Introduction: Reverse Engineer Resin Encapsulated High Voltage Module From China
Every one loves these modules with their long spark distance of around 25mm (1 inch) :D
and they are affordable available from China for about 3-4$.
But what is the problem Nr.1?
They can be easily damaged with just 1 Volt above the rated Input of 6 Volt. So using 2x Lithium cells for more output power is not possible (For example 2x 18650-Batteries in Series=7,4 V) Another common problem is overheating when used too long, but I don't have exact numbers when it's too long.
What is the problem Nr.2?
the PCB is encapsulated in hard black resin so it's not possible to fix broken modules or to understand which component failed What is the solution? I searched the internet how to remove resin since my first tries with boiling water and acetone didnt worked. I found a guy on YouTube talking about removing resin based paint with a heat gun. Bingo! a first hint, if its works on paint it should work on resin too.
So let's try that.
Step 1: How to Start
First I gathered some tools i thought might be useful.
1. a vice to hold the resin module
2. the heat gun with small nozzle 10mm (~1/2 or 3/8 inch)
3. several hand tools i wanted to try out
4. safety glasses (better safe than sorry)
5. gloves to not get burned
6. and just for precaution a dust mask
it's a good idea to have some ventilation since there will be a more or less smell from the heated resin.
Step 2: The Glass Is Half Full (semi-succsessful First Try)
I used the heat gun at nearly 80% of it max temperature (400 degrees celsius)
The trick is this: heat the resin not too much, when you see smoke its too hot, and when you can't peel the resin off the temperature is too cold.
The best tool is a screwdriver which is not sharp. The reason I stopped using sharp tools is it damages the parts of the PCB which I want to recover as undamaged as possible. The heat itself damages the parts on its own so better use a bit more pushing force than too much heat.
On the last 2 pictures you can see the result of my first try.
I ran into a problem, the parts are so close together that even a small 10mm(~1/2 inch) Nozzle was too big and would damage the parts before it was possible to remove the resin.
So a new idea was needed...
Step 3: Second Try
Since the nozzle was to big i switched from the big heat gun to
my SMD de-soldering heat gun with the smalles nozzle i had: 3mm (1/8inch).
I also figured out that 340degrees celsius is enough to remove the resin.
Then I continued with a small screwdriver (without sharp tip)
and worked my way throught and around the PCB and transformafor.
It's a mess :)
Step 4: Make Photos, You Will Need Them Later
Make photos as soon you see the PCB since there might be Parts damaged until you are done.
The reason is for example :
1.wires might unsolder or loose their colored insulation which makes it more difficult to understand the circuit later
2.the surface of components might get scratched or burned and later you cant identify them (from 3 capacitors only 1 survived with unburned markings)
Step 5: Measure Components
Unsolder parts while still making photos before and after.
Then use your multimeter(s) and the famous transistor tester (7$ from china) to find out
1.is the part damaged or not (useful to now where the ciruit failed)
2.type,pinout and charcteristics of the component if the markings are missing/unreadable.
Step 6: Reverse Engeneer the Paths of the PCB With 2 Tools
1.install an EDA Programm (Electronic design automation) of your choice to draw the scematic
There are many free options out there, i used FidoCadJ since its very easy to learn and uncomplicated.
2.now use a continuity tester to follow the paths on the PCB.
Tips:
Now its helpful to use the photos you made before to know which component was at which place on the bare PCB.
Info: The PCB has to be without components otherwise you can't track the paths with continuity tester properly (you would get false positives)
Step 7: Final Result (sort Of)
Now there are only 3 missing pieces left to know to complete the initial goal.
but only one is critical.
1. the voltage rating of the 100pf capacitor on the voltage multiplier part is unknown,
soloution: look at similar circuts or take an educated guess. The voltage might not be lower than that of the 8n2 Capacitor and not higher that 3 of them in series. Answer 3-5kV
2. What is the black SMD component? (one leg broke off when i tried to unsolder it, 2x in 2 cases)
(half :) )Answer: there could be only 2 answers: transistor or mosfet.
But which one? use a standart type and try booth, only 2 possibilities are easy to work out.
But a hint later.
3. the high voltage tranformer is hard to unwind and count its turns so i measured the ratio of input to output resistance.
But the Soloution to the final last 2 Question comes now.
I ordered also some other high voltage kits from china which seem to have a very high similarity when i compare it to my drawn scematic.
1.there was a scematic included which gives us a hint that the damaged SMD part is a transistor.
2. the transformer looks very similar to a popular ebay item and can be ordered form china ebay
("15kv high voltage transformer")
I call this a succsess, now its time to improve the circuit so it doesnt fail so easily.
But this is part of a future instructionable.
I attached the scematic file too. You can open it with FidoCadJ
http://darwinne.github.io/FidoCadJ/
I hope you liked this documentation and have a nice day :)
3 Comments
2 years ago
Good work. Thanks
3 years ago
Good work. Thanks
for sharing. I often wondered about those "400KV" transformers. If the output caps are 3 x 1600V then the voltage cannot be much more than 5KV. And yeah from the schematic the transistor is a MOSFET. I've seen that schematic used several times for the MOSFET oscillator.
Using what you wrote in the schematic, the turns ratio is 1.6H/(.05m)^0.5 = 178. You cannot use the resistance values to calculate this because the wires are of a different gauge and the wire lengths are very different.
Reply 3 years ago
thanks for the info on the transformer to use Henry instead of Ohm.
But on the Output issue, it is higher than 5kV, because there is the rule in physics that a lightning discharge per 1mm air travel distance is 1kilo volt (for pointy electrodes), and 3kVfor non-pointy electrodes like ball or torus. The modules can sparc across 25mm on average which means the output is 25000 Volts. It seems they just "overclock" the caps, but like i wrote they are the last part which breaks, they all where fine.Not sure why but that is what i observed so far.
I have seen booth oscillator circuits, so it can be possible to use Mosfet or Transistor (mosfet is just a special type of the transistor family). But i don't want to break a third module to find that out, since i know that it doesn't matter. Since the goal was to find a basis to copy/improve the circuit.