1
\$\begingroup\$

Ok so I have burned about 10 MIC4605 (Datasheet here ) gate drivers this weekend. I am unsure as to how they should be PROPERLY wired.

Also have burned 8 out 10 MOSFETS (Datasheet Here )

Page 1 of the datasheet shows a typical application which is interesting because the motor I am using is 140mA @12V

enter image description here So I wired it just like that...Magic Smoke came out I tried to PWM the low side mosfet of one side of the bridge while the other side had the high mosfet activated. Things blew up.

So i figured this must be some sort of circuit diagram where "things are assumed" like schottky Diodes. After some research I found this on Microchips own website and the funny thing is the circuit diagram looks a lot different than the datasheet.

Sample circuit

So now I have this wired up and I am down to 1 Gate driver left alive after burning 9. This circuit seems to work. I am not using a 24V bus, My motor and my gate driver are both being fed by a 12V coming from my power supply. In the circuit above I am not using the lowside mosfet, instead I am just testing the high side mosfet and my motor's other lead goes directly to ground.

This works just fine at 12V and I am PWM'ing perfectly fine and works well, until I turn the voltage up to 16V and then my driver dies.

My only guess is that BACK emf becomes too large and kills something, but I have all the diodes in the schamtic in their rightful place..

I am PWModulating at 32Khz.

My ultimate goal is a Full Bridge Motor driver with all N-channel MOSFETS and having atleast 2 Mosfets in parallel for each high side and low side, but i cant even get past one mosfet at a conservative voltage.

\$\endgroup\$
  • \$\begingroup\$ Are you using the -1 version or the -2 version of MIC4605? \$\endgroup\$ – Tom Carpenter Jul 1 '18 at 22:15
  • \$\begingroup\$ Ive managed to burn up both, i Had 10 of the 1 version and 4 of the 2 versions \$\endgroup\$ – Edwin Fairchild Jul 1 '18 at 22:15
  • 1
    \$\begingroup\$ Also, have you tried with something like a 1K resistor as a load instead of the motor? Just to rule out it being some form of EMF or over current issue. \$\endgroup\$ – Tom Carpenter Jul 1 '18 at 22:23
  • 2
    \$\begingroup\$ Cross-conduction due to poor dead-band design is chief failure mode. Inductance delays included \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Jul 1 '18 at 22:25
  • 1
    \$\begingroup\$ Are you using a GROUND PLANE? The inductive kick of 0.14 amp switching in 10 nanoseconds in 100 nanoHenry (4" of wire) is V = L * dI/dT = 100nH * 0.14A / 10nS = 14 volts. If your local bypass capacitor (on the 0.14 amp rail) is a meter away, you'll get 140 1-4-0 volts. \$\endgroup\$ – analogsystemsrf Jul 3 '18 at 3:55

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.