2
\$\begingroup\$

I have designed a PCB and soldered some SMD components on to it. When I check it, it doesn't seem to work. It was my very first SMD soldering experience and I think I used too much hot air. While investigating the issue, I found out a moisture level factor in ICs. So I read this article ICs with humidity or moisture sensitivity - bake recommendations and found out that water vapors can reside damaging the ICs from inside.

I had the ICs weeks before soldering and they were also the very first time I experienced a TQFP44 brand new IC (PIC24EP256GP204). I was careful enough not to touch the ICs by hand but to use an anti-static tweezer. Now all that is done and the ICs not working, I found out about this MSL(Moisture sensitive level) factor.

My problem is, when we're cleaning the boards with IPA, prior to soldering, does that affect the MSL as IPA could go inside the IC package damaging it when heat is applied? (Popcorn effect?)

Thank you very much in advance!

\$\endgroup\$
5
  • 2
    \$\begingroup\$ I would imagine IPA would evaporate before entering the IC. Just a guess though. Anyway, if that is the issue (be it with water or IPA) you can prevent it by baking the parts. \$\endgroup\$
    – Wesley Lee
    Jan 29, 2018 at 3:17
  • \$\begingroup\$ Did you inspect solder joints under a microscope? Did you use a hot plate? \$\endgroup\$ Jan 29, 2018 at 3:40
  • \$\begingroup\$ Does the Xtal or Internal Clock work? \$\endgroup\$ Jan 29, 2018 at 4:11
  • 4
    \$\begingroup\$ While that is a remote theoretical possibility, unless you want to throw everything away and start over you should focus instead on determining what specific aspects of your circuit do or do not work. The overwhelming likelihood is that the problem is a simpler error of design or assembly with your current prototype. Better to identify that and correct it before you build another. \$\endgroup\$ Jan 29, 2018 at 4:20
  • \$\begingroup\$ This question looks fine for on-topic-ness to me \$\endgroup\$
    – Voltage Spike
    Jan 29, 2018 at 7:03

2 Answers 2

4
\$\begingroup\$

I ran a web search using keywords like "IPA moisture issues" or "IPA moisture reflow", etc. and came across various blogs, company websites, etc. that touch on the use of IPA to clean PCBs prior to (and after) board assembly.

From the articles I read, IPA isn't that great at removing nonpolar residues (including various oils, grease, and other hydrocarbon residues), so its usefulness as a board cleaning agent prior to board assembly is somewhat questionable.

Other articles explain that PCBs, after being cleaned (e.g., with IPA or detergents), should be baked to remove residual water moisture prior to board assembly. FWIW, Link Hamson's website (linkhamson.com) currently has a page titled Moisture Sensitive Device Handling with a subsection titled "COMPONENT BAKING OVENS" that identifies standard IPC-1601 Printed Board Handling and Storage Guidelines as well as some recommended bake times and temperatures for PCBs prior to board assembly.

An article published by Michael Watkins of Chemtronics titled IPA as Universal Cleaner: Advantages & Disadvantages discusses IPA's hygroscopic properties, and it mentions how IPA exposed to air absorbs moisture until it reaches equilibrium at 65% IPA and 35% water. Yikes! Furthermore, after the IPA evaporates the water stays behind on the board. In some cases there is sufficient water residue to cause serious problems both before and after board assembly. For example, after cleaning an assembled board with IPA, water residue can get trapped between the leads/lands of fine pitch packages and can stay trapped there for days if it's not removed by baking. This water residue can cause corrosion (imagine packaging the board while it's still wet) and short-circuiting (powering up the board while it's wet).

Various articles mention that IPA isn't particularly good at cleaning some fluxes used in soldering processes, and IPA can indeed be absorbed by and damage plastics, contrary to popular opinion. Mike Jones, V.P. Micro Care, in his response to a question titled Cleaning an assembled board with IPA states that IPA saturates at flux concentrations of around 2%, so LOTS of IPA and scrubbing is required, and IPA has a tendency to smear the flux residue around the board, not to mention flow underneath components and behind fine-pitch leads where it cannot be reached and absorbed with TechWipes, for example.

With regard to handling static-sensitive parts. Handling the parts with ESD tweasers is not sufficient. The entire workspace needs to be designed for ESD assembly, including—at a bare minimum—a properly grounded ESD worksurface/mat, and you wearing an ESD wristband (or ESD jacket/smock) that is connected to the ESD mat. ESD-sensitive parts must never be removed from their ESD protective packaging or handled except at the ESD workstation. Boards containing ESD-sensitive parts must only be handled at an ESD workstation, and must be stored inside closed ESD-protective packaging (e.g., inside an anti-static re-sealable bag) when being stored or transported. The soldering equipment—handpieces, hot air reflow, etc.—must be ESD qualified also. CMOS components like microcontrollers are VERY, VERY, VERY susceptible to ESD damage. Your clothes (e.g., poly-based fabrics) sliding around on your body as you sit in a chair can easily generate static voltages on your body and clothing of sufficient magnitude (through the triboelectric effect) to destroy CMOS devices, especially in dry climates.

Hope this helps.

\$\endgroup\$
1
\$\begingroup\$

I'm no chemist, but I'm willing to bet that the standard flux remover is a stronger solvent than Isopropol Alcohol and that leaves IC's untouched. Most IC's are protected by the epoxies surrounding the silicon. I use water to clean off my PCB's with water soluble flux and that is an industry process (but not as much these days, no clean is more widely used.)

Intuition tells me that soldering right after rinsing is not a good idea, I'd wait for them to dry out. As long as you didn't submerge the board for a long period of time I don't think much liquid would get inside the package.

There are parts that have a higher MLS (like 4 and above) that I'd think about applying liquid to. And there certain electro-optical, MEMS parts and temperature and humidity sensors (especially any IC with a diaphragm or entry to the silicon) that I would never apply liquid to.

In your case you should be fine, start troubleshooting with power, then clock then JTAG against a reference schematic. Or put a known good IC on your board.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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