Low-side MOSFET switching issue with 3W LEDs

Using Arduino LEDs and Multiplexing
1

I have an 8x8 matrix of Cree 3W LEDs. Only one LED is on at a time, and only for a few seconds. I switch on the high-side of every column with an IRF9540 P-Channel MOSFET , and the low side with a CAT4101 constant current LED driver that lets me run a controllable current up to 1 A through the LED. An Arduino is controlling the matrix. Here's a basic schematic for one column/row (open the image in a new tab for a better view):

1307×554 31 KB

For a number of reasons, I'd like to use an N-Channel MOSFET as the low-side switch, connecting the source of every N-Channel MOSFET to the CAT4101; the CAT4101 would still let me control the LED current, using the variable resistor:

1319×567 34.1 KB

With the first schematic, everything works fine, and I can set the current up to the maximum 1A level. With the second schematic, the current maxes out at 0.7 A when the N-channel MOSFET gate is controlled by the Arduino (+5V). If I use a +9V voltage to the gate, the current maxes out at 0.8A. I've checked the datasheet for the IRL540 N-Channel MOSFET, and from the curves I think I should be able to push a full 1A through without issues; tried an IRLZ44, and got the same results. So clearly I'm missing something. Any ideas?

RTI-2.png
RTI-2.png1307×554 31 KB

RTI-3.png
RTI-3.png1319×567 34.1 KB

2

So clearly I'm missing something. Any ideas?

Yes in the bottom circuit the IRL540 is wired in the source follower configuration, so the voltage on the source is never going to be higher than the voltage on the gate, and is normally a tad less.

Your diagram is effectively upside down and difficult to follow. Try and have the positive voltages at the top and ground at the bottom.

3

Thanks, Mike. Don't understand your explanation, but that's ignorance on my part - will do some research. Sorry about the schematic - it's the first one I've ever posted here.

I also have to apologize. I just re-breadboarded the circuit, with the idea that perhaps I'd be able to get it to work with a TIP120 (not a preferred solution because of the increased power consumption). Thought I'd try my original configuration with a higher voltage, just to see what would happen. And this time, it worked perfectly, not just at the higher voltage, but also at the original voltage of +9V; I can adjust the current all the way up to 1 A on the LED. And the MOSFET turns on/off nicely with just a +5V voltage on the gate. I must have screwed up the original breadboarding somehow. Sorry.