RGB led problem.

[Brechtix]

Hi,

I built an on board lighting system in a WAW, based on these RGB LEDs from DX, but I'm running into a few problems I can't solve. My knowledge of electronics is basic.

Battery used was a 8.4V Sigma iion battery. The rear light and brake light (with reed contact) were respectively regulated from a 350mA and a 500mA driver to the red cathode. The turn signals got 500mA through a digital relay to the red cathode as well. It worked well for a while: these large surface LEDs are discrete and aerodynamic but very visible when lit.

But alas: It all breaks down after a while.

1)Why? Did I do something wrong with the regulation? Are these LEDs very vulnerable to road vibration/shocks? Is it the humidity?

2)How can I improve the system?

3)Is it possible to use the blue component as a running light and have the red component blinking on top of that at the same time?

Thanks.

 

[AnAppleSnail]

DIY builds are usually vibration-sensitive until proven otherwise. Check your solder joints if you haven't, then connect (low) separate power to the LED leads to see what's broken. Then let us know what you find ;)

 

[jdp298]

You're not wrong about the vibration thing. I appear to have solved it now, but it took a redesign and an entire stick from a hot glue gun.

 

[AnAppleSnail]

Do you have multiple accounts? Anyway, you can't go wrong potting components once they are securely soldered.

 

[Brechtix]

A picture of the application hereunder. It crossed my mind that it could be the way I mounted them. The diameter of the silicone cover of the LED is 10mm, this peeps just through a 10mm hole in the +- 0.2mm Kevlar skin. I mounted the leds with a little tape and then whacked a thick dollop of silicone or Tec7 on it and the inside of the skin. A coworker used heat glue by mistake and the leds came off rather soon. Apart from a good laugh this also means the LEDs do get quite hot. Is it possible they cook because the silicone dollop insulates them too much? If so, is there a conductive equivalent, a thick glue that evacuates the heat?

 

[Brechtix]

...I appear to have solved it now,...

This came from somebody else. Great he solved it but I still don't know :-)

 

[alpg88]

so how many volts go to leads of each color??

 

[Brechtix]

7.2 V battery > (Velleman?)driver labeled "1W" > Red cathode. According to DealXtreme specs: Red LED light: 2.0V~2.5V, 50 lumens brightness.

I presume the source voltage doesn't matter as long as the driver regulates the current?

 

[AnAppleSnail]

AIs it possible they cook because the silicone dollop insulates them too much? If so, is there a conductive equivalent, a thick glue that evacuates the heat?

I presume the source voltage doesn't matter as long as the driver regulates the current?

Yes on all 3. (I'm curious about your WAW) I suggest that you bang on a thin metal piece and use thermal adhesive so the LED heats up the outside-facing metal bits. Even a small fin should do some good here, at speed the air will slurp up lots of heat from exposed metal.

 

[Brechtix]

All right. I bought thermal paste today, I'll glue the leds to a chunk of aluminium with thermal paste, and then tec7 glue the front of the large surface to the back of the kevlar shell. So the back side is free too cool.

The WAW in the picture is from a client in Germany. We (www.fietser.be) are building them.

Thanks!

 

[AnAppleSnail]

All right. I bought thermal paste today

Thermal Paste won't be glue, it will be goo. Thermal adhesive (2-part epoxy meant to be thermally conductive) may serve you better. I have a one-part 'silicone thermal adhesive' to try out.
Arctic's Thermal Adhesive