PIC16F15323 Keeps Blowing
I have a circuit in which the aforementioned PIC keeps blowing for no obvious reason. I'll explain the four major parts which could potentially have an issue and see if anyone can spot a mistake.Area 1
The first is the power supply. It can rise to 100v AC but is current limited to 500mA. To deal with this I have an AC FET disconnect circuit with a 2.4kohm bypass. The AC goes through the usual bridge rectifier. The PIC controls the FET disconnect which enables at 33v and it's implemented purely in the PIC's peripherals. First though there is a SMAJ28A TVS after the rectifier. It has a clamp of 7A at 44v but initially starts at 28V. So what happens is the TVS begins absorbing transients and warms up until the 500mA clamp point is at 33V. At that point the FET disconnect kicks in and now the voltage is going through the 2.4k resistor so the TVS cools down. This part of the circuit is well tested via oscilloscope and seems to work flawlessly. The TVS is only luke warm.Area 2
The next part is taking the circa <33v down to 5v. I do that with a simple transistor voltage regulator. As the PIC isn't driving much I use a 56k resistor, a 5.6v zener and a MMST3904 transistor (40v rated). The transistor has a maximum gain of 300 so the maximum current draw I compute to be 0.03A at 5v (approx) before voltage drop occurs. Area 3
I use the ZCD on the 100v AC. As there's a 6x limit I use a 51k resistor which limits operation to between 6-36v (roughly). However, after the 36V the ZCD is in "overload", that is it's current regulators can't keep up. I imagine the PIC's protection diodes kick in which the spec says are rated to 20mA. 100v and 51k is only 0.0019A so well within the limit. Area 4
The bypass is done using an isolated driver. I use the PIC to generate a PWM which is half wave rectified. See the attached picture. The only issue I can see here is when the FETs gates are discharged the PIC has a low path of impedance until they charge. These are very small FET's though with minimal gate charge. I can keep them active considering the 100k discharge resistor using a 125KHz square wave. The diodes in the circuit are 100v rated.
This is the only area I can think which could be the problem. Could a high voltage be going through D83 and C82, taking the PIC pin way above the absolute?
post edited by acharnley - 2019/09/20 13:07:41