#DIY

It doesn't quite work but part of the problem might be the EL wire I'm testing with, it's a bit long. Another issue might be that my trimmer values are never getting properly syncronized, adjusting them separately it's possible to change the output voltage measured at the EL terminals a bit but I guess I should calculate proper starting values and then move on from there.

nope, the problem is that I misread the datasheet and connected signals to wrong pins when adding the IC as part to Eagle library, so some rather inportat pins got into wrong places.

Oh well, I noticed some other improvements I need to make to the board when soldering (like not having a ground plane under the inductor when you don't have soldermask, very easy to short to ground there).

With the new layout got the MIC4830 board working for a while but then something went wrong (I already tried replacing IC and inductor and the diode seems to be fine) when I tried with lower input voltage.

MIC4826 board works fine (didn't try with low voltages as 1. see above 2. that driver isn't even meant to work with low voltages, with 4830 one should be able to go as low as 1.5V)

Just to clarify, something is definitely wrong since the board should take 10-20mA of current when operating at 3V and this wanted over 100mA..., no obvious shorts anywhere, nothing gettin really hot, at least not quickly.

Now (again) it would be handy to have something that can do thermal imaging, too bad that kind of equipment is horridly expensive...

Fried the other test board as well when raised the input voltage a bit too much at #hacklabfi, desoldered the output caps and measured them, they seem to be OK so it's either the IC or the diode that is kaputt.

I guess a zener (and probably a reverse-voltage protection diode too) is in order...

Also noticed that there is no way a simple CR2032 is able to supply the required current, so I guess I will end up using very small single-cell LiPos...

Replaced the ICs today, now both boards work again. It seems going to too low input voltage (at least without adjusting RSW to match) will cause the inductor to spike voltage so high that the IC will get fried...

Shame I didn't yet get my LiPo shipment yet, so I could not test these drivers with them.

On the board the RWS is typo, it should be RSW (for Resistor adjusting the SWitching frequency).