DC Motor & IGBT running HOT!!!

Discussion in 'The Projects Forum' started by AresROC, Feb 17, 2011.

  1. AresROC

    Thread Starter New Member

    Jan 3, 2011
    19
    1
    Since my treadmill's IGBT transistor burned out and I replaced it with a new one, the DC motor & its driving IGBT has been running very hot! I put a CPU heatsink on the IGBT, but both the motor and IGBT would become hot to the touch after very moderate use. Furthermore, the motor has noticeably less torque now. :mad:

    Apparently something else is wrong in the TD-1600N motor control board, which probably burned out the original IGBT in the first place.

    Does anyone have an idea where I can begin troubleshooting? I believe the motor itself is fine, and I don't have access to an oscilloscope to check the IGBT's gate signal...

    TIA
     
    Last edited: Feb 17, 2011
  2. R!f@@

    AAC Fanatic!

    Apr 2, 2009
    8,750
    760
    Check the big caps for high ESR
     
    AresROC likes this.
  3. AresROC

    Thread Starter New Member

    Jan 3, 2011
    19
    1
    Don't have an oscilloscope & ESR meter... :( hopefully they are not too expensive.
     
  4. edgetrigger

    Member

    Dec 19, 2010
    133
    19
    I did not get why you had to add a heat sink, was it not there at the first place in the board?

    Hope you have used the same IGBT as the one which was blown out.

    If you have a multimeter check the voltage across the motor when it running, the voltage has to be around the rated voltage of the motor which is 90 vots as per the schematic.
     
  5. AresROC

    Thread Starter New Member

    Jan 3, 2011
    19
    1
    The original IGBT was sleeved in a soft rubber tube and attached to the frame of the motor control board for cooling. I thought that may have not been sufficient. It is now attached directly with arctic silver (but collector electrically isolated) to a CPU heatsink so I can use some draft from the motor's cooling fan.

    Top row from left to right: FEP30 Ultrafast Rectifier, IGBT, BR356
    [​IMG]

    Yes I managed to find identical IGBT. It is the Toshiba GT60M303, 900v 60A continuous, 170w. However, during the process of troubleshooting I did replace the 3 optocouplers, as well as replacing SMD MMBT8050/8550 with through-hole SS8050/8550 transistors.

    Below is a graph of motor speed vs voltage. The max speed is 12mph but I usually "walk" it at 3mph... :rolleyes:

    [​IMG]

    There are 3 pots on the board labeled "Speed" "Limit" & "IR". Not sure what IR is, but it seems Speed is speed calibration and Limit is current limit.
     
    Last edited: Mar 7, 2011
  6. R!f@@

    AAC Fanatic!

    Apr 2, 2009
    8,750
    760
    Try replacing the Big capacitor and check.
     
  7. AresROC

    Thread Starter New Member

    Jan 3, 2011
    19
    1
    Thanks, it would be cheaper too! :cool: I sure would like to replace the 85°c CapXon caps with some Panasonic 105°c 3000hr caps.

    So far I have turned up the "IR" pot a little, as well as the Speed pot. The tread has more power now and seems to draw a little less amp with no-load... I am guessing the higher speed makes it more efficient, and it seems adjusting the "IR" controls the torque output under load.
     
  8. AresROC

    Thread Starter New Member

    Jan 3, 2011
    19
    1
    The "IR" & "Speed" POT has no effect on the efficiency, since it still runs very hot (pun intended). Time for Digikey/Mouser to deliver a few more parts! hopefully they have MMBT8050/8550's...
     
  9. BillB3857

    Senior Member

    Feb 28, 2009
    2,400
    348
    The IR control on a DC drive compensates for the voltage drop due to winding resistance in the armature. In an ideal situation, speed is controlled by armature voltage compared to a fixed field strength. The counter EMF of the winding due to speed, approaching the applied voltage regulates the speed by reducing current, which reduces torque to a value that maintains a balance. Without IR compensation, the voltage drop attributed to the resistance of the windings tends to reduce accuracy of that balancing act.

    Google is also your friend..... www.[B]reliance.com[/B]/prodserv/standriv/appnotes/d7728.pdf
     
    AresROC likes this.
  10. AresROC

    Thread Starter New Member

    Jan 3, 2011
    19
    1
    Thanks Bill, this document explained it very well. I will do some search on the "LIMIT" pot and see what that does.


    I replaced the 200v 1500uF large cap with a Panasonic unit, but everything still runs pretty hot! I suppose it makes sense things run hot since there are 10~20 amps going through the motor for an extended amount of time. The treadmill seems to survive just fine for now! :)
     
  11. DrNick

    Active Member

    Dec 13, 2006
    110
    2
    There is no pull down on your IGBT gate which is probably causing it to switch off slowly, thus increasing switching loss...Do you know if this circuit design ever operated acceptably?
     
  12. AresROC

    Thread Starter New Member

    Jan 3, 2011
    19
    1
    The circuit was working OK, if not very efficiently. I wish if there was a way to improve it.

    To clarify, attached is the datasheet for the GT60M303 IGBT.
     
  13. DrNick

    Active Member

    Dec 13, 2006
    110
    2
    The gate capacitance of the IGBT has to discharge through whatever leakage is available through the transistor (2T) that is driving the gate. This is presumably high. Th therefore when the PWM signal turns off, the gate will slowly decay to zero. during this slow decay, which is likely a significant portion of the switching cycle, the IGBT is still on, conducting the full motor current. During this time the voltage from collector to emitter will increase as the voltage decays. Power = V * I, so as the motor current continues to conduct the power dissipated in the IGBT increases. The faster your switching frequency, the more pronounced this effect will be. I would suggest driving the IGBT with either a complementary emitter follower or a dedicated gate drive IC to provide faster turn off.

    This could also be causing your machine to heat up if excessive peak currents are being provided by the poor switching. Look up "gate drive circuits" on google; this may be able to shed more light on the topic than my short blurb...
     
    AresROC likes this.
Loading...