3
\$\begingroup\$

I recently bought some 0805 Multicomp PTC/resettable fuses from Farnell. They were MC36213 (500mA) and MC36206 (200mA). I plan to use these in product which may exist in a high humidity (~80-90%) environment, although that is not my main concern: instead, what is written on the package:

CAUTION: MOISTURE SENSITIVE DEVICES!

The product in this package is a moisture sensitive device.

The package in which it is supplied will offer physical and 
electrical protection but will not offer sufficient protection
to prevent moisture ingress.

These items should be baked for a minimum of 48 hours at 125C
prior to reflow.

Moisture sensitivity level: 4

Now, I have no idea how I will be able to bake them at 125°C for 48 hours. I do not have the ability to do so. Especially not at a reasonable price.

Do I have to do this, or is it a mistake? How can I prevent damage to these fuses from moisture? What happens if they do get moisture on them? Will they not work properly to protect a power supply against short circuits and over-current? How can I bake them at such a temperature? I plan to reflow them on a hotplate for the initial prototypes, but I'm not sure if that's the "reflow" they talk about.

\$\endgroup\$
  • 2
    \$\begingroup\$ You don't know how to bake things? You could put them in at 115°C/240°F for a few days in a standard consumer oven (the lower temp to account for the relatively terrible temperature control). \$\endgroup\$ – Nick T Mar 1 '11 at 17:59
  • 1
    \$\begingroup\$ @Nick T, I have to eat too, using the same cooker. Otherwise I would have. \$\endgroup\$ – Thomas O Mar 1 '11 at 18:30
9
\$\begingroup\$

The Moisture Sensitivity Level (MSL) is a measure of how much, how fast, and how sensitive a part is to moisture before being attached. It has nothing to do with moisture sensitivity after the board is assembled, for such information you need to consult the datasheet for storage and operating conditions.

If too moisture much enters the part and it vaporizes during reflow, the high pressure can cause sufficient stress to cause the part to explode, causing a "popcorn" noise. Such failures are usually visible with careful inspection as the package will be fractured, usually near the leadframe. It may be cracked on the bottom with DFN-type packages, however, though the device may appear to be sitting at an odd angle, canted up from the PCB.

Depending on the device, other, more subtle failures may occur. Internal wire bonds could be broken causing intermittent or no connections, or in the case of a PPTC, an effectively reduced current limit. Such errors are probably rare, and of course very dependent on the construction of the device, which you don't know. For prototypes, usually people can deal with it, but for production it's important if you are concerned about reliability and/or manufacturing rejects.

You can keep moisture-sensitive parts in resealable containers if you control the humidity with desiccants or dehumidifiers, and you can use indicator cards to monitor that the humidity did not exceed some defined level. 3M makes them, and there may have been one in the bag from your supplier.

If you do end up with a big production batch of moisture-sensitive parts that have been exposed, you can "bake" them to drive off the moisture. Check with the manufacturer, but bakes can be anywhere from 40 °C at 0% humidity for 1-4 weeks (gentle, but complicated), or 125 °C for some number of hours (more aggressive, but your parts may or may not like it). If you're just making prototypes for development...just make some spares (as one does anyways) and fuggedaboutit.

\$\endgroup\$
  • 4
    \$\begingroup\$ I have to say that in several years of being cavalier about prototype IC storage, I have yet to see a a single failure could definitely trace to, or even suspect, moisture issues. \$\endgroup\$ – Connor Wolf Mar 2 '11 at 7:30
  • \$\begingroup\$ @Fake: I haven't either, but unless you/your company does protoduction (a dubious blend of prototyping and production) on an even bigger scale than I've seen at mine, the low frequency of failures in the relatively few moisture-sensitive parts on the BOM probably won't manifest. \$\endgroup\$ – Nick T Mar 3 '11 at 1:44
6
\$\begingroup\$

The reason for having to pre-bake the parts is to prevent damage during the thermal gradient applied during automated soldering (wave / reflow / whatever) caused by expansion of moisture absorbed by the part during its pre-assembly storage.

There isn't an issue once the part is installed in the circuit.

Here's some more information on the subject.

\$\endgroup\$
  • 3
    \$\begingroup\$ It's not really the thermal gradient that will cause damage in moisture sensitive parts, but the vaporization of said moisture causing mechanical problems (e.g. popcorning). Good article link \$\endgroup\$ – Nick T Mar 1 '11 at 18:02
  • 1
    \$\begingroup\$ Clarified, for clarity. :) \$\endgroup\$ – Adam Lawrence Mar 1 '11 at 18:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.