Some open source antispark designs that work

I couldnt get any of mine to work, one blew a diode, the others just output 0.2v where the voltage gets dropped by the switch driver, I really can’t tell if the ground is soldered but I use my largest tip and hottest temperature on the via with hella flux. My ear is ringing pretty bad but i really want to fix this.

I dont think I am going to try making any more, but I will find someone with better soldering skills to make the rest of mine.

Did you solder the ground pad from the underside of the PCB?

I did, I really don’t know what’s wrong you can see there’s no solder bridges but I digress, a swing and a miss.

did you check to see if the regulator output is present?

Did you check to see if the bootstrap capacitor has charge on it?

Inputting 18~v through 9v batteries, regulator outputs 12v, however gate driver outputs 7.7v and outputs it regardless of switch position. C3 at input voltage, C2 at 12v, C4 at 9v and C1 at ~10v. What should I check?

Edit: I got the switch-off drain down to 4.4v after resoldering the gate driver but still doesn’t give any more than that with the switch on, the voltage just gets muddied after the gate driver I really don’t know what to say.

Sounds like the gate driver got busted somewhere along the way, I suspect the ground pad wasn’t properly soldered and you tried to enable it. This has happened to me a few times.

C1 should have around 12V across its terminals, C2 should be regulator output. C4 should be 0 volts when off, and at the input voltage when on.

If you have any parts remaining you can try doing a clean assembly, double checking all the solder connections.

Is there a trick to soldering the ground pad? I try my hardest to fill the via in with flux and then using a tiny bit of solder to cover all of the copper and ground pad but I feel like its just not soldering the ground pad, I got 3 PCBs and parts left I’ll try one final time.

What soldering iron are you using? You might need to hold it there for a minute or two. You should see flux seep from under the chip on the other side.

60w weller station, using a knife tip for smd stuff and a big chisel tip for the ground pad, Im pretty sure thats it I’ve only been holding it on there for 5-10 seconds I figured I’d burn the chip but this gives me confidence.

Yeah I think that was it, going to try a fresh one tomorrow morning when im coherent and hopefully thats the one.

Hi, I built switch with ltc7004 and irfs7730 and i am getting a very weird problem. It cant keep the charge on mosfets, just oscillating very slowly. Voltage on Cg sweeps between 2.8 and 3.9V. I have calculated different values for these mosfets and it should work without problems. I also tried low leakage diode between vcc and bst, also no effect.

My values are Cb=4.7uF, Rg=220k, Cg=100nF.
Ground pad is soldered properly.

Nevermind, propably cold joint or something, it works now

Please use better (and cheaper) MOSFETs.
IPB014N06N is a drop-in.


I had some 7730s on hand so I used them.(4 parallel 2.6mohm, so 0.65milliohm max. total, that’s only 130mW power losses on 200A, plus I can populate another two.)
Your choice is 60V only, I know that lot of antispark switches died in the past with this voltage rating on 12S.
Seems enough for a switch:

Past antisparks died at 12S because they were half-assed in their design.

International Rectifier is overpriced and overrated, there are cheaper MOSFETs available with superior electrical characteristics.

Isn’t this caused by transient over 60V? Have you measured this? I want data!

If you paid attention to any of the other antispark threads you would know that previous designs failed due to thermal shock from the inrush current of charging the decoupling capacitors as the MOSFETs passed through their miller region during turn-on. They also had other problems because they switched on the ground side of the circuit. This design switches on the high side.

The actual capacitance cut-off will vary between MOSFETs, the smallest ones will survive up to 6,000uF of load capacitance before failing short circuit. I tested this experimentally. I can also estimate this based off the controlled rise time from the RC network, corresponding power dissipation, and SOA/Maximum power dissipation of the MOSFET.

And how do you explain that 75V mosfets will not fail?