February 18, 2020 by Robert Keim
This article examines the PSMNR51-25YLH and discusses MOSFET drain-source leakage current.
An idealized MOSFET is a marvelous device.
We apply an appropriate voltage to a perfectly insulated gate and current flows freely from drain to source. Once we start to learn about all the little details of MOSFET operation, we might almost feel a bit of nostalgia for this vastly oversimplified model.
Drain-to-source resistance, gate resistance, parasitic capacitances, the body effect, avalanche breakdown…. There seems to be no end to the imperfections that complicate the design and implementation of MOSFETs, and in this article, we’ll explore a nonideality known as drain-source leakage current and abbreviated IDSS.
MOSFET Drain-to-Source Leakage
We expect a MOSFET to prevent current flow from drain to source when we apply a gate-source voltage that is far below the threshold voltage, and this is indeed what the device does—well, more or less. Some current will flow even when the MOSFET is in the off state, and manufacturers convey information about this “leaking” current by reporting values for IDSS.
In an application note on power MOSFET parameters, Taiwan Semiconductor defines IDSS as the current that flows from drain to source when the gate is shorted to the source and the device’s maximum drain-source voltage is applied.
This definition seems to be fairly standard, though the datasheet for Nexperia’s new PSMNR51-25YLH reports IDSS at a drain-source voltage of 20 V, rather than at the maximum, which is 25 V.
Diagram of IDSS. Image used courtesy of Taiwan Semiconductor
Drain-source leakage increases significantly as temperature increases. Nexperia indicates that the PSMNR51-25YLH exhibits good IDSS performance even at high temperatures.
Should We Be Worried?
Drain-source leakage is undesirable simply because it leads to wasted power. The leakage current multiplied by the drain-source voltage represents power dissipation that is occurring whenever the transistor is inactive.
However, drain-source leakage currents tend to be very small, and this makes me wonder why Nexperia emphasizes low IDSS in its description of the PSMNR51-25YLH. Reduced channel length contributes to drain-source leakage, and consequently, IDSS is a nontrivial concern with newer integrated circuits, since the numerous short-channel transistors in a VLSI chip will create significant amounts of static power dissipation.
The PSMNR51-25YLH, though, is a discrete FET intended for power applications. Thus, I can only assume that the importance of IDSS in this context reflects the desire to maximize energy efficiency even in systems that are not constrained by the need to extend battery life or reduce heat generation.
Logic-Level Power Control and Load Switching
The PSMNR51-25YLH supports fairly high drain-source voltages and extremely high drain currents (380 A), and yet its gate-source threshold voltage is only 1.55 V (typical) and its on-state resistance starts to flatten out at around 5 V.
RDS(on) as a function of gate-source voltage. Image used courtesy of Nexperia
Thus, you can control the FET using a typical 3.3 V logic signal from a microcontroller, and if you want to reduce RDS(on) almost to the minimum, all you need to do is translate that 3.3 V signal to 5 V.
And while we’re on the topic of on-state resistance, I really was surprised when I saw the RDS(on) spec for this FET: 0.82 mΩ maximum at VGS = 4.5 V. Less than one milliohm is some seriously low resistance!
One thing that I like about the PSMNR51-25YLH is related not to the part itself but to the datasheet. Thermal issues are a critical and easily overlooked aspect of power design, and most electrical engineers don’t have the expertise or the resources needed to perform advanced thermal analysis. Thus, the assistance offered by the manufacturer is greatly appreciated.
In the case of the PSMNR51-25YLH, the datasheet includes two examples of PCB layouts and provides specs for junction-to-ambient thermal resistance that corresponds to these examples.
Diagrams and specifications relating to PSMNR51-25YLH’s thermal resistance. Image used courtesy of Nexperia
This is valuable information that helps the designer to estimate the amount of copper area required to maintain acceptable temperatures in a given application. (For more information on this topic, refer to my article on thermal design for linear regulators.)
Feature image (modified) used courtesy of Nexperia
I’m still curious about why IDSS might be considered a prominent spec in the context of a transistor such as the PSMNR51-25YLH. If you have any thoughts, feel free to share them in the comments.