2
\$\begingroup\$

I can always find the Rds(on) values for the MOSFETs I use, but the Rds(off) value is hardly ever listed. Is there a special reason for this?

In the datasheet listed, the Rds(on) equals 0.54 ohms but Rds(off) can't be found.

http://www.vishay.com/docs/91015/sihf510.pdf

\$\endgroup\$
  • 2
    \$\begingroup\$ Because there is "Zero Gate Voltage Drain Current" - Idss. \$\endgroup\$ – Todor Simeonov Jun 8 '17 at 16:45
  • 2
    \$\begingroup\$ That's actually a pretty good question, but you can figure it out from Idss max which despite what @TodorSimeonov says, is never zero. Minimum resistance is \$4G\Omega\$ at 25C for the listed device. It IS however highly temperature sensitive, dropping to \$0.4G\Omega\$ at 125C, which is probably why they don't mention it specifically \$\endgroup\$ – Trevor_G Jun 8 '17 at 16:52
  • \$\begingroup\$ Umm no, firstly you can't just divide voltage by current at a single point and assume that the result is a meaningful resistance. Secondly it seems you screwed your units up. \$\endgroup\$ – Peter Green Jun 23 '18 at 0:02
4
\$\begingroup\$

The concept of "resistance" assumes that voltage and current are at least roughly proportional.

Diodes and by extension transistors tend to have a roughly constant leakage current over their normal operating voltage region (though there can be significant variation with temperature), so it doesn't make sense to characterise that leakage in terms of a "resistance".

\$\endgroup\$
2
\$\begingroup\$

There a drain and source dependence, as well as a temperature dependence. For a nFET and starting from the EKV model, I came up with

$$R_{off}=\frac{L \kappa e^{\left({\kappa V_{T0}-V_{g}}\right)/{U_T}} \left(V_d - V_s\right)}{W U_T^2\mu C_{cox} \left(e^{{-V_{s}}/{U_T}}- e^{{- V_{d}}/{U_T}}\right) }$$

where \$\kappa\$ is the channel divider, and \$U_T\$ is the thermal voltage. Most people do not really care that much about actual performance when a device is ``off"; however, due to the drain dependence, it can be significant. (I've ignored \$\sigma\$ on the drain term)

\$\endgroup\$
  • \$\begingroup\$ Your equation could do with some more definitions, in particular what are the voltages relative to? substrate voltage? \$\endgroup\$ – Peter Green Jun 9 '17 at 13:15
  • \$\begingroup\$ I assumed that the substrate was 0v and the device was an nFET. This was me literally starting from EKV and moving through it in 5min. It's possible that it's not correct, but zeroth order it had the correct behavior. \$\endgroup\$ – b degnan Jun 9 '17 at 13:34
0
\$\begingroup\$

Rds(off) is so high that it is not relevant for the vast majority of MOSFET applications (mainly power switching applications). Rds(on) is normally used to determine the on-time power loss. The power loss is always assumed to be zero when the FET is fully turned off i.e. Rds(off) is infinite.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.