irlb8721pbf.pdf- datasheet
- (267.13 KiB) Downloaded 3 times
Power MOSFET IRLB8721PBF
Moderators: adafruit_support_bill, adafruit
Please be positive and constructive with your questions and comments.
-
Dpm206
- Posts: 1
- Joined: Wed Dec 13, 2017 4:26 pm
Power MOSFET IRLB8721PBF
Adafruit promotes this MOSFET as suitable for operation - with only 3.3 volts on the Gate. This is a misinterpretation of the Vgs(th), Gate Threshold Voltage, specification in the data sheet. That threshold voltage is when the MOSFET just begins to turn on - not fully on in the desired minimum resistance (RDS), switch mode of operation. In fact, that Vgs(th) parameter is more applicable to the linear/active region mode of operation - as an audio amplifier, for example. Note that the conditions for the parameters in the data sheet - are with 10volts - or - 4.5 volts - on the Gate. When a MOSFET of this vintage (2009) is considered logic-level compatible - as suggested by the "L" in the part number, it assumes 5 volt logic levels. Hence the parameters on the data sheet with 4.5 volts on the gate. I don't consider this part to be suitable for operation with 3.3 volts on the gate. "Conventional" BJT's, Bipolar Junction Transistors, would be a better choice. Figure 12 in the data sheet has the gate voltage curve starting at 4 volts.
-
adafruit_support_mike
- Posts: 67485
- Joined: Thu Feb 11, 2010 2:51 pm
Re: Power MOSFET IRLB8721PBF
We've tested the IRLB8721PBF directly. Its Rds.on is well below 1 Ohm at Vgs=3.0V.
Datasheets are legal documents, and as such represent a negotiated compromise among a vendor's engineering, marketing, and legal departments. The marketing department wants to make every device as attractive as possible at first glance, the engineers want to provide useful technical information so they don't have to deal with tech support phone calls all day, and the legal department has a hard requirement that every fact in the datasheet has to be provable in court.
As such, the table of min/max/typical values in the Electrical Specifications section tends to be conservative. Most companies use 6-sigma standards, which sets the tolerance at six times the standard deviation of the process average. That puts the chance of finding a device that doesn't meet spec at about 1 in 500 million, which makes replacing anything near the edges trivially inexpensive.
That leads to different practices in design and provisioning. Worst-case design adds all maximum tolerances together, which forces a tradeoff between poorly-specified performance and high cost for tight-spec devices. Standard design treats tolerances as statistical values and takes the RMS sum, which provides a much better expectation of good performance from reasonably priced components. Short-run and one-off design can afford to characterize devices individually, which allows the best possible performance from any device, regardless of its price.
Of course, short-run and one-off projects can also afford to spend more on the best possible devices because the economics of scale doesn't apply.
Datasheets are legal documents, and as such represent a negotiated compromise among a vendor's engineering, marketing, and legal departments. The marketing department wants to make every device as attractive as possible at first glance, the engineers want to provide useful technical information so they don't have to deal with tech support phone calls all day, and the legal department has a hard requirement that every fact in the datasheet has to be provable in court.
As such, the table of min/max/typical values in the Electrical Specifications section tends to be conservative. Most companies use 6-sigma standards, which sets the tolerance at six times the standard deviation of the process average. That puts the chance of finding a device that doesn't meet spec at about 1 in 500 million, which makes replacing anything near the edges trivially inexpensive.
That leads to different practices in design and provisioning. Worst-case design adds all maximum tolerances together, which forces a tradeoff between poorly-specified performance and high cost for tight-spec devices. Standard design treats tolerances as statistical values and takes the RMS sum, which provides a much better expectation of good performance from reasonably priced components. Short-run and one-off design can afford to characterize devices individually, which allows the best possible performance from any device, regardless of its price.
Of course, short-run and one-off projects can also afford to spend more on the best possible devices because the economics of scale doesn't apply.
Please be positive and constructive with your questions and comments.