P Type MOSFET Driver Problem

Discussion in 'General Electronics Chat' started by Tobias, Nov 2, 2011.

  1. Tobias

    Thread Starter Active Member

    May 19, 2008
    158
    0
    This used to work quite well. Instead of the MCP1403 driver, I had a Clare 1035 relay to trigger the P type MOSFET with a PIC. Now I try a MDP1403 connected like the attached schematic and the drivers overheat and fail.

    I thought maybe I was overworking the MOSFET so I soldered another one on top of the first driver, same result.

    Spec sheets on the MOSFET and driver are attached.

    Any ideas?
    Thanks in advance
    Toby
     
  2. crutschow

    Expert

    Mar 14, 2008
    13,023
    3,236
    I don't see an obvious problem.

    Is Vss (GND) connected to the PIC common? It must be for proper operation.

    Can you look at the output signals of the MCP1403 with an oscilloscope?
     
  3. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    You don't show resistors between the MCP1403 and the MOSFET gates. Without a resistor, the C of the gate and the L of the wiring make a series LC circuit that will ring like a bell. Having a small value of resistance in there will dampen the ringing.

    You don't show bypass caps across the MCP1403's power pins.

    You don't show bypass caps from the MOSFET source terminals to Gnd/Vss. Inductance in the wiring to the source terminals can also allow ringing.
     
  4. Tobias

    Thread Starter Active Member

    May 19, 2008
    158
    0
    Yes VSS is connected to the PIC common. There are also by-pass caps on the MCP VIN/VSS. I am using this to activate solenoids. It isn't a PWM scenario, either on or off. On state is for a couple seconds at the most.

    The board worked great when I had the cpc1035 relay activating the MOSFET. I swapped it out for the MCP part.

    I am really not sure what to do. One thing I didn't mention in my first post is that without a load on the MOSFET, the MCP works just fine. When I have a solenoid hooked to the Drain, the MCP will then fail.

     
  5. crutschow

    Expert

    Mar 14, 2008
    13,023
    3,236
    You didn't answer my question about the signal levels. Do you have an oscilloscope?

    If the signals are slow then you can just measure them with a multimeter.

    Without such information it's difficult to know what the problem is.

    Have you double-checked all the wiring?
     
  6. Tobias

    Thread Starter Active Member

    May 19, 2008
    158
    0
    On my scope the output of the MCP is 12v or 0v. When INA is 5v from the PIC, OUTA is 0V and when INA is 0V, INA is 12v

    The circuitry is on a PCB all at that looks to be ok. Like I mentioned, all is good with the MCP activating the MOSFET without a load on the MOSFET.
     
  7. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488

    Well, the solenoid part must be the key to the problem, I cannot figure out why, though. Even if the solenoid doesn't have a freewheeling diode I don't know why this would affect the driver.

    But I would like to know:

    1. Does the solenoid have a freewheeling diode?
    2. Can you post a picture of the PCB layout including the solenoid and power supply showing where exactly is what connected to? Like distance from power supply to MOSFET / driver / VSS to solenoid
    3. Did you try to put an equivalent resistive load and to see if it still heats up the driver IC?
    4. I didn't see if you answered Sgt's question regarding the gate resistor. With your o-scope try to trigger on the falling edge of the gate signal and see how it looks like (negative spikes or ringing). Also with an o-scope measure Vdd to Vss directly on the IC pins at Gate Turn-Off. (trigger on gate turn-off and monitor Vdd/Vss on the second channel)
     
    Last edited: Nov 2, 2011
  8. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    The CPC1035 has very slow turn-on and turn-off times; 1mS and 2mS. I suggest that if you add 33 Ohm to 1k Ohm resistors between the gates of the MOSFETs and the MCP1403 driver that much of your problem will go away.

    If the gates are ringing, your MOSFETs will be spending a lot of time in the partially conducting state, generating heat.
     
Loading...