformer affected the properties of the circuit. Also, one of the transistors in the CFL circuitry began to smoke.

I tried connecting the transformer in series with the bulb. This worked using either of the outer pins. I could create an arc of about the same length as before, but this time the bulb stayed on - except when the arc got to its maximum length, and then the bulb started flickering and the arc made a buzzing sound.

I didn't know very much about flyback transformers before this project. Apparently they are so named because they capture energy from the deflection coil in the horizontal "carriage return" movement of the electron beam between scan lines in a TV.

I didn't look for a schematic, but followed the linked instructions for identifying the pins of my transformer (it came from a big HD color TV). It was interesting that there is a 20-30V drop when DC is applied to the high voltage secondary of the transformer, I guess this means that the rectifier consists of many diodes in series. I also could have identified the high voltage ground by process of elimination - it was the only pin on the base which was not connected to some other pin on the base. Selecting the primary was done by attaching to the various pins a 24 VAC, 1.25A power supply with a 22 ohm resistor in series, and looking for the biggest AC signal at the secondary (it wasn't very big).

mb45380 (author)2016-05-17

Please see "Fire Extinguisher using Voltage Multiplier"

aadi8007 (author)2016-05-13

Can i use external electronic ballast ???

nadgoB. (author)2016-03-14

So by 65W bulb you mean it takes 65w from the grid and makes it into like idk 200W or does it take like 11W and makes it into 65W
I hope you understand what i mean
Because when you buy a bulb its says like 11W (shines 100W) or what ever

landraka (author)nadgoB.2016-04-30

What they mean by 11W shines 100W is that the florescent bulb only consumes 11 watts of power but is as bright as a 100 watt incandescent bulb. They put that on there because people were used to how bright a 60W or 75W or 100W incandescent bulb was and didn't know which florecent bulb to get.

Also, power is mesured in watts and power is concerved so a bulb could not take in 65W from the grid and turn it into 100W because it would have to be making power somehow. What it can do however, is convert the 120 volts from the outlet into 600 volts but to achieve this the current drops so power is conserved.

nattyaku (author)2016-03-12

can anybody help me to get this pdf. i can't have pro account.

1_64phiGUY (author)2016-03-06

Can this get a small ac voltage increased? Or must the input be 110v? I only got about 70volts to play with. Way small amps too.
I,m spinning an ac synchronous step motor by a tiny dc motor by 9volt battery. Speed controller us only circuits I use in my contraption. Its a Gerard Morin replication, BASICALLY. Resonance.
Yes, looping is possible, I,m able to spin another tiny dc motor, but need more boost, then I could hit sweet spot- then disconnect battery=looped:)
So????????? Is this circuit able to function in a way relative/& or beneficial? Ty....
Please don't be rude if your to closed minded. You will not get me like that. Ty

mazhargulam (author)2016-02-16

Thanks a lot will try and see good luck

mazhargulam (author)2016-01-19

Hi i would like to know how long can you keep the arc going, like can you keep the arc going for about2 to 3 hours?


GyurikaK (author)mazhargulam2016-02-14

The fly back would be able to handle it...but I burned a CFL in 20-30 mins...

mazhargulam (author)GyurikaK2016-02-15

Thanks a lot for the reply can i use the older model ballast that they had in the floroscent tubes,and also what if i use a very high wattage cfl?

GyurikaK (author)mazhargulam2016-02-15

The ballast will work better from what I've seen, not myself...i seen it in videos, the guy said since it runs for days it can run a lot...but I am not sure...he didn't try it more than 30 mins too and... The cfls run a lot of time...but they fry out...damn easily... You could try, and you can find adjustable ballasts...which is nice and good to have the +adjustability :)

silverassassin3 (author)2014-10-29

Hey, so after completing this, how could you use this to power something else? like is there a way to connect it to another source?

GyurikaK (author)silverassassin32016-02-14

You can make a Tesla coil...

pravinjoshi57 (author)2014-12-25

Can I get circuit with Tesla coil....?

GyurikaK (author)pravinjoshi572016-02-14

Yes you can

nickglowsindark (author)2009-06-07

I understand that occasionally the lead configuration in the CFL will be different, but the one I got looks nearly identical to the picture you posted. By the end of the project, I wasn't getting a spark any longer than about 4 or 5 millimeters, and so I started troubleshooting. As far as I can tell, everything was hooked up properly- I'm confident that I identified the correct terminals on the flyback. I used a multimeter to check the voltage output of the CFL circuit (expecting it to go up in flames- the meter was only rated for 600V) and instead saw a value that started around 120Vdc and dropped as soon as I hooked up the meter (like a capacitor would have). I decided that I must've hooked it up to the wrong wires, but after checking voltages across all combinations of the 4 wires, the most I saw was a constant 120Vdc output across an outer and next-to-outer wire. What sorts of values should I be expecting at the "high voltage" leads? Do I need a different kind of CFL, or did I manage to fry something internal accidentally?

GyurikaK (author)nickglowsindark2016-02-14

I don't know what kind of CFL you have...but all my CFL s had the two out put...thingys...parallel with the input wires, which are the closest to them...

I am having the same problem but at about 270v (my wall plug is rated at 230v) I checked all of the pins, some combinations were like 10v, 40v and 270v (if I remember correctly)

Biotele (author)nickglowsindark2009-06-09

Hmmm, You may have a fried CFL, The transistors are just passing along the voltages from the rectifiers. Do you know how to test a transistor? Or.. the circuit is not oscillating and you need to add a feedback capacitor like in step 4.

nickglowsindark (author)Biotele2009-06-09

I ended up having to add the capacitor to get any voltage at all out if it, and I know how to test transistors, but I'm afraid it's a moot point now, since a little bit more fiddling around resulted in an interestingly colored flame and a terrible-smelling apartment. I'll be picking up a new one later today probably and giving it another go.

i_build_stuff (author)2009-02-03

i just took the board out of a large CRT television very very carefully it had been unplugged for only a day or so. The board i got is riddled with large capacitors and I'm to scared to touch anything let alone try to discharge it. On the biggest one it says on the side 200v 820uf 5258(m) can anyone tell me exactly how dangerous this thing is? Until l get a better understanding of this thing I'm puttin it in a box and stashin it.

GyurikaK (author)i_build_stuff2016-02-14

You can connect a plug to the wall and use the ground wire to discharge whatever you want, I use it too, I is really painful being hit by those... I'm telling this from experience, but the caps and anything on that wont zap you, but the fly back, if you get close to it, you're doomed xD

If you are scared you should not be doing this. That means you do not know what you are doing.

Biotele (author)i_build_stuff2009-02-03

it can give you a jolt. Take a screwdriver with a rubber handle. Hold the screwdriver by the rubber handle and try to short the cap by touching both pins with the metal part of the screwdriver. You might get a loud bang (or maybe nothing if it is discharged). Keep one hand in your pocket. Do this to all the big caps.

GyurikaK (author)2016-02-14

So...I have a big...fairly big fly back, but in my small city I can't find but 30w CFLs so I was wondering if I can hook up 2 to the fly back... I suppose I have to connect them in series :/, with only one I can squeeze out just... Around 7kV wich is not enough to power my SGTC :(

handyman29 (author)2010-08-23

Is it 50mA or 50MA, because I don't think 0.05A is enough to kill someone. I would think 5A is enough to cause damage. I've worked with wires connected to 500mA sources and I am still here.

DeusXMachina (author)handyman292010-08-23

It's 50mA, across the *heart*. Remember Ohm's law, I = V/R. A 12v car battery is capable of producing several amps peak current. However, the resistance from one hand to the other is around a few thousand ohms. So, it is possible to grab both terminals of a car battery, contrary to what you see in movies. You need enough volts to push pass the resistance, AND just enough current to move through the heart, to risk electrocution. Wall voltage is enough to do it, and definitely some parts of this circuit are able to.

factors affecting body resistance, sweat, humidity, sea water, chlorinated water, ac or dc. the lowest voltage, reportedly causing death is 6V dc. so do, you feel lucky? the body skin resistance, can change according to the situation. and of course once the initial punch through talks place, resistance is dramatically decreased.

plus there are the other effects, of physical clamping or vaulting. which can cause, physical trauma that can cause death or injury. but at 50ma, flash burn would not be a factor. but as voltage increases, the possibility of soft x-rays becomes greater and depending on the type of targets used such as tungsten. and at 50kv you are, entering medium x-ray electron volt range. which is only briefly, for the arc start period.

and remember in short circuit conditions, voltage decreases and amps increase. so your 50ma could quite possibly become, 100 ma sufficient to stop the heart of cause fibrillation or freeze the lungs causing suffocation in case of clamp on. because of the two scenarios of either freeze or being thrown away. if your muscles contract you freeze, if they expand your kicked away.

this all plus, the danger is not necessarily over after being shocked. since people have died, even after a week of being shocked and only thought, they were ok. so it is best to minimize the risk, of receiving any shock to 0 chance.

Something else to note: the 20mA range can be more dangerous than the 50mA range. 20mA will disrupt the heart's operation (even with current removed), leading to a potentially painful death. 50mA+ clamps the heart; remove the current and the heard will resume pumping.

Again, this is lethal Primarily if the current path crosses the heart. Those who have worked with (and even taken) more current and still post here did not have a current path across the chest.

A simple and effective precaution when working around high voltages is to always keep one hand in your pocket: worrisome situations occur when one hand is grounded and the other is touching a hot lead.

tristantech (author)DeusXMachina2010-08-24

yes, you are absolutely right. Thanks for clearing this up.

skrubol (author)handyman292010-08-23

5-10mA is enough to kill, but it's not likely to. I've been hit by 9kV, 30mA, as well as shocked myself multiple times with 120V house current (on a 10+A fuse,) and I'm still here. That doesn't mean either of those situations can't kill you.

About This Instructable




Bio: Make a better Instructable, and the world will beat a path to your website.
More by Biotele:Easy to Build WIFI 2.4GHz Yagi AntennaHand Cranked Flashlight Mod plus Joule ThiefGanzfeld: Hack Your Brain the Legal Way
Add instructable to: