Temperature Controller

[lloydi12345]

Hi guys, I am having problems with my project Temperature Controller. It is a device that controls gas valves with timer and displays temperature of a cooking fryer. The circuits involved here are microcontroller, piezo speaker, crystal with two 22pf ceramic caps for clock, tack switches, 7 segment displays, switching regulator, transistors and mostly discrete components.

Few months ago, I made this prototype project using locally purchased components and everything worked fine at temperature 360F. The prototype was used for two months without any failures.

The project later on was passed to China to fabricate it. Testing one unit, it fully functioned at first operated at 340F. Several hours later, a problem showed up. Everything slowed down, including the sound, buttons pressed, displays, clock. I checked the switching regulator, touched it and felt no heat at all. It perfectly looked good! So, I suspected that it could be the 4MHz crystal's problem. I am currently testing it again at 340F and currently it still functions as it should be.

Can anyone assume possible problems here too? Thanks.

 

[3v0]

Are you sure they did not swap out parts for cheaper components.

 

[lloydi12345]

Are you sure they did not swap out parts for cheaper components.

Hi 3v0, That's another thing too that I am currently considering. The units made were incredibly cheap and I have no idea how they managed to build it in small cost considering the parts and labors involved. I am using PIC18 for microcontroller so this should be removed from the list since it has no crappy version. I could be wrong but it couldn't be the power circuit's fault since the circuit doesn't reset or lose any power. I am thinking about the low quality crystal. Using the hot air rework station, I tried heating up the timing circuitry and the problem popped out. I haven't tried it replacing though yet. Still have to confirm it. Could it be also the crystal's capacitors?

 

[alec_t]

I tried heating up the timing circuitry and the problem popped out.

So you've probably located the source of the problem. If the whole system runs slowly that suggests the micro clock is slow. Or perhaps the micro is getting repeatedly reset or stuck in a loop, or a loop counter is getting set wrongly (flash memory only marginally programmed?) when the temperature rises.