With all due respect, if the gate threshold does not meet the specification, the device does not meet the specified range of operating parameters. I want to power an LED strip (160mA). The Fairchild datasheet indicates this device should power up to 20A with 3V at the gate (yes, it should supply higher currents at higher gate voltages; and yes, at some point a heat sink is required). The datasheet indicates a gate threshold of 1V min to 2V max. And the datasheet also indicates that these can be controlled directly from an IC.
The vendor's images clearly show these marked as a Fairchild device. And the vendor even has a video showing one operating a small motor (perhaps 1A?) while supplying the gate from the 3.3v output of an Uno through a pot (Fairchild mark clearly visible on the device in the video).
The vendor may not even be aware that the devices do not meet spec. One popular vendor notes that they supply one of two "nearly identical" variants. The other variant listed is also obsolete. My guess is that they are sourcing them from ostensible NOS supplies somewhere, making their supply vulnerable to counterfeits or rejects.
The IRLB8721PBF that I have now ordered are the same ones referenced by Wawa as available from Adafruit. They are a current production device and the big houses show stock levels in the thousands. Hopefully that increases the odds they are genuine.
FWIW: I have very limited experience with the Arduino line, all with Unos. That does not mean I am a complete newbie when it comes to tinkering with electronic devices. I started with repairing tube radios and TVs in the 60's, progressed to nuvistors and transistors. Built light organs and audio amps before building my first computer from a bag of chips in the late 70's. Etc. etc. Lately I've been building robots.
SMDs give me trouble, even with readers and magnifiers. My eyesight ain't what it used to be, But I still know how to read a datasheet.
Sadly, the data sheet value of Vds(th) does not actually mean that.
It simply means that this is the lowest voltage at which the mosfet's turn-on process begins.
If you are interested in a nice detailed review, please look at
250uA is listed as the Drain current at the threshold voltage at the gate -- but that threshold is between 1 and 2 volts (precise value of individual samples depending on manufacturing tolerances, etc.) not 3.3V.
No, it's not like a mechanical switch. It's a curve. I get that, never said any different. But the drain current should climb exponentially once the threshold is crossed, reaching into 20A territory by the time gate voltage hits 3.3V (depending on temperature, etc. etc.).
For those that are interested, the IRLB8721PBF that I ordered (from a large supply house with >44,000 in stock) arrived today. They meet their spec. I spent a little more time experimenting with memory depth and horizontal bandwidth settings on the digital scope, and managed to find a closer approximation to the clean lines displayed on the analog Tek. The left cursor marks the 1V point (although the threshold is actually slightly higher on this one -- 1.35v to 2.35v) and the right cursor is marking 3.3v to show the range I have to play in. The uppermost curve is about 250mA (I need 180mA).
Look again at the graph in post #10. The x-axis (horizontal) is for Vds. I thought you were using 12V as a drain voltage supply. Why are you looking at Vds limit of 3.3V?
The upward steps in the parametric curves show the device conduction for increasing values of Vgs.
I wasn't trying to replicate the format of #10 (which steps discrete values of Vgs). I was interested in characterizing a varying Gate voltage against very low values of drain current* so the X axis on my curves are the Gate voltage, not Drain. Generated in the form of a 10v p-p triangle wave. (The completed circuit will be powered by a nominal 12v supply at the Source, but that won't affect gate voltage)
The Y axis represents Drain current. (Generated by applying a series of ascending pulses up to 10V across a precision resistor per cycle of X). Sorry for any confusion, I guess I probably should have clarified that from the beginning, or marked up the screen captures with Photoshop or something. I didn't anticipate this thread would dive so deep into technicalities.
The original point of this thread is best illustrated by Wawa's question: I bought them through an Amazon store because they are discontinued; Mouser, Digi-key, etc. don't have stock left and can't get them anymore. I learned long ago to be very wary of eBay "NOS" components, I guess I need to add Amazon to that list too. Caveat Emptor.
*Before someone asks the obvious question "Why choose a device that can handle 10s of Amps to drive something that requires mA?" Heat dissipation. These will live inside motorcycle luggage with very limited airflow and subject to operation during warm sunny days. I'd like for them to have a fighting chance.
