My partner saw a neat tutorial on how to make a simple cellphone powered LED light, basically it uses a DC-DC converter to step the μUSB port voltage from 5V to ~3.7V. The problem I saw with the simple design was that there was absolutely no current limiting or over-voltage protection on the phone μUSB port, this can be a bit of a worry if your phone was designed without them in mind. So another idea was born, make a similar device that offered these safety features.
Having a look through DigiKey I found this neat little IC, the PAM2863ECR, which fit the description perfectly and even had the ability to control the current flowing though the LED (thus adjusting it's brightness). For circuit protection I added a 5.6V zener diode (MMSZ5232B-7-F) on the μUSB power line followed by a 0.5Atrip resettable fuse (MF-PSMF020X-2). Finally for the light source I went with the TL1F1-NW1,L(S which was a cheap yet powerful LED I managed to find on Element14.
Next up I drew up a schematic in Altium and started the PCB design. One thing to note is that the LED and IC both have power losses with the LED being so significant that sufficient cooling (usually in the form of big copper planes) is required for the LED not to cook itself. As an experiment I decided to completely ignore these and see if I can get away with a tiny board (man I must have some sort of fetish for small PCB's...), here is that the board was going to look like:
And here is how it actually came out to be, minus the μUSB connector:
Surprise surprise the device did not function as expected here is why:
- You guessed it the tiny PCB did not have adequate heat sinking capability, this caused the LED & IC to heat up to ~150°C, a fair bit over their rating... One interesting this I did get to witness was the LED slowly failing, you could see the current being constant but the brightness slowly fading away.
- Something else I should have expected was the voltage drop across the resettable fuse. This reduced the IC input voltage from 5V to ~4.5V, and since it's minimum voltage is 4.5V it did not behave as designed. Removing the resettable fuse solved the problem, until the whole thing got hot again...
All in all i r8 an 8 outta 8, good reminder on the things to look out for when designing a circuit like this.