How does air pressure affect electronic components such as MCU, Crystal and others?

[BasePointer]

Hi,

We have a product that is dust and water proof.

There is no way the air inside to go outside. And also, from outside to inside.

The product also heats yourself. The air expanded inside over temperature causes pressure to rise accordingly.

At that state if I open the cover of the product, the MCU loses the clock. and generates loss of clock condition.

The MCU datasheet(LC60) doesn't indicate any values or test result for pressure. That means MCU works properly at all pressure condition?

If you have experienced, can you share other components behaviours at pressure such as electrolytic capacitors, smd resistors, crystals?

10x,

BP.

[mke_et]

When I meant equalizer, I meant building the case more to still be sealed but allow for expansion and/or contraction. But that also limits how much pressure differential would be allowed and still be practical.

As to the crystal, I was thinking more along the lines of sudden changes flexing the metal can, and that leading to undesired geometry changes inside the can...

[mke_et]

If it were me, I'd build up a 'test circuit' without the MCU, but all the other components. Then see if I can get clock/osc problems in a simple circuit. If so, then
I'd see if I could build a 'walled box' where the PCB is the wallm with nothing but the crystal inside, and see if it still failed.

Yeah, I'd suspect the crystal only because it's a 'case' with space. Next I'd suspect any electrolytics...

On the other hand, you could do the same thing with the MCU fed from and external
osc input with nothing but the MCU in the box and see if just the MCU configuration fails.

Hmm, another option would be to put some kind of 'equalizer' in your case to limit air
differential.

Message Edited by mke_et on 2008-03-26 01:34 PM

[bigmac]

Hello,

---

mke_et wrote:
.
Hmm, another option would be to put some kind of 'equalizer' in your case to limit air
differential.

---

I would be cautious about providing a smaller "breather" hole in the enclosure, if this is what is meant by an "equalizer".  The reason is that humid external air would enter the enclosure, and with reduction in temperature, may condense.  This could eventually result, over a lengthy period, with an accumulation of moisture in the bottom of the enclosure, unless drainage is also provided.  Alternatively, a desiccant could be provided within the enclosure.

If the crystal you are using is with a metal can, changes of external air pressure shouldn't be a problem since the can provides a hermetic seal.  My understanding is that this type of crystal may be filled with nitrogen, at normal pressure, during manufacture.

Regards,

Mac

[BasePointer]

Hi,

Yes the part broken is crystal 32768Hz:

http://www.raltron.com/products/pdfspecs/crystal_R38_R26_NPB_series.pdf

The crystal is broken as soon as we open the case. And the mcu couldn't lock to it again.

We will test more samples with different configuration to identify repeatability, faulty ratio.

Unfortunately the broken sample was working over three months

Regards,

BP.

[Denn]

Hello BP,

You can raise a service request with FSL for more accurate details regarding affects of pressure on MCU components.

Regards,

Denn.

[bigmac]

Hello BP,

How much increase of pressure is actually occurring?  I seem to recall a distant memory that, for a constant volume, the pressure will be directlly proportional to absolute temperature (the name "Charles'  Law" comes to mind).

If the temperature of the air within the enclosure were to increase from 20C to 60C, a rise of 40 degrees, I calculate a 13.7 percent presure increase (i.e. 2 psi).  I would be rather skeptical whether this small increase alone would be likely to cause a problem.

You seem to be implyng that there was no problem as the temperature (and pressure) slowly increased, but only with a sudden decrease of pressure.  Did the clock automatically resume operation after the event?  Is the loss of clock consistently repeatable?  Does the event also occur when using the internal oscillator in lieu of the external crystal?

Regards,

Mac