Despite being a fairly mature design, there are quite a few wrinkles in the kicker board. Furthermore, it seems to be much lower impedance than the last design and produces far more power (250V @ 360++A). This kills our silicon.
The current limit measures current through the silicon not through the solenoid. When using the circuit
- The IGBT switches on
- The IGBT switches off when a programmed current limit has been reached. Thanks to the flyback diode, that current still runs in the solenoid.
- Thus, when the IGBT switches on again (nearly instantaneously) it sees the current running in the flyback/solenoid + new current. This destroys the silicon
The solution is to measure the current running in the solenoid. I'll put up a modified schematic.
Also, there are a few problems associated with this section of the circuit
- Shunt resistor is underrated. Burned up one during testing.
- The gate resistors may have too little power though this is probably related to the failure of an IGBT during testing.
Barebones Layout Issues
The prototype was manufactured on a barebones process so some errors have been incurred as a result.
- Traces running under isolated power supply can be shorted on mounting feet of supply
- Solenoid connector interfere with underlying screw
- Battery connector interferes with underlying screw
- Fuses soldered into holders
- Jumper for the safe side logic power is far too large
- Space to add additional IGBT with separate gate drive resistor for each output
- The ability to easily disable the current limit by removing or adding a part