1
\$\begingroup\$

This may be a question strongly depending on application and opinion, but I'll give it a try.

I am designing some controls for an industrial application, including cap touch buttons and LED 7-segment displays. As I have many of those LED-segments (about 280) I won't be able to control them directly all the time, instead I will multiplex them.

The brightness of the led-segments will be indirectly proportional to the multiplex factor, I am wondering how bright a LED display "normally" should be, and if I can get away with my heavy multiplex design.

I am planning to use this, which has a typical brightness of 9.0 mcd. How deep can I multiplex this? could I get away with a 1:7 multiplexing (the brightness per segment will be 1/7th)?

If the above is not answerable: Is there anyone regularly working with displays? how is the brightness with those in general? The device I am building is only used inside. Sun-radiation directly onto the display is unlikely.

\$\endgroup\$
3
  • 2
    \$\begingroup\$ try and test it \$\endgroup\$
    – Pedro Quadros
    Jul 22, 2015 at 14:10
  • \$\begingroup\$ Those currents can be easily handled by 74HC595 and similar chips. Maybe you don't even need to multiplex the LEDs, as long as power/cost/surface budget allow you to do so. \$\endgroup\$
    – FarO
    Jul 22, 2015 at 15:21
  • \$\begingroup\$ @OlafM I looked at the 74HC595 and on GPIO Port extenders like PCA9505/06. Certainly I will use one or two, but having a bazillion of them in my design will be much too expensive. \$\endgroup\$
    – jwsc
    Jul 23, 2015 at 6:23

2 Answers 2

Reset to default
2
\$\begingroup\$

It has a lot to do with the environment it will be operating in and the optical filtering (eg. a colored lens) in front that prevents much of the ambient light from splashing around inside the LED display and reducing the contrast.

As Pedro said, your best bet is probably to test it.

Above the human eye's flicker fusion frequency (and you'll want to operate well above that or the display will be very annoying) the brightness will be approximately the same for the same average current. So if it's okay at 5mA DC it will be okay with 35mA at 1/7 duty cycle. It's usually not good to go higher than 1/8 for a display if you want maximum brightness or 1/16 if you can tolerate substantially lower performance for saving some parts. 1/7 is a good duty cycle.

That's a pretty efficient display series as well as a relatively small segment area for each LED so I suspect 5mA average / 40mA peak (1:8) or 35mA peak (1:7) will be more than acceptable for a typical industrial environment given a reasonable colored filter in front of the display.

\$\endgroup\$
3
  • \$\begingroup\$ Thank you very much. Testing is not an option right now, unfortunately. The figures you gave was all what I'm looking for. \$\endgroup\$
    – jwsc
    Jul 22, 2015 at 15:31
  • \$\begingroup\$ @jwsc datasheet here I'd be nervous about a 9 mCd LED and I'm surprised it's that low. Cd is brightness level - that seems immensely low. The saving grace is that Avago don't make junk (almost ever) so if it is viewable anywhere it should be OK in your application. | FWIW - the datasheet is wrong as it says the output is 9 megacandela typical. Those are only used in Star Destroyer main batteries. :-) \$\endgroup\$
    – Russell McMahon ♦
    Jul 22, 2015 at 16:26
  • \$\begingroup\$ @RussellMcMahon 9mcd is actually pretty bright for an LED display. 500ucd is usable. It's not supposed to be a flashlight. ;-) \$\endgroup\$
    – Spehro Pefhany
    Jul 22, 2015 at 17:48
1
\$\begingroup\$

The datasheet for your display tells us that it produces 9mcd for 10mA.
It also tells us that it can handle 100mA for a 1/10 duty-cycle and 0.1ms pulse width and although the graph which shows relative luminous intensity vs forward current only goes up to 50mA, a guestimated extrapolation seems to imply that you could get as much as 4x as much output - so possibly in the region of 36mcd during a 100mA pulse.
Apparently perceived brightness is not simply the average of on brightness over time (I don't have a reference at hand unfortunately) so a 1/7 duty-cycle of 36mcd should appear brighter than 5.14mcd. How much - I'm not sure...

\$\endgroup\$
1
  • \$\begingroup\$ Oh... That are some nice thoughts. I wouldn't want to test out the max ratings of that display, but it is definetly a fallback. \$\endgroup\$
    – jwsc
    Jul 22, 2015 at 15:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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