Shop OBEX P1 Docs P2 Docs Learn Events
Problems with Transistor heat — Parallax Forums

Problems with Transistor heat

brooka0brooka0 Posts: 13
edited 2011-02-03 10:23 in Robotics
hello,

I have an issue with driver transistor ckts. I am trying to build a simple ckt where the NMOS will act as a gate. I have a 12 Volt source and a 3 ohm load. The source and is connected to the load, the load to the drain, the source of the transistor is connected to ground. Everytime i build this ckt the transistor that I am using heats up and dies. I also need 4 amps of current to drive the load.
1ClBq.png

Comments

  • LeonLeon Posts: 7,620
    edited 2011-02-03 07:08
    Schematic?
  • RobotWorkshopRobotWorkshop Posts: 2,307
    edited 2011-02-03 07:42
    What transistor are you using?? Without a part # for the parts you're burning up we have no idea if you have a part that should handle that kind of current or if you are using something meant for small currents only. To be of any help we need to know:

    - Actual part # of the transistor/FET/whatever you are burning up

    - What is the 12V power source? Battery, wall wart, etc

    - What is the 3ohm load? Is it just a large resistor? Heating element? Motor? or some device that uses 12V and draws 4 amps?

    - What is controlling the switch and how is it wired?

    The more details you provide, the more likely you'll get useful answers and some help.

    Robert
  • brooka0brooka0 Posts: 13
    edited 2011-02-03 07:49
    Transistor is http://www.fairchildsemi.com/ds/FQ/FQPF13N06L.pdf
    enercell lead acid battery
    3ohm is a SMA wire
    haven't hooked it up to the board yet but i am planning on wiring to a Javelin stamp with a BOE, 500k resistor parallel to the gate and ground
    is it overheating because the gate is floating?

    I calculated wattage given off by the transistor on is only around 1.3 Watts, it is rated for 60V and around 40W so I don't see why it should be heating up without any voltage on the gate pin
  • LeonLeon Posts: 7,620
    edited 2011-02-03 07:53
    Schematic?

    What is an SMA wire?
  • brooka0brooka0 Posts: 13
    edited 2011-02-03 07:56
    Shape memory alloy wire. Does it really matter how the voltage is supplied to the MOS? There is no current flowing into the gate anyway
  • Mike GreenMike Green Posts: 23,101
    edited 2011-02-03 08:07
    Yes, it really matters how the voltage is supplied to the MOSFET. It really matters what the voltage is. People here tend to have a lot of experience and ask these questions so they can give proper advice.
  • LeonLeon Posts: 7,620
    edited 2011-02-03 08:08
    Schematic?
  • brooka0brooka0 Posts: 13
    edited 2011-02-03 08:12
    Sorry about that still new to these types of circuits. attached is a updated circuit showing the intended gate hook up. Sorry again if I offended.
    I build most of this circuit but stop because the transistor used gets really hot.

    XdU8h.png
  • Dave HeinDave Hein Posts: 6,347
    edited 2011-02-03 08:25
    This transistor isn't fully on until the gate voltage is about 6 volts. 5 volts should be OK, but if you have a resistor between pin 0 and the gate the voltage will be less than 5 volts. If the gate voltage were low enough where the transistor had an effictive resistance of 3 ohms it would be dissapating 12 watts of heat. Without a heat sync it would probably burn out.

    So the question is, are you really driving the gate at 5 volts?
  • brooka0brooka0 Posts: 13
    edited 2011-02-03 08:35
    I haven't really ran the program yet. I am definitely sure that the output for a pulse width setup on the stamp is defined as 5 volts. As i said before when I have been building the circuit I haven't been connecting the gate pin on the transistor to anything. I just tested an idea and connected the gate pin to the ground on the battery and the transistor did not overheat. I seem to think that because I was leaving the gate pin floating it was causing the transistor to conduct at a high resistance which would explain the heat experienced when i tried to pull the transistor. (to visualize the schematic look at the last picture and take away the pin connection)
  • Mike GreenMike Green Posts: 23,101
    edited 2011-02-03 08:36
    Let's see ... from the datasheet ... With a gate voltage of around 5V, the MOSFET has an on resistance of around 1/10 Ohm. With a 12V supply and 3 Ohm load, we're talking about a load of around 4A. That's 0.4V across the MOSFET for a power dissipation of 0.4V * 4A = 1.6W. That alone can cause the MOSFET to become hot without a heatsink.

    Now, I'm not sure about SMA, but most similar conductors have a temperature sensitive resistance where the resistance when cold is much lower than the resistance when hot. Are you measuring the resistance or is the 3 Ohm value the specification? If the cold resistance is much lower, the current through the MOSFET will be much higher and the heat generated will be a lot higher. I suggest you check out the cold resistance and the initial on current. You may need a heatsink.
  • Dave HeinDave Hein Posts: 6,347
    edited 2011-02-03 08:43
    brooka0 wrote: »
    I seem to think that because I was leaving the gate pin floating it was causing the transistor to conduct at a high resistance which would explain the heat experienced when i tried to pull the transistor. (to visualize the schematic look at the last picture and take away the pin connection)
    So is the gate pin floating or is it pulled down to ground using the 100K resistor that you show in the schematic? If it's floating the gate voltage could be in the partially on region.
  • brooka0brooka0 Posts: 13
    edited 2011-02-03 08:45
    hard measurements aren't really possible the composite alloy is hard to read on my multimeter. I am able to calculate from the datasheet that the amount used would give me 3 ohms. Also, the datasheet for the device says the device is rated for 24W power dissipation and can facilitate a continuous drain current of 10 amps.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2011-02-03 08:46
    With a gate voltage of 5V, the transistor's drain-source resistance is 0.14 ohms (max). With a 4A load, it will be dissipating 2.25W. Without a heatsink, that could be enough to overheat and self-destruct.

    Also, if you're trying to PWM the gate, your upstream circuit may not have enough instantaneous drive capacity to overcome the gate capacitance on each pulse and drive the transistor to full saturation. If that's the case, the gate will be seeing a lower effective voltage and the drain-source resistance will be even higher. This can be remedied with a MOSFET driver IC or with a lower PWM frequency.

    Another thing to look at is your wiring. If there are IR losses between logic ground and the transistor's source, it will see a lower effective gate voltage. It's always best to connect logic ground and the load's return supply at one point and one point only: right at the transistor's source.

    -Phil
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2011-02-03 08:48
    brooka0,

    The maximum power dissipation rating of 24W is at room temperature, i.e. with a heatsink. Above a die temp of 25C, you have to derate that 24W figure. Also, the on-resistance increases rapidly as the temperature increases, further exacerbating the dissipation issue.

    -Phil
  • LeonLeon Posts: 7,620
    edited 2011-02-03 08:52
    Just put the wire across the supply and measure the current through it. You will then be able to calculate the resistance when it is hot.
  • brooka0brooka0 Posts: 13
    edited 2011-02-03 09:08
    Ok i have some of the resistance readings, looks like there is a 4.5 ohm at cold state and a 9.6 ohm on state resistance of the wire. It still contracts even though less than 4 amps is activating. This works out to 2.8A through wire while off and 1.3 Amps while on. With the Rds max at .14 i got the wattage dissipated in the mos as 1.1W from the off state and .24W on state
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2011-02-03 09:12
    Why would there be any current flowing when the transistor is off? What are you not telling us about the "off" state?

    -Phil
  • LeonLeon Posts: 7,620
    edited 2011-02-03 09:12
    That makes sense. The resistance increasing when the wire gets hot means that you won't get "thermal runaway".
  • brooka0brooka0 Posts: 13
    edited 2011-02-03 09:15
    Thanks for all of the help!
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2011-02-03 09:36
    I'm not sure we solved anything yet. Please let us know what you had to do to make your circuit work.

    -Phil
  • brooka0brooka0 Posts: 13
    edited 2011-02-03 09:53
    The gate wasn't pulled down it was in the partial on position. I have pulled it down with the 100k and everything is ok right now I am still waiting on the last parts for other regions of the system which I have tested.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2011-02-03 10:05
    Oh. It's always helpful when seeking assistance that your description/schematic match your actual circuit. Anyway, 'glad you got it working! :)

    -Phil
  • Dave HeinDave Hein Posts: 6,347
    edited 2011-02-03 10:10
    Phil, it would be too easy if they told us everything. Where's the fun in that? 1087.gif
  • brooka0brooka0 Posts: 13
    edited 2011-02-03 10:23
    Yeah sorry about the run around thanks for the patience!
Sign In or Register to comment.