Damaged Crystal

[ilia]

Hi.
I have a Nuvoton NUC123 microcontroller, it runs an ARM cortex M0, and it came on a developement board (NUC123 SDK) which contains a 12 MHz crystal.

It recently started freezing every other time upon reset, and in the debugger I saw that it is stuck in a while loop, waiting for the external 12 MHz clock source to become stable.
After a while it stopped starting at all, always freezing on that same line. So I thought maybe the crystal was damaged, so I removed it and soldered a new crystal and, voila! it worked. However after a while it stopped working again with the same problem as before.

My question is: Does anyone have a guess at what's been happening? Is the chip damaged? was the crystal damaged at all?

Does anyone know of a way to test a crystal if it's broken? I have a 2 Mega Samples Per Second oscilloscope at my disposal.

And also, does anyone know how do microcontrollers usually decide whether or not a clock source is stable?

The tech support of Nuvoton haven't answered me yet to these questions, so I'm asking in this forum.

In terms of what I've been doing with the board:
I connected a self made 3.3V UART to +-7V RS232 converter (it might have electrocuted the board?), but during the period between the soldering of the second crystal and the chip freezing again I didn't use the converter.
And I also worked quite a bit with its USB.
I also soldered pins to the board to gain access to the pins of the microcontroller.

I made my best to prevent ESD, by working with an ESD protection bracelet.

I would buy a new board, but I need to know what the cause was so that I don't repeat the same error again.

So that's it. anyone can help?

 

[ilia]

OK problem solved.

While soldering the extra pins to the board I used a water soluble flux. As it turns out, upon drying it leaves a conducting residue. I tested two random points on the board with an ohmmeter and to my surprise it was around 100kOhm.
So I washed my board in a bath of distilled water and then in isopropanol, and then dried it and tested it and it worked.

A very annoying problem indeed.

P.S. actually it could be the flux inside the solder wire I was using rather than the water soluble flux, because the two points with 100k resistance between them were near the area where I used a different solder. And also, after washing the board, the area around my recent solder (the one with the flux inside it) became covered with some white paste that I scratched away with a knife.

 

[davenn]

never heard of water soluable flux that leaves a conductive residue ??

The flux inside cored solder ISNT conductive

before your second post, I was going to suggest an oscillator circuit to test the crystals

cheers
Dave

 

[ilia]

before your second post, I was going to suggest an oscillator circuit to test the crystals

cheers
Dave

Please do, it is still useful, especially since my solution seems absurd and it may break again (I think)

 

[ilia]

never heard of water soluable flux that leaves a conductive residue ??

And BTW, I thought about it, and the residue it leaves behind need not actually be conductive, it could simply collect moisture from the air and become conductive...

 

[hevans1944]

And BTW, I thought about it, and the residue it leaves behind need not actually be conductive, it could simply collect moisture from the air and become conductive...

I always clean a board after soldering by using a so-called "acid brush" dipped in 99% (anhydrous) isopropyl alcohol. Soak and scrub vigorously and rinse with same. Then blow dry and inspect. Repeat until there is NO visible residue. But I guess you know that procedure by now!