Sunday, February 13, 2011

Tips on Testing MOSFETs

Imagine the flow direction
Source to Drain

This is how I test FETS 
What I will try to do here is to provide a little insight on how to test MOSFET's in circuit the easiest way I can think of. Essentially the FET acts as a switch, and the GATE either opens or closes the switch. When power is supplied to Pin 4 (usually the GATE) it will either open or close depending on the type of MOSFET it is. The arrows shows the direction for reference when using a meter to test.  

Setting the meter in DIODE mode.
On the N-Channel FET place the NEG lead on Pin 6/7 and the POS lead on Pin 2/3. You should get a reading of some value ~100 to 500 or higher.

Imagine the flow direction
Drain to Source
On the P-Channel FET place the POS lead on Pin 6/7 and the NEG lead on Pin 2/3. Again you should get a reading of some value ~100 to 500 or higher.
The reason I say pins 2/3 and 6/7 is because they are common, and it's less of a shorting risk than saying pin 3, which could slip and give you the wrong reading if pin 4 is touched or blow it if it has power on it (from battery or charger).

Most companies make it fairly easy to identify P-Channel from N-Channel parts by part number.  This is usually the case with U.S. companies.  National Semiconductor, Fairchild and others. But it is not a hard and fast rule with some Chinese manufacturers, and may need to check the datasheet before assuming they will be the same.

ODD number MOSFETs are considered to be P-Channel
EVEN numbered MOSFETS are usually N-Channel. 
Example: FDS6679 would be a P-Channel and FDS6690 would be a N-Channel MOSFET
FDS stands for Fairchild part
SI stands for Siliconix/Vishay part
AO stands for Alpha and Omega Semiconductors part
IOR  stands for International Rectifier part (Their numbering system may not adhere to above numbering system)


  1. see dat my blog

  2. see dat my blog

  3. Reading mosfets in circuit can be tricky. Make sure the CPU is removed or many n-channel fets will read shorted. A question I have however, is why is it some in circuit tests reveal partial forward and reverse voltage (400-500). Even with the fet lifted on one side, testing the pads from source to drain and vice-versa have these same readings. It's not a complete short... and its not completely open either. Once you move past the charging circuit, its extremely hard to trace (at least it is for me). Any ideas on how or why this is happening? It drives me crazy probing these things sometimes. Also, is voltage across the fet determined solely by the gate voltage? I ask because sometimes I've read 19v at drain while only 5v is at the gate but yet on others, the drain voltage is equal to whats being applied to the gate.

  4. MattBrad2: The FET should only read one way in diode mode, since it basically a diode that is turned off or on. The GATE voltage may be different based up on the part specs. Such as the charge circuit, may use 12v FET gate to verify that POWER IN is correct. If the power is too low, it will not charge. This also accounts for incorrect power supplies not working or able to turn on particular models. FETS also come in different current ratings, so a DV6000 charger will not charge a DV9000 because a 65w power supply will not charge a 90w unit. But will run. If it does charge the battery, the charge will not have sufficient current to the LiION particles. So you only get 1/2 hour or more use on battery alone.

  5. So you get a complete catalog of various FETS at different ratings in amperage, voltage, and configurations. P-Channel, N-Channel and dual configurations based up requirements of the circuit, and trace lines. See examples of SI4914, SI 4812, and SI4916

  6. Sir, I just want to say that we started a PC repair shop and you have helped us infinitely. I can't express enough gratitude to you for sharing your knowledge. Thank you!

  7. Thanks Shawn. Hopefully I can save you some of the headaches that I had.

    Feel free to chime in.

  8. Hello just curious if an ao4413 can be a suitable replacement for an ao4433 I had no power or lights replaced the pq101 and now the charger light blinks and still no power supply light or power up. Thanks your bolg is very helpful.

  9. Andy: They have considerably different values. When the rest of the circuit has set the R/I by resistors and caps, it should be swapped with an equivalent rated part. So in a word "NO"

    AO4413 and the other is AO4433

    Hold shift key to open in new window.

    You can also download the datasheet if need to print out.

  10. Thanks for the response, I Will get an ao4407 out of another board if I can find one and post my results.


  11. I replaced With an AO4407, still the say results, Something weird happens now, when the plug is unplugged the led on the dc jack blinks real fast.

    Thanks for the help.

  12. Andy: With your meter check the orange/yellow ceramic capacitors in the area, you may find a short there.

  13. Hi Lindasy and thanks for the kind words. Could you contact me directly. We may have something to discuss, and/or verify that you are a real person and not a 'hacked' account. I just may be overly cautious.

  14. hi
    i have laptop hp dv9000, with lines and scratch on screen, i heated it (gpu) and works fine for one day, but the next day another problem is shown:
    when pressing power it works for just 3-4 seconds then turn off
    charge is ok
    when removing ran it stays on with sound indicating no ram.
    any help??

  15. whaT a nice job sir,.

    thanks for sharing,.

  16. Philip1980: If you have a DV9700 series, then you are not heating the GPU properly. I insist that flux is used. I has several purposes.

    1: helps to spread heat evenly
    2: removes oxidation
    3: assists in heat-bridging from solderball (from GPU) to contact pad (on motherboard).

    When no flux is used, the solder may not 'jump over' and make contact with board, or jump from board to GPU. With the number of balls under the chip, it is difficult to get it enough heat evenly before scorching the topside of the GPU essentially damaging it.

    I guess in a mental image it is similar to the way magnets attract to each other when too close. The same thing happens when two heated solder joints 'bridge' and connect to each other. Flux helps bridge that space between them and makes the contact happen.

  17. Thanks Life2087 for the kind words. Just trying to help others see things a little clearer. With some training, it's not that complex to repair most problems.

  18. Also Phillip: if you had lines on the screen and a reflow cleared them up one day, but the next day the system was down and would not start, it was an improper reflow in 2 ways.

    1: it only warmed the video memory to make a minor connection.
    2: the GPU was not at fault, but instead it was the video memory.

    This happens a lot on the DV95xx, and DV97xx series because it has a dual purpose GPU with southbridge combined.

  19. Great pic with mosfet function

  20. Thnx for helping us all out. I have a dv7 with the exact power issue you described in a different post but not the same mosfet. The diode test on pu102 fails... Looks like part no 87743. Could this be the problem? I can't find schematics anywhere...

  21. im newbie on laptop repair and i found your blog lots of great tips for newbie like me...thanks you so much sir....

  22. i only get 100 in doide mode.. Is that short on fds6679 thnx..


I cannot repair laptops for you via the internet, I can only answer questions related to posts. This is because I have not had experience on every motherboard that is out there. Bear with me.

If you do need it repaired. Contact me for quote (US/CDN Only)

Note: Only a member of this blog may post a comment.