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.

CSD18540Q5B: How do I determine if MOSFET can handle a single current pulse in saturation mode at ambient temperatures other then 25°C? Requirement: ID = 4.5A, VDS = 16.5V, T = 20ms, case temp = 71°C or 105°C

Part Number: CSD18540Q5B

I’m trying to figure out if a MOSFET can handle a single current pulse for a given time duration, at a given VDS and at a given ambient temperature. The CSD18540Q5B for example.

I’m not sure I can just look at the SOA curve because I’m not sure if this curve only applies to devices at 25°C. How do I figure out if the device can handle the pulse (single) at 71°C?

Pulse handling requirements:

  • Single Pulse (not repetitive)
  • case temp = 71°C or 105°C.
  • Time duration = 20ms
  • Current = 4.5 Amps
  • VDS = 16.5V

Thanks,

Nick

  • Hi Nick,
    Thanks for your inquiry. We need to use both the transient thermal impedance curve and SOA curve from the datasheet. Starting with the transient thermal impedance curve, if I draw a straight up from the 20ms pulse duration, I conservatively estimate the normalizing factor is about 0.7. I can use this to calculate delta-Tj = 0.7 x 0.8degC/W x 16.5V x 4.5A = 42 degC. The estimated Tj = 113degC @ Tc = 71degC and 147degC @ Tc = 105degC. We should be OK if we assume there is no thermal runaway or local hot spot. However, there is not much margin for Tc = 105degC.

    Using the SOA curve, I estimate that the max ID @ VDS = 16.5V is 7.5A. However, since we are in the saturation region, we should derate the SOA curve. Since we don’t have a 20ms line, I used the 10ms line and derate by a factor of (150deg (TJmax) – 71deg(Tj)) / (150 – 25deg (measured)) = 79 / 125 = 0.632. Mutliply the 7.5A by that factor and you get 4.7A so it’s pretty close at this temperature. Definitely wouldn’t recommend doing this pulse at 105deg where the derating factor would be 0.36 and the SOA curve would get derated to 2.7A.
  • Thanks John,

    So it sounds like I can assume the part has enough thermal capacitance such that the case temperature is not changing during the pulse since we are using a constant 71°C case? Would this be true all the way up to a .1 second pulse?

    Does the following look correct?

    I think this part has a max junction of 175°C.

    So for 71°case: (175°C-71°C)/(175°C-25°C) = .693 and for 105°case: (175°C - 105°C)/(175°C-25°C) = .466

    So per SOA max current for 71°C case = .693 x 7.5A = 5.19A

    per SOA max current for 105°C case = .466 x 7.5A = 3.5A

    What was your math for getting the 7.5A?

    Thanks,

    Nick

  • Hi Nick,
    The assumption is that the case temperature is fixed and does not rise appreciably due to the short pulse. It also assumes a single pulse event. If this is repetitive, then we'd need to know the duty cycle to see how that affects the calculation.

    You're correct that this part is rated to 175°C. I incorrectly assumed 150°C. Your calculations look correct as well. To get 7.5A, I used the SOA curve, drew a vertical line at VDS = 16.5V (approximately) and a horizontal line where it crosses in between the 10ms and 100ms SOA lines. It's an estimate!