Having trouble with voltage multiplier circuit

[jerryk]

I have salvaged some parts from an old laser printer and made a voltage multiplier circuit following the circuit diagram below as best as I can.  The problems I'm encountering are that it keeps blowing the IRFZ44 mosfets.  I have switched to IRFZ46 and that works but I'm not sure why.

The gate voltage seems to be the problem with a Vmax of 4V and Vmin of -52V.  The transformers are taken from the multiplier circuit that was on the board in the printer.  In fact I just cut that part of the board out since it had slots cut under transformer, diodes, and capacitors.

The goal is to make a portable battery powered Bee zapper.  Right now the current consumption is too high (.5 amp) to have it last as a portable application.

Questions are:

1.  Can any one tell me what I'm doing something wrong that is causing the mosfet failure?  I would like to use the IRFZ44 since I have a pile of them.

2.  Is there a way to limit the current without sacrificing the punch of the output.  FYI - the .5 amp draw is with the added 1 ohm resistor at the source.  Otherwise the circuit consumes .8 amps.

I have attached the circuit that I'm following,  the scope readings of the gate and drain of the IRFZ46 mosfet, and a Falstad diagram of the circuit as I have constructed it.

Any tips would be appreciated.

Thanks - Jerry

[Jwillis]

Maybe one of the experts could chime in here but the only thing that stands out to me is the body diode recovery  is nearly twice as slow on the IRFZ44 at 180ns  compared to the IRFZ46 at 99ns . Every thing else look fairly similar between the two . I'm not sure if you want to rely on the body diode in a fly back circuit . But it seems to work in the video. Could be the IRFZ44 you have could be faulty. Can't be certain about that.
To reduce current consumption you could try increasing the winding ratio on the transformer.
I have limited experience on fly back circuits and these types of oscillators so that's all I have . I'll have to try this circuit and see what happens .

[T3sl4co1l]

Do you know that the salvaged transformer is 1:1 like in the screenshot?

The waveform suggests an unloaded output, and excessive negative gate voltage.  I don't know why one survives and others fail; it could be manufacturer, age or utter luck.  You're lucky it works for one cycle at all, let alone (apparently) thousands.

Tim

[Terry Bites]

Don't zap bees!

[Doctorandus_P]

The yellow trace goes up to 70V while the IRFZ44 is only rated for 60V. It probably goes into avalanche mode there and clamps the voltage. According to vishay this is an avalance rated FET, so it should be OK if you stay within dissipation energy limits. But I do not trust that your cheap pile of MOSfets are really IRFZ44 fets.

The most obvious however is the gate voltage. Gates of MOSfets are made of a very thin oxide layer, and should never be subjected to a voltage exceeding + or - (usually) 20V. Clamping the drain with a bidirectional TVS to the gate may greatly improve the chance of your Fet surviving.

medium size TVS diodes are great for experiments. If your diode is around 3mm to 5mm long, you can directly solder it to the MOSfet, which is very handy for experiments on breadboards and such.

[jerryk]

Thanks for the informative replies.

I modified the circuit (bottom attachment) and now have the gate voltage under control and have switched to the IFRZ44N with no issues other than it appears to be clipping the the peak voltage which I assume is when it is in avalanche mode.  As it stands I could sail forward but would like to tame the overall current consumption.  Another issue is that when I put a 100 mega ohm resistor across the zapper electrodes this seriously dampens the zappers output.  All circuits that I have seen like this have some sort of bleed resistor across the electrodes and I thought 100 M would be adequate.

The mosfets are all either purchased from Mouser in the case of the IRFZ44N or salvaged from various other legitimate donor boards in the case of the IRFZ46N.

I think the transformer winding ratio is way more that a 1:1 ratio but have not tested it.  For some reason I can't use my signal generator to calculate it.  Maybe I need to take the transformer out of circuit.  I do know that the primary is 38 turns or thereabouts and the resistance is on the new falstad sim picture below.

The unloaded output comment has me curious because I don't know what to call the load.  Is it the Bee when it gets zapped?  I did hook the bleed resistor up across the electrodes and all it does to the wave form is increase the frequency with the same voltage peaks.

Is there a proper way to put a snubber circuit on the output to keep the peak voltage within the limits of the mosfet?

I have attached some new pictures of the output wave forms, and the little board arrangement that I cut out of the laser printer.  It is a meaningful project in that these Bees which are actually meat wasps have terrorized our property during the past three summers.  So much that you can't even sit out on the porch and have a beer without being attacked.

The goal is to modify some small grates so that I can just put the grate over a tupperware tub (most likely yellow) full of a bait like salmon or some other fish and see if they will be kind enough to complete the circuit I'm working on.

[camila]

Must have tried this circuit. Have a look on this