Can you identify any inherent problems with the following LED driver circuit?


simulate this circuit – Schematic created using CircuitLab

Ignore the part numbers as they are just for symbolic purposes. Assuming that the n-channel MOSFET in question can handle the 60mA called for by the LEDs, is this a sound LED drive circuit?

It's not typical in my experience to put an n-Channel MOSFET on the high side, but I'm hoping that's just my inexperience talking. I'm willing to lose some efficiency in exchange for not having to use inverse logic on the control point. What type of n-channel MOSFET is appropriate in this context? One that I'm considering using is NXP 2N7002, will it do the job?

Update #1

To make this clearer, I am interested in using a circuit like this as part of the topology for an LED matrix. My thought was something like this (where the current sinks are implemented by, e.g. TLC5916) for a 3x3 configuration:


simulate this circuit


1 Answer 1


Your instincts ("It's not typical in my experience...") are correct. That circuit won't work. The IRF530 needs 4 volts Vgs worst case, and that is simply not going to be possible when you include LED forward drops.

However, it's not clear what your justification for doing this is. You mention that you don't want to "use inverse logic on the control point". If this means that you want to use a 5 volt signal to turn on the LEDs, put the MOSFET on the negative side of the current sources, with the source at ground. If you want a 0 volt signal to turn on the LEDs, replace the MOSFET with a p-type MOSFET, with source at +5. Either way will work.

  • \$\begingroup\$ It's for an LED matrix high side drive, using constant current sinks on the low side \$\endgroup\$
    – vicatcu
    Commented Jan 20, 2015 at 0:07
  • \$\begingroup\$ So put the MOSFET in the usual place. It's true that the current sinks won't work properly when the MOSFET is blocking the current, but that won't matter, since there's no current flowing anyways. When the MOSFET turns on, the current sinks will do their thing. Unless there are control lines for the current sinks that you're not showing, and even then they ought to work OK. \$\endgroup\$ Commented Jan 20, 2015 at 0:10
  • \$\begingroup\$ That doesn't really make sense to me, in a matrix configuration all the cathodes of a column are wired to a current since, and all the anodes of a row are wired to a voltage source. So for a 3x3 matrix there would be three duplicated circuits I depicted in the question, and each of and those current sinks could be turned on / off by the microcontroller individually. \$\endgroup\$
    – vicatcu
    Commented Jan 20, 2015 at 1:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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