Trying to figure out this bug zapper inverter

[cte7ds]

I found it so intriguing how those tennis racket style portable bug zappers can generate such a big spark and a loud bang from only few batteries that I had to get one off ebay just to see what makes it tick. What I found was 2 D-cell batteries in series and a really interesting inverter board with few high voltage parts. I was surprised how simple it is.

Heres what I got so far:


Parts list:
CY1 - The big red capacitor
CX1 - The big blue disc, a 250V~ X1Y2 capacitor
D1,D2,D3 - RFC3K, 3kV fast recovery rectifier
C1 - 2kV high voltage ceramic capacitor, says 471K on it, so probably 470pF
TR1 - has 3 pins on both secondary and primary side, the coils in the schematic are probably not accurate
Q2 - TND017MP, N-Channel Mosfet for lowside power switching

I've been googling around and found out that the transformer is probably some sort of step-up/flyback transformer that drives the transistor to make an oscillator that switches the incoming 3VDC for the transformers primary coil. The windings of the step-up are probably something like 10:2000.

Another thing I found out is that the C1 and D2 form a Villard circuit that doubles the voltage output of the transformer. What I didn't figure out is why the D1 and D3 back-to-back diodes and the so low-rated CX1 capacitor are there. Couldn't you just charge the main storage capacitor (CY1) straight after the Villard circuit? Or perhaps the CX1 acts as some sort of current limiting?

I haven't yet measured anything because I don't have an oscilloscope or a high voltage probe for a DMM, however I was thinking of building one myself with the instructions found from here: http://www.repairfaq.org/sam/hvprobe.htm

Some things I would like to try and use this circuit for:
1. Attach a bigger capacitor bank, possibly those big electrolytic filter capacitors found from computer SMPS supplys, like 220uF 400V and 820uF 200V, maybe if I put enough of them in series I could charge them up with this? Or removing the Villard circuit so I wouldn't need to many in series to handle the voltage?

2. Use a voltage multiplier with a lot of stages to get bigger sparks. But I'm not yet sure which diodes or capacitors to use?

3. Combine the above ideas to get something crazy

Oh yeah and I'd like to keep it... fairly portable, simple and cheap and fun, so if you have any ideas, please do tell.

[amspire]

Your schematic is not quite right. The 220 ohm resistor is meant to be in series with the LED.

Also the transformer schematic has to be wrong. Your Pin 1 has to be connected to a primary winding, not just the secondary winding. It could be that the primary and secondaries connect together on that pin.

The TND017 is an IC containing a mosfet, not just a mosfet, and you have the pins mixed up a bit anyway.

See if you can sort out the transformer a bit better - try and work out where the winding on Pin1 connects to the other primary windings.

Richard.

[cte7ds]

Your schematic is not quite right. The 220 ohm resistor is meant to be in series with the LED.

Also the transformer schematic has to be wrong. Your Pin 1 has to be connected to a primary winding, not just the secondary winding. It could be that the primary and secondaries connect together on that pin.

The TND017 is an IC containing a mosfet, not just a mosfet, and you have the pins mixed up a bit anyway.

See if you can sort out the transformer a bit better - try and work out where the winding on Pin1 connects to the other primary windings.

Richard.

Thank you a bunch for your input, your absolutely right, I had a lot of mistakes on that one. I made a new one with hopefully less mistakes :p

I also did some basic continuity and resistance measurements on the transformer (in circuit). There seems to be a dead short between the pins 1 and 6.

There was a 1.1ohm resistance between pins 4 and 5, so that's probably the primary coil. Also there was a 0.571k resistance between the pins 2 and 3 so that's probably the secondary coil. I found no other continuity or resistance between the pins.

I updated the TR1 symbol on the schematics to reflect my findings.

Hopefully now someone smart can shed some light on how this whole thing works?

[T4P]

It's a lowside power switch which means,
it switches near ground and if you read the datasheet last page you can see below the threshold it switches off and above the threshold it switches on
so it means to switch it to high frequency ( as it is ), it depends on the induction loop of the transformer to bring the voltage down along with the resistor or bring the voltage up above the threshold to switch it on and off at high frequencies ...
Okay ? The chip is near ground therefore being 'low side' AND but not limited to drawing the power from the out (The load)
So the 'IN' pin is the 'feedback' loop that's where it was designed to use pin 1 and 6 of the transformer
It's not neccesarily a feedback loop, it can use a MicroC, but the inverter was designed as a feedback loop to depend on the induction loop around the transformer to give the feedback loop
This is interesting, i believe what i said should probably be right

IF you do notice why they used a villard circuit on purpose is because the ripple characteristics is poor and that makes a good one for even higher voltage generation
As for CX1 i think it is a return path to ground and it doesn't need so much voltage ( I might be wrong ! )
D1 and D3 might be a full wave rectifier ... but at the same time i also see a cockroft walton voltage multiplier
Meh confused

[cte7ds]

Here's how all the component leads connect to each other on the board:

I've been trying to search wikipedia about voltage multipliers and voltage doublers but I can't understand which circuit this one is.

[CampKohler]

I re-drew the diagram in an attempt to understand the secondary's operation. Sometimes re-drawing to the simplest layout helps to figure things out. I suppose it doesn't really matter, but I always start with the input on the left, so that the current flow is to the right. (Insert favorite "Australia is upside down joke here.")

Notice that CY1 is physically bigger than the other two caps. Part of this is due, of course, to it's higher voltage rating, but also because, I suspect, it has a larger capacitance. Over time (several charging cycles), this will be hold a bigger surprise for the poor bug. So, the first two caps get fully charged on the first alternation (after turn on or zapping), but, if the third cap is larger, it will take several steps to charge it fully.

Perhaps you could supply us with the missing capacitance ratings of the caps.

Notice how crappy-looking the soldering is; it looks like there might be some cold solder joints in there.

------------

If you are going to try building some voltage multipliers, be very careful of how much storage you build into to it (capacitance). Even with a puny circuit like this one, over time it can charge up the largest of caps and kick you through the goal posts of life. Using some dry (waxed? varnished?) wooden sticks with metal probes (paper clips?) and a well-insulated resistor and suitably-rated wires (check!), you can build a discharging apparatus inexpensively. It also is a good idea to place a permanent resistance across the output that is high enough not to load down the output, but low enough to discharge the output cap after the power is off for a few seconds. (But see below about voltage ratings of the resistors.)

Another cheap safety device is a NE-2 neon bulb and series resistor across the output that will let you see when it is discharged, or, more importantly, when it is not. These can be salvaged from old appliances such as electric blankets, etc., but you will need to up the resistance to limit the current when operating at higher than line voltage; add additional resistors in series with the original one, because each resistor (1/2 or 1/4 watt) should not be subjected to more than 500 volts or so.  So, if you were getting really frisky and putting out 5KV, you would need at least 10 resistors in a string for the voltage rating, but the more the better. For resistance value, use Ohm's law to calculate the current drawn when the voltage across the resistor is line/mains minus 65 (the lamp voltage). Then use that current to calculate the needed resistance for the output voltage minus 65. See http://www.giangrandi.ch/electronics/neon/glowui.html for more about this lamp. (Note how they are using an R as a shunt to read the lamp current with a scope.)

Finally, make sure you have a signed will on file and read this.

------------

With regard to your lack of a scope, try getting in touch with the amateur radio operators in your area. Someone might have an old (tube?) scope squirrelled away that is "too good to throw away, but too old to use." Even something like an old kit 1MHz scope would be fine for this work with the proper voltage-rated probes. (Watch Dave's recent video on scope grounding first!)

--------------

PS: Would somebody please tell me why my drawing came out so small (before clicking it) as compared to the others? It was of a nice enough size when created in MS Paint and would be a pain if I had to make it say, ten times bigger. Because I can't see the data of your photos, I have to assume you can't see the data for my drawing, so here it is: 33.61 kB, 519x432. What was the data for your schematics?

[Peterp1per]

Hello I'm on a long camping holiday in southern Europe just now and being a bit hard up for a project and also beset by flies I've started a bit of a collection of these. I find them in Chinese' sell everthing' shops which are all over the place
What surprises me is the wide variation in design, there seem to be two main threads, a voltage trippler charging a modest cap of 10 to 22nf to about 3kv and a straight rectified output into a 220nf cap at about 1500 volts. The second one is more lethal than the first but every one I have had (it's a premium brand) stops working quite soon due to the main cap losing value.

Following on from this there is also the issue of 'cooking' current, once stunned on the grids some source more current to kill the bug by heating than others.

I don't have access to components, and can't use mail order as we constantly move, currently I am stripping CFL bulbs for bits, also throw away cameras have some high voltage parts.

So far I have concluded that the cheap ones are best, maybe because they don't comply with any safety regs (I think it's 47 micro coulomb, ha ha, no chance of a humane kill)
They will also tolerate an extra cell in the case of the 2 cell mains recharging version.

Ideally then about 3.5KV any more constantly arcs across the murder mesh.

A decent short circuit current maybe 5mA

Half a micro farad or more, though this would likely be physically too big at 3KV.

Any ideas where a cap that size might be scavenged?