Here is the second of two stories about relay failures caused by our assembly process:

We have another product with mostly surface mount parts, along with one through-hole relay and some 48-pin through-hole connectors.  After the SMT parts are reflowed, the boards go through a wave soldering operation to attach the four connectors and the relay.  Our first batch of 50 boards had 6 fallouts that were traced to "bad" relays.  Since I had recently been burned by water in "sealed" relays (another style, from another manufacturer), my paranoia kicked into high gear.

The datasheet shows these relays as "Fully sealed", with a note to "Contact your OMRON representative for the ratings on fully sealed models."  Searching the relay manufacturer's website turned up a few more documents.  One document, titled Electromechanical Relays Technical Information shows that for Fully Sealed relays:
Boiling cleaning and immersion cleaning are possible. Ultrasonic cleaning will have an adverse effect on the performance of relays not specifically manufactured for ultrasonic cleaning. The washing temperature is 40°C max.

My guess is that our wash process (and probably yours, too) is above 40°C.  This is probably how I screwed this one up . . .

Opening up the 6 failed relays showed significant water inside.  Since these boards were produced last month, they had been sitting on the shelf for several weeks, with the water and other chemicals causing problems, including mechanical binding.  My real fear was that other boards that tested "good" might have water inside and would soon go bad.  The failed boards showed several different problems, but a common feature attracted my attention.  During the initial board testing, as I ramped up the DC power supply (that activated the 24V rated relay coil), all of the good boards made a little "tick" sound as the relay armature took up slack in the system, but didn't have enough force to overcome the return-spring force.  This sound occurred at about 8V, except for the "bad" boards.  I touched a 9V battery to the coil drive pins while holding a bad relay against my ear.  I could only hear a slight "tick" sound.  Holding a good (new, never been washed) relay up to my ear, then touching the battery to the pins gave a louder, stronger "tick" that had a little ringing character that was not present in the bad relays.  The difference was subtle, but similar to the difference in sound from dropping an old (Copper) penny and a new (Zinc) penny.  I cut the top from a good relay, listened to the sound as I touched the coil terminals to a 9V battery (reference, good).  I ran water over the exposed (good) relay, and shook off the excess.   Activating the dampened "good" relay gave a sound that was more muffled, lacking the "ring" of the little vibrating parts as the water film damped the vibrations.  With this test, I was able to check all of the "good" boards, finding that they all had the "ring" of goodness (hopefully with no significant internal water).

I'll send some pictures to Steve for his wonderful picture hosting service.

Once again, our planned solution to the problem is to add the relays after the final board cleaning, using a no-clean flux solder.


Once again, our conclusion is: don't wash relays, regardless of the datasheet information.



Later,

Carl



p.s.  there will be no follow-up water-in-relay stories from me because I'll never allow a relay to run through a wash system. This goes for potentiometers, too, since I've had "sealed" pots fail after washing.





Carl B. Van Wormer, P.E., AE7GD
Senior Hardware Engineer
Cipher Engineering LLC
    21195 NW Evergreen Pkwy Ste 209
    Hillsboro, OR  97124-7167
    503-617-7447x303
    [log in to unmask]<mailto:[log in to unmask]>     http://cipherengineering.com<http://cipherengineering.com/>

This message may contain confidential and/or proprietary information, and is intended for the person/entity to whom it was originally addressed. Any use by others is strictly prohibited.  If I sent this to you by mistake, please be nice and delete it, and then tell me of my mistake so I can send it to the right person.



______________________________________________________________________
This email has been scanned by the Symantec Email Security.cloud service.
For more information please contact helpdesk at x2960 or [log in to unmask] 
______________________________________________________________________