Facebook
Google
GitHub
Linkedin
If you are a hobbyist and/or lack good troubleshooting skills, you should buy from reputable suppliers so you don't waste time with counterfeit/reject components.
If the MOSFETs have been subjected to ESD, they're likely to exhibit abnormally high gate leakage current (if they work at all). That's easy to verify with an ammeter while applying the maximum gate-source voltage allowed (both polarities).
Manufacturer data indicates that the relatively high gate capacitance of power MOSFETs allows them to absorb some static charge but it's still wise to use proper handling precautions.
https://forum.allaboutcircuits.com/...osfets-applicationnotes-v01_02-en-pdf.320871/
You need to test to insure that the devices function (turn on and off, and are the right polarity). To do that, I use a PWM circuit at the maximum rated current, or the highest current you can manage. That test can also be used to get an approximate on resistance. On resistance could read higher than the datasheet specification because to measure it accurately, you need to follow the manufacturers specifications. For IRLZ44, that's a pulse width ≤ 300uS with a duty cycle of ≤ 2%.
I used a single comparator PWM, so the frequency varies a bit. Since pull-up isn't active, that can affect the on resistance measurement. Waveform here.
What I found measuring on resistance was that an analog scope was better than a digital. On an analog scope, you can overdrive the scope input and increase resolution to be able to measure on resistance. The digital scopes I had (Hantek 20MHz and a cheap FNIRSI PRO weren't able to give accurate information when overdriven). My Tek 7D20 (from the 1980's fared better), but I stuck with an analog scope.
I also measured Vgs(th). Initially I did that with a current sink (built to do load tests on voltage regulators and test power transistors) set to 250uA (schematic here). Then I read that not shorting the gate to drain could cause the threshold voltage to measure lower than what it actually is (due to drain voltage interaction with a short channel). So, I designed a circuit that would measure threshold voltage with the gate and drain shorted. What I observed was the opposite of what I was led to expect. Instead of the threshold voltages measuring slightly higher, they were slightly lower. I had binned the devices by small ranges of threshold voltage variation, so I was only able to verify that the threshold voltage of some of the devices were low enough to be in a lower bin.
EDIT: Of the 130 power MOSFETs (IRFZ44, IRLZ44, IRF9540, and IRF9405; all packaged improperly) I've recently purchased on AliExpress, only one was bad, and it was my fault. When I breadboarded the initial test circuit, I had the gate of an IRFZ44 floating and it died from over dissipation before I noticed the wiring problem (and before I smelled something getting hot and turned the power supply off). The on resistance of the IRF9540 were higher than spec, so I suspect rejects or relabeled inferior parts (e.g. IRF9510 or IRF9520).
EDIT 2: corrected IRFZ44 (was IRF44)
Even though they were packaged improperly, I wore a grounding strap while handling them and repackaged them in antistatic bags after testing them. Since they all are suspect, none will be used in important circuits. I never learned about MOSFETs in school. Until recently, I had only used them as switches. I had one superficial chapter, so I'm learning about how to use them in analog circuits.
Last edited:
