Monday, May 21, 2007

555 Timer-Based Flyback Transformer Driver

By leorick  |  5/21/2007 09:04:00 PM 41 comments

My 24kV high voltage "Jacob's Ladder" from DIY flyback transformer driver using 555 timer.

Creating an electric arc

I've always wanted to create an electric arc but don't know how. Then I come across the theory that air breaks down at about 1MV/m (Mega Volts per meter) (24kV/in). That mean you need 1kV in order to get 1mm arc. So you need a higher voltage. One of the method is to use a flyback transformer that can be found from an old TV or an old CRT PC monitor. It could generate about 10 to 30 kV. Other method is to create a "tesla coil" which is quite complicated. Maybe it will become my next project.

Flyback transformer and preparation

Flybacks can be found in all types of monitors and screens that use a cathode ray tube (CRT), e.g. TV sets, computer monitors etc. It has a big red cable with a suction cup. It looks something like below.

Next, you need to identify the primary and secondary pin out. Thanks to "Lab HV-PS page" for providing an instruction on how to find the pinout. The main HV out on the secondary coil is a big red cable with a suction cup. Now we need to find the 0V pinout for the secondary coil. The trick is to use a DC power supply. This is because the flyback secondary coil resistance is much too high. There is no way you could find it with ordinary digital multimeter.

So use your own understanding on the circuit below to find the 0V pinout. Give it about 12V and your meter should show some volts when you find the 0V pinout. For me, just to be safe, try to find a datasheet of the flyback transformer or try to find the TV or old CRT PC monitor service manual/schematics diagram to find the pinout like below. Most modern flybacks include built-in HV rectifier diode(s) and/or voltage multiplier (tripler) so output without additional components will be high voltage positive or somewhat smoothed HV DC. So, make sure your polarity is correct.

Unless you have one of these multimeter, you should get the resistance reading out of it. From my FLUKE 189 multimeter you can see that it shows more than a hundred Mega Ohm. That is why ordinary meter could not measure it because of it's limit. Below I test two types of flybacks with 112 Mega Ohm and the other about 522 Mega Ohm. Again, polarity is critical to get the reading.

To find the primary coil is a much simple than the secondary coil. The primary coil resistance is about 1 ohm and again I confirm this with a TV or old CRT PC monitor service manual/schematics diagram. In my case I could only get 0.45 ohm.

Creating the flyback driver (20kHz with 90% duty cycle)

Thanks to "Jonathan Filippi" for the idea. My circuit is quite different. I try to fix up the frequency and duty cycle with help from simulation software. I use "Electronic Workbench" to simulate the circuit which can generate about 20kHz with 90% duty cycle and I come out with this. Using 555 timer to generate 20kHz with 90% duty cycle. Next I try to put it on the breadboard and test the output from it. I get about 18kHz with 85% duty cycle. Jonathan Filippi is using 2N3904 and 2N3906 but I'm using c1815 (npn) and a1015 (pnp). I found out that you can use any multipurpose transistor and I could find it on my old TV board. For the MOSFET, Jonathan Filippi is using IRF840 but I'm using IRF630. You may try to find it's equivalent and experiment with it. Just make sure it is compatible if you want to use other types of MOSFETs. Find it's data sheet and compare the characteristic for both types of components.

Before assemble it, I test this circuit with a small transformer which I can find it on the same old TV board. Since I'm getting too excited, the quick test is to connect the output to the lowest resistance coil. I test it with a limit resistor and surprise, I can get a hundred volt out of that.

Next is to plan to transfer it to the stripboard/veroboard. Here is the stripboard layout and the assembled circuit board. Make sure you mount the MOSFET to a heat sink since it going to heat up while running/powering it up. Note that I put the 150 ohm "snubber" resistor and diode near the flyback. This is to suppress ("snub") electrical transients that might damage of the circuit.

Again, I test the assembled circuit board with the same small transformer and I could get a neon to light up. This mean I'm getting about hundred volts. Neon needs about 80V or higher in order to light up.

Test it out

Now it time to test it out. Get a high power supply for this test. Don't use an expensive lab power supply for this test. It might burn or damage. For me, I'm using a 12V DC battery charger that can give about 5 Amps. You may also try a car battery if you have one.

There is an arc!. At last, I could get an arc out of it. I try to measure the initial max. length and I could get about 24mm. Thus, it is about 24kV. Remember the theory 1MV/m?.

I measure the DC operating current. It is about 5 Amps. I've blown my DC power supply fuse in the process. Maybe I need a bigger power supply :-) . At least I've got some arcs.

Good luck!


Anonymous said...


Anonymous said...

Nice work. I love the way you explain each step. Chris form Texas

I really enjoy this as well. I don't think I would be able to do it myself but I like the way you wrote this and added pictures for each step.

Circuit Labels

Anonymous said...

You can't have 24kV there because the air humidity is higher than 0% so you have less than 1MV/m. I could get 2cm arcs using a small b/w TV flyback and only 15W input power. The 13kV HV diode that came with the flyback is about 6cm long. Another HV diode from another b/w TV rated at 20kV is 8 cm long.

Anonymous said...

nice i hope it works because it is apart of my graduation project.

leorick said...

Good luck. :-). Just make sure your power supply is powerful enough. Also, don't forget about humidity factor as mentioned by etiquette when you come to the theory part.

Anonymous said...

Very cool. I have 3 flybacks from TV's and computer monitors. Actually, i have two now, because i blew up one because i didnt know what i was doing. (NEVER plug a flyback into wall current btw)
Will this circuit work with all of them?

Anonymous said...

Your site is interesting.I am however disapointed at your use of the 555 timer,since a more robust PWM driver circuit can be built with discrete components.This would serve to further the education those studying.
This comment is particularly relevant given the current trends in component availability and the decline in electrical restoration as part of wastfull society..

Anonymous said...

Looks like a great project - can I put a link to it on our web site? My customers are always looking for new projects to try.

i want generate the output about 2kv just for shock lizard can u show me the circuit?

i use L DR sensor for detect circuit.triangular wave for switching mosfet and connect to flyback circuit,can u show me the idea of this circuit just to get output 2kv for shock the lizard

Unknown said...

One quick thing about the breakdown field; the electric field is the gradient of the electric potential (E=∇·ΙΈ.) What this means for your setup is that the electric field near the wires is not simply the voltage you achieve divided by the distance between the tips, but is actually a function of the geometry of the individual leads. If they are sharper, then the electric field in the vicinity of the tips will be higher than if, say, you replace the end of the wire by a conducting sphere or plate. A simple method to get a more accurate estimate of your voltage might be to build a voltage divider with a sufficiently high resistance that you don't bleed off too much power, and then to measure the rms voltage across that.

Unknown said...

Thank-you so much! This is the fourth or fifth circuit I've tried and the first thing to work! And all of the components were scavenged from other devices (except the perfboard).

I used a 3904 and 3906 for the transistors, and substituted the TVR10G with a 1N5399. Also, I switched up to a 4700uF 25V power cap. And it all works perfectly, using voltage adapters from 9-19 volts from 900mA to 3.5A! Thanks again!

is great to have so easily to read these blogs, they are too useful, this one in particular helped me because it explained step by step how to properly work the equipment!

Anonymous said...

A word of caution: If your removing a flyback from a crt, make sure you short out the crt first. Do this by inserting a flathead screwdriver under the anode cap on the crt with a jumper connected to the "dag" ground. Crt's hold voltage. Some have a bleeddown circuit. But don't take any chances.

Anonymous said...

Actually, why have those transistors in there at all? The 555 doesn't have an open collector like the 558, so it should be able to drive the MOSFET directly

leorick said...
Andrew said...

I have tried many circuts and with the transistors this one seems to be a little nicer on the FET and cause them to turn off properly.

Previously my fets seemed to continue to conduct and would start heating up however this circut seems to reduce the heat and swings the fet all the way on and off.

Im going to reduce the frequencey so as to use automotive coils also I have found using an MOV seems to quench the back emf quite well however have also tried usign a neon indicator lamp and resistor that was set up to strike at about 100V and this also seemed to clip the back emf really well

JJArif said...

blog yg keren mantep keep rokin bro

Anonymous said...

what if i remove the two npn and pnp transistors and i connect it directly to the circuit

Anonymous said...

nice project, i built the same circuit as you did and combined it with a sony flyback transformer. I got some nice fat sparks, the longest being like 4 cm. It kept me busy for a while... Thanks for the schematic. Greetings from Belgium.

Anonymous said...

Hey there, I recently built this circuit but it does not oscillate fast enough. i can hear the constant clicking on the flyback and even feel it vibrate but no arcs. it gets very hot when run from a car battery even with a heatsink but still no oscillation at the right frequency. what am i doing wrong?

leorick said...

If you can feel to vibration, meaning that the frequency is not high enough. I think there may be something wrong with the oscillation circuit.

Paul said...

Having fun with electronics ? Me too , spent countess hours ! I have built many electric fence chargers , A switching power supply ( 120Vac to 12V/150A ),car amp with 90V ( +45/-45V )switching power supply that can handle 1000W easy .
Have fun !

Anonymous said...

I got everything working well, but I've noticed that the output voltage is starting to decrease. I was getting 11.4kV with ~20V input, then 8kV, and now 4kV. Any ideas on why this would be happening?

leorick said...

Flyback operates ~15 kHz. Frequency dropping perhaps. Try to measure freq. while the output voltage is dropping. You might see a trend there.

Anonymous said...

I modified your circuit slightly and separated the timer circuit from the power MOSFET and Flyback. So, now the input into the primary is 20Vdc. I get the 12V, 18kHz output from the timer, but once I connect the MOSFET to the flyback, the output drops to 7V, 17.5kHz. I'm unsure how to post my circuit schematic, which may give you a better idea of what I'm doing.

leorick said...

Hmmm! Try to connect the MOSFET to a small transformer just like I did before, just to verify that the circuit is OK. Maybe the flyback it self is having a problem? Have you try other flyback?

Anonymous said...

I got it figured out. When I was measuring 11kV, the HV lead wasn't close to ground so the voltage reading was high. Once I came into close proximity with an Ion Cord connected to ground, the voltage starts to drop and forms a small corona in a helium atmosphere. A very nice and simple way to do some electrostatic spray studies.

Unknown said...

I have made the circuit, it is working fine.I have used two mosfets in parallel instead of one and used a potentiometer in place of the resistor going directly to the oscillator supply rail.
I found that the system works without blowing the 3 amp. fuse I used between the supply rail and mosfets.
Perfork--Arjun R. Nair

Unknown said...

What did you test the circuit with before you put the re-charger across it and how much current should the circuit draw?

Would a 12V / 1A wall socket be enough to power a small transformer of a TV circuit board (one associated with the 240V input)?

I tried a couple of lantern batteries in series but I seem to get bugger all volts through the circuit and pretty much nothing at the transformer.

I take it the 5W resister and diode combination is a snubber to stop larger back currents from the secondary coil.

In which case it could be left out while testing with the small transformer?

leorick said...

What did you test the circuit with...?
12V / 1A ... be enough to power a small transformer...?
Using bench power supply, I think 1A is more than enough for low current testing.

5W resister and diode combination... could be left out while testing...?
During my testing, the snubber was already there but I beleive you can bypass the snubber with low current testing.

Unknown said...

Problem is that using a 12V / 1A wall plug for testing results in it shorting out and the current ceasing to flow.

So I presume the total resistance of the circuit is too low resulting in a high current draw from the wall plug thus tripping the cut off inside it.

You say you used a current limiting resistor during this sort of testing but you have not specified where you put it in the circuit.

And my electronics knowledge is not strong enough to deduce where you might have put it.

I have made a copy of your vero board layout and pasted it into a word document....which I was using while setting it up on a bread board.

If I email you that word document, would you be able to place a red line, representing you limit resistor, over the image in the position where you put it while testing at low voltage?

If so what is your email address?

leorick said...

Now I remember about the limiting resistor. As you can see in this image, it was soldered directly to the transformer. Please refer to this schematics (R?). But I forgot the value.

Realy I got it on my need,I want to make a plasma ball.Thank you....

Vaseem Mehran CP

Anonymous said...

i love this too... can i use it as my ignition coil for my mc?

Aniran said...

I have used this circuit and the flyback is on for sometime after that it is getting off. Any suggestion

leorick said...

Aniran: any components went damaged/burned when getting off? especially the MOSFET?


What is the name of the software to made the nice stripboard layout?

Thank you

leorick said...

strip board layout: autocad :)

Anonymous said...

Hello Leorick,

Thanks for the awesome blog. Any idea how to modify so I still get high frequency for ozone but only need less than one amp of current total (for small amount of ozone) (this device needs corona spark)..will continue to experiment!

Recent Articles

Distributed By My Blogger Themes | Created By BloggerTheme9 | © 2014 DIY Electronics