Hi there.  There are a lot of cheap 8x8 LED matrices out there (a grid
of 64 LEDs).  Most of them have only 16 pins -- 8 for the rows and 8
for the columns.  This lets you turn on any combination of LEDs in a
single column (apply positive voltage to one column and negative
voltage to the rows that you want to light up in that column). 
Typically you'll hook this up to a chip controller that scans across
the rows so that each row is lit for 1/8th of the total time.

My question: does this mean that the LEDs are 1/8th as bright as they
could be if you had 128 pins (two to each LED)?  Or is there some
capacitor in there that lets you load up enough voltage to keep the
LED lit for the other 7/8ths of the cycle?

Thanks.

Hi megacz,

First, I will answer your last question, because that is the one with
a straightforward answer. You ask if there is a capacitor that keeps
the LEDs lit when they are not being driven. The answer to that is
"no". To provide enough capacitance for this purpose (for 64 LEDs)
would increase the package size considerably, and greatly increase the
package price due to the extra fabrication complexity.

Your main question is whether an LED is only 1/8th as bright if it is
only being driven for 1/8th of the time. To answer this we need to
consider the behavior of the human eye as well as the behavior of the
LED itself.

The human eye responds to peak levels of light, and also to average
levels of light. The brightness that we perceive is somewhere between
the peak and average brightness. We can pulse an LED so that the
average brightness is the same but the peak brightness is higher, and
this will increase the perceived brightness without increasing the
power consumption of the LED.

This is described in an application note from ON Semiconductor
(Acrobat reader required):

"...Oscillator Increases the Brightness of LEDs..."
https://www.onsemi.com/pub/Collateral/AND8067-D.PDF
(see the section "Why are pulsed LEDs brighter than DC LEDs?")

But we also need to look at the light output of the LED at different
current levels. This is different for various kinds of LEDs, and you
would need to consult the data sheet for the specific display that you
are using. The above application note gives an example of a specific
green LED which is brighter when driven by a pulsed signal with
average current of 10mA then when driven by a steady 10mA DC, but then
cautions that if the same LED is being driven at lower current levels
then it would be brighter using DC than pulses.

So, although you would need to check the specific data sheet for your
LED matrix, you can be assured that pulse-driving your LEDs is not a
bad thing to do, and may even improve the perceived brightness of your
display.


Additional links:

"Seven Segment Displays for High Light Ambient Conditions"
http://www.zipernowsky.hu/~kami/data/optical/hdsp3530.pdf
(recommends strobing at 1kHz or faster for maximum brightness)

The LED Frequently Asked Questions Pages
http://www.pioneernet.net/optoeng/LED_FAQ.html
(lots of good information, in particular questions 22, 1, 2 and 3)


Google search strategy:

led multiplexed brightness maximum average
://www.google.com/search?q=led+multiplexed+brightness+maximum+average

"64 LEDs" brightness current frequency
://www.google.com/search?q=%2264+LEDs%22+brightness+current+frequency


Regards,
eiffel-ga

Thanks!!!  Quite helpful.  I guess there is no perfect answer since
this depends on the LEDs I use, but knowing that I'm probably not
losing anything close to 7/8ths of the percieved brightness (and might
even be gaining some) is good enough.  Thanks!