Re: LEDs, current and PWM
Hi, Robert zeu Great Moderator
What's your question ??? : I have 267 Led segments on my circuit and I want to test them. But @ 20 mA each, and all together ... that draws 5.34 amps ...
OK ...
Now ... the questions
1) is it compulsory to test them all together at the same time ( and is it possible ? - some multiplexing used or not ??? )
2) you absolutely need 20 mA per segment ???
PWM won't do anything for you ... cause you have to give some energy to light on your leds - whatever the way you provide it - just to be able to see them glow.
a switching supply only could allow you to start from a higher voltage DC "mains"
From that you could understand to reduce the current draw, you could quickly test groups of leds after groups ... ( say 7 seg, then 14 seg, then leds ... i.e )
second way would be to use lower current for the test ... but that needs some scheme watching to tell if possible !
No miracle when energy speaking ... never ! :p
Alain
Re: LEDs, current and PWM
You may want to use this one.
MBI5026
16-bit Constant Current LED Sink Driver.
Output led current can be adjusted with a single resistor.
and also you can use OE pin for PWM input to dim the leds.
Re: LEDs, current and PWM
Alain, I can't test a few LEDs at a time. The idea is to light up the hole board to make it easy to spot non-working LEDs. I have 4 preliminary panels drawn so far, at least 1 more to go and they look like this:
http://i991.photobucket.com/albums/a.../Overhead1.jpg
I guess I could turn them on and off in sequence real fast, that should be just as good.
Robert
Light blue = rotary switches
Purple = rotary encoders
White = pushbutton
Red = pushbutton
Re: LEDs, current and PWM
Hi, Robert
just consider you need a certain amount of energy to make your lights glow ...
WHATEVER the way you give them that energy.
NO miracle, as I told you ...
so, the only way to get them all together lit is to reduce current to the minimum " readable " ... ( here PWM can be useful ... :wink: )
or use very high efficiency leds ... as there's a lot here ... current can be reduced to ~ 5 mA per Led to get a nice visible glow.
Alain
Re: LEDs, current and PWM
You can increase the current to the maximum the LED's can accept in pulse operation and then light them in turn with a low duty cycle.
Ioannis
Re: LEDs, current and PWM
May be I should have been more specific.
Check MBI5026.
This is a shift register. Use as many as you need.
And the current can be adjusted.
1 Attachment(s)
Re: LEDs, current and PWM
AN234, figure 11
http://ww1.microchip.com/downloads/e...tes/00234a.pdf
Attachment 6373
This is the technique I meant by blinking the LEDs in sequence (multiplexing right?).
- Let's say a complete 7-segment LED consumes 20mA.
- If I turn Q1 on/off, then Q2 on/off, and so on, really fast.
Wouldn't it only consume 20mA since only one 7-segment is lit at one time?
The human eye would see them as all on, no? Kind of like screen refresh rates.
Robert
Re: LEDs, current and PWM
If you light a LED for 10% of the time, fast enough, then your eye will of course see it ON but not bright. Maybe not 10% of the LED normal brightness but sure dim. So you have to increase the drive current.
That is what Alain meant by "No miracle when energy speaking ... never ! :p"
Ioannis
Re: LEDs, current and PWM
Ok, so multiplexing LEDs really is the same as PWM but in a group.
Time-sharing lightbulbs. :D
I guess I won't have a choice but to make a compromise on what Alain suggested; PICs go through a pre-programmed sequence lighting a group of LEDs across all panels.
Robert
Re: LEDs, current and PWM
Quote:
Originally Posted by
Acetronics
...
or use very high efficiency leds ... as there's a lot here ... current can be reduced to ~ 5 mA per Led to get a nice visible glow.
Alain
I put 16 high efficiency LEDs on a MCP23016 just to see what it looks like. They were just a bit too bright full on, so they come out just about right when pulsed in sequence, and current is kept at a minimum.
I could even use a bit lower than the regular 330ohms too.
Robert
:)