0
\$\begingroup\$

I use this 40V, N-Channel NexFET Power MOSFET http://www.ti.com/product/CSD18504Q5A

in

schematic

simulate this circuit – Schematic created using CircuitLab

The solenoid takes 25A peak current. (Rated drain current is 50A)

In this setup, MOSFET overheats and shorts after some time. What is wrong with this setup? Please advise.

P.S. I am a noob.

\$\endgroup\$
  • \$\begingroup\$ Actual drive voltage on the MOSFET? Also PCB layout or photo of your arrangement. \$\endgroup\$ – Spehro Pefhany Mar 17 '16 at 16:11
  • \$\begingroup\$ 5V to Gate is applied \$\endgroup\$ – Karthik Nishanth Mar 17 '16 at 16:13
  • \$\begingroup\$ What kind of monster solenoid dissipates 300 W? \$\endgroup\$ – jms Mar 17 '16 at 16:24
  • \$\begingroup\$ @jms This is used to actuate a gear switching assembly. The ON pulse lasts for about 350ms. \$\endgroup\$ – Karthik Nishanth Mar 17 '16 at 16:25
  • \$\begingroup\$ Link to the heatsink data would be useful. \$\endgroup\$ – Brian Drummond Mar 17 '16 at 17:53
2
\$\begingroup\$

With a 4.5V gate drive this MOSFET has a maximum Rds(on) of about 10m\$\Omega\$ at 25°C, perhaps as much as 15m\$\Omega\$ at a safe higher junction temperature. If we use that value, power dissipation can be in the 10W range. See the datasheet.

That's rather high for such a tiny package and may be virtually impossible to achieve with a reasonable setup. At the reference 50°C/W (see below note as to specifics- that's 1 square inch of 2 ounce copper- not insignificant) it obviously will burn out rather quickly (500°C rise is not going to be possible). At 125°C/W (second layout below) it will probably fry in seconds.

You need a MOSFET with much lower Rds(on) and/or a MOSFET that is in a package capable of dissipating more power.

Note that if you're feeling fooled by the datasheet- that is an 'absolute maximum' rating- not something you should be designing to. Constraints often turn out to be other than the first number you run across- in this case it's thermal. All constraints must be simultaneously respected or you may have a bad day.

(2) Device mounted on FR4 material with 1-inch2 (6.45-cm2), 2-oz. (0.071-mm thick) Cu.

enter image description here

\$\endgroup\$
  • \$\begingroup\$ Thanks, this answers my question. I will wait for more opinion before marking yours as answer. \$\endgroup\$ – Karthik Nishanth Mar 17 '16 at 16:23
  • \$\begingroup\$ This MOSFET was a TO-220, attached with a heatsink (the one that you bolt to the hole on TO-220). Are there any packages that could dissipate more power than this setup? \$\endgroup\$ – Karthik Nishanth Mar 17 '16 at 16:24
  • \$\begingroup\$ That does not match the MOSFET in the datasheet you linked and I have idea what a 'TO-22' is. Of course if you have a totally different part than in the question the specifics will vary- just use the methods I outlined and do the calculations. You can get parts that will dissipate 1000W comfortably. \$\endgroup\$ – Spehro Pefhany Mar 17 '16 at 16:27
  • \$\begingroup\$ Sorry, that was a TO-220. The same MOSFET is given in different packages by TI. \$\endgroup\$ – Karthik Nishanth Mar 17 '16 at 16:28
  • 1
    \$\begingroup\$ @KarthikNishanth - And the fact that it takes a while to overheat is an excellent indicator that whatever heatsink you're using is marginal. \$\endgroup\$ – WhatRoughBeast Mar 17 '16 at 16:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.