max7219 & high voltage 7 Segment LED

I've been trying to figure this out for the past week, so I figured it was time to seek out some help.

I'm trying to use a max7219 & arduino to drive a large 2.3" 7 segment LED. I'm getting the correct digits displayed, but they are very, very dim, probably 1/10th the brightness the display should be. The important specs for the red 7 segment LED are: forward current per segment: 30ma (within the 40ma maximum of the 7219) VF typical 6.8V VF maximum 8V The 7219 can only directly drive displays that have a maximum 5.5V / 40ma per segment, but Maxim does describe a number of methods to use the 7219 to drive higher voltage displays. This is the info they supply: I've wired the 7 segment to the 7219 according to figure #3: I'm using a 2N2222 as the Q3 transistor and a 60v p-channel mosfet as Q2. The driver current is 7.5V.

Anyone have any idea why the LEDs are nowhere near full brightness? I know the problem isn't with the 7 segment, because I wired it directly to the power supply to see what the brightness would be like (REALLY BRIGHT). I'm also certain that there is no trouble with the arduino code or in the order the 7219 pins are connected to the 7 segment, as I tested all of that with a lower current 7 segment that I ran off the arduino's power.

any help would be appreciated.

whats Iset?

maybe the chip is current limiting on the cathodes?

I have a 10k on iset. On the max7219 datasheet, the lowest resistor they recommend is 9.53k, which I was afraid to go below. Maybe I'll try, risk the burnout. I'm pretty sure it's not the chip, because the driver power isn't passing through the 7219. I starting to think I need a transistor more robust than a 2N2222. That could be limiting the current. Maybe?

Well the chip hooks up to both positive and negative on the LEDs in a grid, seems like you've upgraded the hookup on one side of that, but not the other...

perhaps, but I did also try boosting the segment connections from the 7219 as in figure 4 of the maxim document: with no change in the brightness. still trying.

Look at the spec of Q2, it looks like this is not turning on fully with the voltage on the gate not going low enough to turn it on. It is a p-channel FET you are using?

Q2 is a p-channel MOSFET. I've tried a 60v min -0.4A & a 55v min -1.8A. What makes you think that it's not turning all the way on? If I were to try a different MOSFET, it means waiting for another Jameco order.

In trying to figure this out, I made a mistake that may prove Grumpy Mike's point: while switching between and testing different MOSFETs, I accidentally put a npn 2n2222 transistor where the p channel MOSFET was supposed to be. With the wiring of both figures 3&4 (though no R1 resistor), The display is roughly at half power, and switching fine. If I wire the segments directly(no transistor as in fig 4), then the LEDs power up full, but never shut off. Now I need to be somehere between the 2 results.

Any more ideas?

've tried a 60v min -0.4A & a 55v min -1.8A. What makes you think that it's not turning all the way on?

Because what you describe is what I would expect to see if it wasn't turning fully on.

You don't say what the mosfet is so I can't look it up but the parameters you quoted are not the important ones here. The voltage required to turn it on is usually found as a condition in the Ron parameter.

The voltage you will get on the gate is determined by the supply Vdriver and how low the transistor's emitter goes. This might be improved by increasing R3 or reducing R2. Make some measurements around these points. The original circuit should work given the right components.

Following this thread with interest. Let us know if you nail it down.

I´m having the same problem, but I´m using 2 ULN2804. One for the segments and the other for the digits. The result was the same of recur´s. I´m need the solution for my graduation project.

Lougard:- Best start a separate thread with a detailed explanation and schematic of your particular problem.

What’s the RDSon of your FETs? I’ve been doing some investigating into using them for high-side drivers, myself, the last few days. And the ones I saw with low-ish max currents like .4A, in TO-92 cases (that would allow you to confuse them with a 2N2222) had resistances on the order of tens of Ohms. It doesn’t take a whole lot of mA to cause a couple of Volts of drop across that big a resistance.

A question for Grumpy Mike, since you seem to have a good knowledge of this: I’ve been thinking of using the FETs to switch voltages on the order of 5V or so. The datasheets say they have threshold voltages in the range of 2-4V, but they always spec the on-resistance and current-handling capabilities at a VGS of 10V or so. Does that mean they need to be derated at 5V? That it’s a bad idea to even use them at that low a voltage?



Does that mean they need to be derated at 5V?

Basically yes. There should be a curve in the data sheet showing gate voltage against Ron or against maximum current. Operating a FET at just above the threshold drives it in the liner region where there is lots of volts drop. Sometimes that is exactly what you want to do.

However, what you should look out for is mention of "logic FET" in the specification and ensure it is fully on at the voltage you want to drive it from. In that respect HEXFETs are often better. Remember not everyone wants to use FETs with logic, in fact it is only a small part of the applications of a FET.

Thanks, Mike, that was a big help. I've used millions of FETs over the years, but they've all been inside digital ICs, so I don't really understand the meanings of the datasheet curves for discrete ones yet. You've given me a good clue to what I should be looking for.

Thanks again,


Any new on this topic? What FET did you use in the end?

Thanks, ralf