• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Subscribers 62 subscribers
  • Views 1527 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

CSD18502Q5B: thermal characteristics for calculation of max dissipation

domenyko
Genius 9655 points
Part Number: CSD18502Q5B
Other Parts Discussed in Thread: CSD18509Q5B, , CSD18510Q5B, CSD18512Q5B

Dear all,

one of my customer is using 4x CSD18502Q5B in a bridge configuration  and 1x csd18509q5b to power the bridge. The MOSFETs are used to drive a motor. In order to calculate the max dissipation he needs more details on the thermal paramerters of the MOSFET. In few words the motor will be ON just for few hundreds of ms, in this time the envelope of the PWM is mostly  a pulse. Is there any tool / model / calculator to avalaute the power dissipation ?

thanks,

Domenico 

  • 0
    TI__Mastermind 18200 points
    Domenico,
    There is currently no tool for calculating losses for motor control, but for the pulse in question, you can just do an I^2*R calculation to determine what the losses will be during the pulse.

    What would be the peak current and duration of the pulse in question?
  • 0 in reply to Brett Barr1
    TI__Genius 9655 points

    Hi Brett,

    The motor will be ON for 350ms, the switching frequency of the FET of the bridge is 20KHz, so the pulse will last 7000 periods. The current at max efficiency is 18A while the stall current is 125A, the operative current can be as high as 50A.

    Regards,

    Domenico

  • 0 in reply to domenyko
    TI__Mastermind 18200 points
    Domenico,
    At 20kHz, while your switching losses can't be neglected, generally they will account for <1/3 of your total losses. And your conduction losses are going to be way to high if any of the FETs can be exposed to peak currents of 125A for up to a few hundred ms.

    Even 50A RMS is going to be way too much current for a QFN. Take the CSD18502Q5B, with typical resistance of 1.8W. Let's say the junction heats up to 100 deg, so based on Figure 8, we would expect the typical resistance to be 1.8*1.4 = ~2.5mOhm. Now we have Pcond = 50*50*2.5E-3 or 6W, way too much for a PQFN to handle. And that's not even looking at the peak current pulse of 125A.

    Sounds like the current levels you are driving will require either paralleling multiple lower resistance FETs or using a TO-220 / heat sink combination. Unless I am misunderstanding something about the application and the RMS current passing through the FETs.
  • 0 in reply to Brett Barr1
    TI__Genius 9655 points
    Hi Brett,
    thanks for the calculation. My original proposal for the customer was to use 2 MOSFETs in parallel, (total of 8 for the bridge) and 2 in parallel to enable the bridge.
    I'll push again for this solution.

    regards,
    Domenico
  • 0 in reply to domenyko
    TI__Mastermind 18200 points
    Domenico,
    One more thing I would point out, both of these devices have newer "replacement devices." The CSD18509Q5B has a better lower resistance device, the CSD18510Q5B (typical 0.79mOhm Rdson) that is going to RTM this week, which is basically the same price.

    Then the CSD18502Q5B has the recently released CSD18512Q5B -which is both better resistance and a little more cost effective, so it should be a win win to use these proposed devices as well, which should reduce your conduction losses and run a bit cooler.
  • 0 in reply to Brett Barr1
    TI__Genius 9655 points
    Hi Brett,

    good hints. Since the customer is sensitive to the price and these decvices have better performances at the same price this will be a good point for us.
    regards,
    Domenico