Micro stuck in reset

dluu13 · Joined: 28 Sep 2018 Posts: 395 Location: Toronto, ON

Hi all,

This isn't a CCS question exactly but I'm hoping to get some info about this.

There have been a couple of instances now where I have some boards stuck in reset even though they program properly. When I start probing around, I notice that MCLR is stuck low. Here is my circuit on the MCLR pin:
https://sta.sh/02do8echup6

After removing the capacitor C8, my chip is released from reset. Do you have any ideas why this might happen? My capacitor is a 50V 0u1 capacitor so I don't think there should be any breakdowns here.

I understand the capacitor is required in order to prevent any random resets due to noise but I'd like to know why I'm getting these failures.

temtronic · Posted: Fri Aug 28, 2020 12:58 pm

random thoughts...

1) maybe a bad cap ??? rare but possible....
2) poor PCB ?? some 'semi conductive stuff' bridging the cap layout ??
3) high humidity causing a semishort ??
4) bad reset switch, kinda stays closed ??
5) it's friday and you want to fly up north for the weekend ???

dluu13 · Joined: 28 Sep 2018 Posts: 395 Location: Toronto, ON

newguy · Joined: 24 Jun 2004 Posts: 1908

Re: humidity. What sort of flux and/or solder paste/flux is used, and how is the assembly cleaned? Leftover flux residue + humidity can become conductive.

dluu13 · Joined: 28 Sep 2018 Posts: 395 Location: Toronto, ON

Ok that's really useful... I used some no clean "halide" flux, and I cleaned it with a MG chemicals flux remover. It's not that great I still see a bunch of streaks on the board. You are probably right though. I hand-soldered all the stuff and there were some things acting up on a couple of these I was testing. Fixed it by reflowing some pins with the soldering iron and hitting it with some more flux remover.

I figure that the capacitor should be ok though, since the pads (0603 footprint) are so big at least compared to the MCU legs.

Ttelmah · Joined: 11 Mar 2010 Posts: 19518

Other thought, is obviously some form of whisker on the PCB.

What chip is involved?.

No clean halide fluxes, can be absolutely 'awful' for cleaning. Unfortunately the
chemicals used in these fluxes, can result in damp being absorbed, and dendrites
forming between the pads. In 'professional' use, these require
very accurate temperature control during the soldering process if problems
are not going to appear. Cleaning after hand soldering can be a real problem.
The white residues are normally salts, that can easily result in conduction.
Have a look at:

<https://www.aclstaticide.com/blog/why-clean-no-cleans-printed-circuit-board-flux-removal>

These fluxes are actually harder to clean than traditional systems.
Have to say, I'd not use a 'no clean' flux on boards that are being hand
soldered.

temtronic · Posted: Sat Aug 29, 2020 4:23 am

In the good ol' days, before SMT was even thought of, guy I got PCBs from in Toronto, put assembled boards in a regular house dishwasher, to remove all oils and water based flux residue(probably just water), then a heat dry cycle. Once cooled for a day, he'd test on a 'bed of nails'.
I KNOW humidity can cause 'random' problems. Had a panel work for 3 years in a rough industrial enviroment only to lose communications after a week of hot, humid weather. Culprit was a 'bad' cap used for the 24hz master clock.
I don't know how they clean SMT PCBs but that HAS to be 'high tech' as anything under devices could be a nightmare to deal with.

dluu13 · Joined: 28 Sep 2018 Posts: 395 Location: Toronto, ON

I have been using an MG Chemicals flux remover on the halide flux which appears to be an acetone based cleaner.

Are there ones that I could get that come in a large bottle format?
Should I be looking for rosin flux? Any preference between non-activated, fully activated, and partially activated rosin?

I don't particularly like flux pens because I always forget to cap them and then they dry out.

newguy · Joined: 24 Jun 2004 Posts: 1908

Ultrasonic heated cleaners are supposed to be the standard for removal of the flux residue. Can't comment as I don't have one yet. In the meantime I use MG flux remover, a pig's hair brush, then a close inspection (microscope) and a dental pick to scrape the white salt residue away.

Seriously looking forward to the ultrasonic cleaner.

temtronic · Posted: Sat Aug 29, 2020 1:44 pm

hmm.... possible option ...
have a PCB 'house',preferably local, make them for you ??? Has to be a few in Toronto still ??
I know not cheap, but if you factor in YOUR TIME to debug, it might be an option ??

dluu13 · Joined: 28 Sep 2018 Posts: 395 Location: Toronto, ON

It's better to have them make it at volume... I only wanted a couple of pieces for testing. I am using pcbway for bare board manufacturing. I might try their assembly service.

asmallri · Joined: 12 Aug 2004 Posts: 1635 Location: Perth, Australia

It is conceivable but unilkely the capacitor is the problem. If it was, then you would be able to measure leakage across it and it would have to be serious leakage given you have a 10K pullup.

I suspect the problem is more likley to be poor positioning of the manually soldered 0605 capacitor on the pad resulting in shorting of the pads.
_________________
Regards, Andrew

http://www.brushelectronics.com/software
Home of Ethernet, SD card and Encrypted Serial Bootloaders for PICs!!

dluu13 · Joined: 28 Sep 2018 Posts: 395 Location: Toronto, ON

I totally believe that. I had a batch of boards that started failing after about a day or so, also stuck in reset. That time I was using 0u01 caps. After removing the cap from the board I measured the caps and they were a dead short. :(

Ttelmah · Joined: 11 Mar 2010 Posts: 19518

What type of capacitor were these?. Failures like this on ceramic capacitors
would almost certainly be down to thermal stresses when soldering.

dluu13 · Joined: 28 Sep 2018 Posts: 395 Location: Toronto, ON