Pic reset and following lock up problem

[Archangel]

Hi Joe S,

I replaced my IGBT driver with CNY17-1 opto-coupler. That brought my circuit back to life. But the old reset problem still exists.
When I touch one of the crystal pins with a needle probe Pic locks up and when I remove the probe it starts working again. When I touch the other crystal pin or any other Pic pin nothing happens.

The needle probe acts as a path to ground for the oscillator and the oscillator stops as a result, just what I would expect that tiny little R/F oscillator to do.

When I trigger the IGBT driver using Pic powered by the programmer/evaluation board. The reset problem doesn’t improve and eventually it blows the 315mA fuse in 12v power supply of the programmer. The Pic is 16f628 positioned on a sim stick that is plugged into the programmer. The 20Mhz crystal, capacitors and MCLR pull up resistor are positioned on the sim stick as well.

Remember 25 ma per output total all outputs . . . oops I cannot remember and I am in the house right now, but probably less than what that IGBT requires.

I replaced the secondary winding earth wire with a high-tension lead and a spark plug and attached the earth to the spark plug outer shell. That fixed the problem. Pic works stable. Though I would still like to get it to work without the additional resistance in the secondary circuit. This would make the circuit a bit more robust. I will try to get the current clamp and graph the primary current and solve the crystal problem.

Then you need a coil who's primary is wired to work on 12 volts. standard automotive 12v coils are made to operate on about 7 volts. The late modle GM HEI coils may in fact be 12v, I do not know for sure, the coil for an 1989 chevy caprice with 350 and F.I. is a square coil with a standard sparkplug tip secondary electrode and may serve you well, they cost about $45 US and make sparks in the lethal range, really big nasty blue sparks! Or check companies like ACCEL or Mallory or MSD for Hi Perf aftermarket coils, their tech support people can advise as to whether or not a primary resistor is required. BTW I have seen up to 60 KV from a GM HEI coil.

[zerosvarka]

Hi Everyone,
While I am ordering more hardware I have got current clamp and graphed the current draw. The clamp was set to 10mv/A and is clamped around all the earth wires just near the battery terminal. Top signal is the current and the bottom one is the trigger.

[skimask]

The late modle GM HEI coils may in fact be 12v, I do not know for sure

They are straight up 12V. The early coils for points type distributors also had an 'S' terminal on the starter. You hit the key, the coil gets it's juice direct from the battery since the starter is already killing off some voltage. You release the key, the coil gets juice thru the ballast resistor, knocking the volts down to 7-ish like you said. Keeps the coil from overheating.
The HEI coils can run 12V because of the inherent variable dwell (not to mention they're designed for it), whereas the older coils for points had a more-or-less constant dwell.

What's really fun is to hook a coil up to an 'old-school' turn signal flasher and a single spark plug gapped to about 1/4"+ and zapping your friends.

[zerosvarka]

Hi Everyone,

I use Usb to Serial adaptor to communicate between Pic and PC. I noticed that when I fire the coil without the high tension lead and spark plug (the spark jumps to the secondary earth wire that goes to the battery negative terminal) my Usb to Serial adaptor locks up and I have to reconnect it on the usb side. But if I communicate with stand alone Pic powered from a separate power supply and trigger the coil from my circuit everything works fine.

I tried to Debug Status and Pcon registers to determine the cause of the reset. When I disconnect power or ground the MCLR pin it doesn’t change the registers. I attached the code. Can someone have a look at it please