Spec'ing MOSFET power

Discussion in 'General Electronics Chat' started by Uber_Goober, Mar 9, 2013.

  1. Uber_Goober

    Thread Starter Member

    Jan 19, 2013
    I built a little circuit that uses a logic MOSFET to drive a solenoid. Its kind of funny this is the first PCB I've ever made, and the first SMD work I've ever done. So I layed out the board on my screen not really having any perspective of how small everything was. I just picked devices usually based on the first one that I found in the eagle library that had the right properties, laid them out on the board, and ordered the same package on digikey. This left me with a board full of 0402 and 0603 parts haha. You can imagine my surprise when the parts came in from Digikey and I realized how small they were.

    I had the boards made at BatchPCB and a solder stencil made at Pololu, and I reflow-soldered on a $20 hot plate. Believe it or not- 100% success. Not a single short or bad connection. Every single board works perfect. My arm is sore from patting myself on the back by now.

    But now I'm looking at some of these tiny parts. The solenoid is 220mA @ 12VDC. I'm just incredulous looking at this miniature MOSFET (and kickback diode) that I have driving the solenoid. Is it really adequate? So I'm running through the numbers to make sure I have it right. Let me add that it is working perfectly and my ultra precision thermal probe (my fingertip) doesn't detect any difference in temperature when the solenoid is on (note: the solenoid doesn't turn on very often in this device, but it does stay on for 1-2 minutes at a time).

    First I was going to switch to an SMA diode (a 1n4001 equivalent) for the solenoid coil kickback protection. I'm currently using a BAS3010A in a SOD323 package. But I looked at it- and all the specs on this diode are just better than a 1n400x and it seems perfectly spec'd for this application (30V, 1A, 470mv fwd voltage and very fast recovery). So I don't think I need to change the diode, even if it is almost invisible. On to the MOSFET.

    I'm using a DMG1012UW in an SOT323 package. Its rated for 20V and 1A. That should be fine but then I noted the power rating: 290mW. But (my first reaction) 12V @ 220mA = 2.64W. Thats almost 10X the rated max... Hmmm... why hasn't it burned up already? So I started researching SMD MOSFETS that handle 3 watts. They exist but these are also rated at about 8A. And the SMD packages are about 100x bigger than the SOT-323. Something wasn't adding up. So I had to learn a little bit more about MOSFETs, and a few minutes consulting Google taught me that the power rating is based on Rds-On and the current (using I^2*R).

    It looks like I'm comfortably in the safe zone on this MOSFET, but I wanted to be sure. Rds-On is dependent on the gate voltage and drain current. So that leads to my question after that long rambling build up....

    The Max Rds-On of the DMG1012UW is:
    a. 0.45 ohm (Vgs = 4.5V, Id = 600mA)
    b. 0.6 ohm (Vgs = 2.5V, Id = 500mA)
    c. 0.75 ohm (Vgs = 1.8V, Id = 350mA)
    (These are max values, the typical are 33% lower)
    (these are all the data points the specs give, no graph or anything)

    So calculating I^2R (0.22A) means I'm loading it with somewhere between 22mW and 36mW - very safely under the 290mW maximum.

    Can anyone tell me if I've done the calculations correctly? Does it look good to you? I just can't get over that I'm driving a 220mA solenoid with this itty bitty MOSFET and diode. Thats crazy.

    Oh and the standard disclaimer: if while reading this you start to think to yourself "this guy doesn't seem to know what he's talking about" there's a very distinct possibility that you are absolutely correct! I'm a software guy :)

  2. antonv


    Nov 27, 2012
    Looks like you got it. Mosfets are amazing h that way.