Help Needed: PIC Full Bridge PWM + Bridge Driver not working

[picbits]

I found it very very easy to fry the HIP controllers when things went wrong. Have you checked to see if they are getting hot (or your FETs) ?

I finally managed to get mine going with the last 2 I had knocking around before I fried those as well by accident and went for a P channel high side using transistors in the end.

Not as nice but a bit more robust.

 

[Speakerguy]

Everything is staying cool. Last night I blew a FET, but replaced those with new ones this morning. The low side outputs of the HIP are good, so either I blew the high side outputs or I'm doing something wrong. It's all SMT and the board doesn't even get warm.

I hate finicky semiconductors. Of course what I consider finicky means "doesn't withstand with my screwups"

Thank you guys for the help. I do my best to provide good info, so thank you for the good responses.

 

[Sceadwian]

FETS are not 'user friendly' simply handling them without basic static precautions will easily fry them, the voltage required to kill the gate doesn't even require a noticeable 'zap'

 

[Ubergeek63]

how are you driving it? the high side FETs can NOT be on steady. if either side is going to be turned on at DC it needs to be the bottom ones and do not expect the other high side to be ready switch instantaniously.

 

[Speakerguy]

Damn, you're right! The high-side bootstrap will never recharge with the PIC only modulating the low side switch in one direction. It's so basic I can't believe I didn't see it. Sigh. Now I get to figure out how to fix this the easiest way. Maybe use the PIC in half bridge mode (P1A and P1B outputs as complements of each other) and use those to drive the full bridge? I hate it when I make a mistake like this.

 

[Speakerguy]

FETS are not 'user friendly' simply handling them without basic static precautions will easily fry them, the voltage required to kill the gate doesn't even require a noticeable 'zap'

I've heard this from many people, but I've yet to kill a FET due to mishandling. I'm young compared to most, only been doing electronics for about 3 years now, so perhaps today's devices are more rugged? I really don't take any care wrt ESD and I've never had a problem with anything. I do live in a fairly humid part of the US so maybe that helps too.

 

[Sceadwian]

I've killed a few through 'normal' handlding. CMOS chips are pretty resilient now days but power fets are still susceptible to static shocks, even the kind that you don't feel. There are so many factors that determine an indivudal persons static shock risk and the environment they're in is a big part of ti, so some people will have a golden touch and some iwll have a 'black' touch =)
Also Speakerguy, do you store your fets properly like in conductive foam or do you just toss them in a plastic drawer? Makes a big difference.

 

[kooloosj]

JK

Hi Guys,

Pays to run your scope off an isolating transformer when probing around H-bridges. The Probes ground lead will catch you out more times than you think!

JK

 

[ericchee]

hi,

i see that you solved the problem. Do you mind to explain further on the problem. I read the suggestion but i do not get it. i also not quite understand of bootstrap function. i did googling on it but still no idea. I just build a driver using l298 and now i am looking at other driver. thank you for your attention

rgds,
ericchee