CHARGING LED no longer lights

[JohnS]

So, next problem.

I left my A13 WiFi on all night running only from an Olimex LiPo battery so that it would fully discharge, which it did.

Now when I power it from 12V 1A PSU it charges but the CHG LED is not on - why?
How do I troubleshoot this please?

(I did this to try to home in on why Linux crashes with the NULL pointer dereference.)

BTW, for anyone searching in the A13 PDF, this LED is sometimes CHGLED but also CHLED.

John

[JohnS]

I can partly answer my own post: the board no longer has any volts on the LiPo connector!

Oh dear.  I wonder why not...

[olimex]

disconnected Lipo battery cable?

[JohnS]

Ouch - the Olinuxino WiFi board has somewhat blackened solder/pins on the underside of the LiPo connector.  (I have a jeweller's loupe with LED light.)

What could cause that?  I am very used to embedded stuff and do know how to handle boards and in no way has it been mistreated.

I measure 2.98V on the battery leads when off the board so it's not even anything like fully discharged (it's Olimex LiPo 3.7V 1400mA).

My guess now is it's open circuit at the pins or blown chip/etc between them and the board circuitry.  Not sure how to go about checking as so many traces are resist covered and/or tiny.

[olimex]

can you post image?

[JohnS]

Well... here's an attempt at a pic from my phone looking into the loupe!



Oh.  How do I attach it?  When I clicked on the "Attachments and other options" it doesn't offer attachments

John

[olimex]

upload to imageshack and post link?

[JohnS]

Oh.

Try this

John

[olimex]

this looks like short

is it possible the boards to have slip over metal object in this area which to shorted the battery pads and caused this sparking of solder around the pads?

[JohnS]

I can't see any way that could happen.  I've not blown a board up for at least 30 years...

John

[JohnS]

Further info.

Looking at the underside of the VGA connector it is not tidily soldered and the area around the PWR jack is also not good.  There are tiny solder splashes.

The underside of the LiPo appears to have very closely-spaced traces and I se no strain relief or locating pegs for the LiPo connector.  That especially surprises me.  It seems to me it could flex and for a live part that should not be allowed.  Have I missed the pegs or what?

John

[jwischka]

Oh.

Try this

John

Carbon = good conductor. That's definitely going to short.

[JohnS]

A meter shows it's not (currently!) shorted but at other times?  Maybe.

It doesn't look a good idea to reconnect the battery

A pity, as that was the only times it booted Linux OK (from each of 2 different SD cards).

[olimex]

mail please this board to us, we want to diagnostic your SD-card issues as well

[JohnS]

Thanks for the offer.  However, I need it to test development software on.  Maybe I can send it if I make no progress?

It looks so far that the SD card issues are one or more of these:
1. kernel bug (race or whatever)
2. noise, perhaps made worse by what the kernel does at a particular stage
3. other - but what?

#1 doesn't seem very likely

#2 fits all the known facts including that the LiPo fixed it

#3 well, maybe tests on Android would assist and I asked what to do in that regard but Lub didn't get back to me

For #2 I plan to remove kernel features / add delays / not turn on WiFi etc

I would have unplugged the MOD WIFI but it seems soldered on!

The SD cards boot into u-boot and work well from u-boot (as well as on my laptop and desktop) so the actual cards do not look to be a problem.

I may well also rebuild u-boot as it would not be a bad thing if it would do, for example, md5sum and far better RAM tests.

John