# Does LED matrix scanning reduce overall brightness?

Using scanning (or multiplexing) of an LED array, such as below:

Such that only one row is powered ON at any given time, would this cause a reduction in LED brightness?

My rationale is: In this 4x4 matrix, row 1 will only be turned on 25% of the time. This seems awfully similar to PWM control, which is also a common method of controlling LED brightness. In this example, would all the LEDs be operating at 25% of their maximum brightness?

• Sure, yes. That's the limiting factor how big the matrix can get. – Janka Oct 18 '16 at 20:06
• @Janka Well that's a disappointing answer. Didn't see anything about brightness anywhere I read, so was hoping I was missing something... Oh well, thanks. – qoou Oct 18 '16 at 20:12
• LEDs have a max continuous and max pulsed rating, that might help. – Wesley Lee Oct 18 '16 at 20:15
• @qoou But you can compensate a bit by throwing more current through the LEDs each time they are turned on (I think that's what Wesley Lee wanted you to guess). – dim Oct 18 '16 at 20:16

Yes, if the LED current remains the same, then scanning will reduce the duty cycle and the amount of light.

You can offset this by increasing the LED current. But, you need to watch out for the LED current ratings (continuous and pulsed), of course.

edit:
As an aside, keep in mind that human perception of the brightness is a nonlinear function of light output.

(from here)

• I wouldn't put too much hope in increasing the current. Better use the brightest LEDs you can get instead of some from the cheapo low-luminosity bin. Brightness difference between different bins from the same manufacturer can be factor 20. – Janka Oct 18 '16 at 21:48
• That chart is probably accurate for objects in a dark environment, but a brightly-lit environment will often shift the y axis considerably. If an LED is shining through e.g. a piece of paper in a brightly-lit office, a small change in brightness can mean the difference between the LED being essentially invisible versus being clearly visible. – supercat Oct 18 '16 at 21:50

Yep, the maximum duty cycle goes with the inverse of the number of groups that get cycled through. This can be compensated for to some extent by increasing the drive current, but then you may have to take some precautions to ensure that the LEDs do not get damaged if they are accidentally run at 100% duty cycle at the increased current. I picked up an LED matrix assembly a while back that has integrated drive electronics (decoders and shift registers) and it actually includes a timer component that limits the length of the enable pulse so that it's difficult to leave one row illuminated indefinitely.

Yes, LED's brightness depend on the current supplied; the bigger the array, the more current you will need otherwise brightness with decrease.