next topic »

[Reply #15 on:]

Very informative page, thanks Mike. I did some research before but didn't find that particular page.

I had seen some references to putting high currents into LEDs (eg. IR transmitters) for a brief time, but on careful reading it appears that such LEDs would be designed to handle that, and in any case, the higher current does shorten their life, it's just allowed for in the design.

So would you say then that 8 resistors, in series with A-G/DP would be the best? Or 4 in series with each of the digits? I'm guessing that the 8 are better because that would give more even current.

So that makes it OK then? Or does it make it yet another circuit designed to destroy components?

Forgive me, but I assumed that a commercially sold board would have investigated the ramifications of this and found out that the design was OK.

[Reply #16 on:]

4 in series with each of the digits?

The problem with that is as more segments come on more current is drawn therefore more voltage is dropped across the resistor. That means that the digit's brightness depends on the number of segments that are on. So you get a bright 1 and a dim 8. The resistors need to be in each segment so 8 it is. That is the attraction of having constant current drives.

I had seen some references to putting high currents into LEDs

Yes a lot of LEDs have a pulse current specification that is much higher than the continuous current rating. The on / off ratio is normally quoted and it is used to compensate, to some extent, for the dimming caused by the on / off ratio. However, even so the current has to be managed to be within that specification, and it does shorten the life. A firm I worked for had a 10% return rate on set top boxes with high current pulsed LEDs even though the current was within the specification. Contracts often mention a epidemic failure clause that kick in at about 3% so there was a lot of fuss about that. When making a million boxes 10% is a lot. There were 19 LEDs on the box and it only takes one to fail for the box to fail, that is be returned.

However, the major problem with high pulsed currents is from the driver rather than the LED. The potential for doing damage is a lot greater because chips are of a physically more delicate construction than LEDs.

but I assumed that a commercially sold board would have investigated the ramifications of this and found out that the design was OK.

Ah if only that were true. Thing is, even commercial designs are made by people, with all the variability that implies. You should have seen some of the things I had to correct from my 'professional' engineers when I was a manager.

[Reply #17 on:]

The Sparkfun serial LED displays have taken this route I think, you can't see all the PCB so I can't verify how many resistors (if any) have been used but the brightness varies a lot depending on the number of segments lit.

And that is supposedly designed by professionals.

______
Rob

[Reply #18 on:]

The problem with that is as more segments come on more current is drawn therefore more voltage is dropped across the resistor. That means that the digit's brightness depends on the number of segments that are on. So you get a bright 1 and a dim 8. The resistors need to be in each segment so 8 it is.

I guessed that would be the case, <sigh>.

In my alarm clock circuit here:

http://www.gammon.com.au/forum/?id=11165

I used 4 resistors. And guess what, the digits were varying brightness depending on how many segments lit up. Ah well, we live and learn.

Thing is, even commercial designs are made by people, with all the variability that implies

I think I am beginning to understand why my electric blankets have been failing every year. They have pretty sophisticated-looking control boxes, but don't seem to last very long. Either they are badly designed, built to a (low) price, or both.

[Reply #19 on:]

Either they are badly designed, built to a (low) price, or both.

In the battle of conspiracy over cockup, I vote for cockup every time.

Most things are built to a price coupled with a reliability target. It is the target reliability that drives most consumer goods in large companies, mainly because return figures are easy to collect as a metric. No prizes are given for over achieving on reliability, in fact in most cases you are criticised for not making it cheap enough. In fact in my last job that was written into our contracts.

[Reply #20 on:]

Code:

It is the peak current that is important not the average current when considering component ratings.

heh, its not really how much force is being applied to your crotch over a matter of time, its the initial kick that drops you

[Reply #21 on:]

... in fact in most cases you are criticised for not making it cheap enough. In fact in my last job that was written into our contracts.

So you would have left the resistors out? Makes it cheaper ...

Then a year later the company can sell another one.

[Reply #22 on:]

Have you any idea what you save by missing out a resistor? In the quantities we used to by them at they were £0.0001 each!