Protecting boards from corrosion

[brabus]

Dear all,

I am currently coating my boards with specific silicone to protect them from moisture and atmospheric agents. The boards are mounted in outdoor equipment.

I am trying to find out alternatives to coating, for two reasons:
- it does not completely solve my issues with aging and corrosion. Some boards come back with corrosion around or in the connectors
- I am planning a stacked board Setup. How can I protect the board-to-board connector?

I tried to take Inspiration from engine ECUs, but I see many different approaches: complete potting, Surface coating, naked board in tightly insulated enclosures...

Anyone willing to share their experience?

Thanks in advance.

[james_s]

I like the tightly sealed enclosure idea, my observation is that unless you completely pot the PCB, it's really difficult to fully protect it, and potting brings its own set of issues.

[Doctorandus_P]

There are clear jelly like (2 component) potting compounds which are easy to remove if needed.
These are often used in hybrid modules and Dave told something about them (including their name) in a teardown of a cypress AAD (Automatic activation Device (for the reserve parachute)) .

Anoter way would be to remove the connectors from the equasion.
Either solder wires directly between the boards or solder a piece of flex between the boards.
Some board houses can make a combination of flex and rigid in a single pcb and these can be folded.

If your pcb does not generate much heat (power) you could put it in a plastic bag and top it off with silica gel.

Completely hermetic rigid enclosures are problematic because air expands and contracts with temperature changes which will cause pressure differences. A bladder can be added to allow for this expanion/contraction.

In industrial settings it is (was?) common to add a heater to scare away condensation.

You might be able to protect connectors with application of vaseline, silicone grease or similar.
The contacts will push the grease away and still make contact, but moisture has no chance to get to the metal.

Using connectors that simply do not rust may be an option.
For example: Car trailers have used brass contacts for ages, and those connectors are out and unprotected in all weather.

[T3sl4co1l]

Potting seems like not a good idea for board-to-board (pluggable) headers.  If they're soldered on both ends, that would be fine.  (And, uh, it's not like you'll be wanting to dig out all that potting just to get at the lower board, anyway, right?  Right?...)

Sealed enclosures are good, but mind they are still permeable -- air and moisture will diffuse into them.  Conformal coating for the boards, and corrosion-resistant platings for connectors (e.g., 1µm gold, not just flash?), are good ideas.

Careful washing of the PCB assemblies will also help reduce electrolysis.  It's fine if water condenses on the board and it's not conductive enough to cause a problem.  But it is a problem if it condenses there, and dissolves salts and fluxes sitting on the board already.

The gold standard, more or less -- typical for avionics (flight instruments) -- is a hermetically sealed metal enclosure.  Fused ceramic/glass feedthrus are used for connectors, and soldered metal joints are used to close the assembly.  The enclosure is purged with dry nitrogen gas, and the cover is soldered on (usually all at once, with an induction heater).  Or there may be a gas fitting, so it can be sealed, then vacuumed, heated (baked out), and back-filled with inert gas.  Or just left at vacuum, for some instruments (refrigerated cameras, for example, or physics apparatus).

Tim

[Bud]

Have you tried conformal coating spray? It has been used in the industry for longer than i can remember.

[brabus]

Dear all,

Thank you very much for your suggestions and for the time you spent writing them.

I will try different approaches on a couple of prototypes and conduct some accelerated corrosion test; I think we will go for the complete potting of the boards and a new set of rugged connectors.

Thank you and Best Regards.

[tautech]

PCB lacquer spray in aerosols.

Clear, dries semi hard and you can solder through it, great on new PCBs with a very light coat until you get around to populating them. Or you can spray thicker coats on.
I often use it after rework on anything that is exposed to climatic variations.

[Styno]

I know it's an old topic, but I just didn't think this warrants a new one...

Just got a sensor back that: 'eats batteries' according to the client. Did an EnergyTrace power measurement (Very nice BTW!) and instead of the usual 6uA this thing required 23 mA   

The solder joints looked a bit dull so I checked with the microscope (see attachment) and to my horror it seemed like the solder joints were 'melted' under components and the green/blue hue hints to severe corrosion.

The thing is: the electronics are housed in an RVS tube sealed with a nylong plug that has two silicone o-ring seals, watertight connectors (M12) and a desiccant bag.

[Smallsmt]

Sometime desiccant bags are wet!
You need to dry before use!
We had a lot of problems with totallly sealed logger system and desiccant bags we took from stock.
The water came out of these bags during installation time.
Another thing for lead free electronics you need to clean the pcb well to prevent tin whisker.

[jmelson]

Did you clean these boards?  If the cleaning did not remove all solder flux under the components, then absorbed water from the cleaning could remain, and eventually lead to corrosion.  We have a few things that run at up to 500 V across single SMT passives, so we have to clean them like crazy until the leakage is acceptable.  What we do is clean in an ultrasonic bath with pure alcohol, several times, rinse with deionized water, then bake dry at 50 C overnight.  Then, test for leakage. Repeat if leakage test is not good enough.

Jon

[mrpackethead]

Another thing for lead free electronics you need to clean the pcb well to prevent tin whisker.

Meh. Tin wiskers were a problem because PCB manufacturers were using pure tin for plating.   no amount of cleaning solved that problem.     You'd be hard pressed now to find a plater who used pure tin, as they all dope the tin with tiny amounts of lead which stops the whiskers.  ANd if you are doign PbF seriously, you're probably not using tin plate anyway

[Styno]

Sometime desiccant bags are wet!
You need to dry before use!

The customer replaces the batteries when necessary and I recommended they place a new desiccant bag each time, so I don't really have control over that. But I'll remind them to use new (or freshly baked) ones.

Did you clean these boards?  If the cleaning did not remove all solder flux under the components, then absorbed water from the cleaning could remain, and eventually lead to corrosion.  We have a few things that run at up to 500 V across single SMT passives, so we have to clean them like crazy until the leakage is acceptable.  What we do is clean in an ultrasonic bath with pure alcohol, several times, rinse with deionized water, then bake dry at 50 C overnight.  Then, test for leakage. Repeat if leakage test is not good enough.

The solder used is a no-clean type from ChipQuick (SMD291 iirc) and we do not clean them nor are these coated. These sensors are low-power, low-voltage (highest voltage is 3.6V). The consumption is usually exactly what I expect based on the component specs so any leakage is insignificant, perhaps because the board isn't high-density helps here.
 
 

Another thing for lead free electronics you need to clean the pcb well to prevent tin whisker.

Meh. Tin wiskers were a problem because PCB manufacturers were using pure tin for plating.   no amount of cleaning solved that problem.     You'd be hard pressed now to find a plater who used pure tin, as they all dope the tin with tiny amounts of lead which stops the whiskers.  ANd if you are doign PbF seriously, you're probably not using tin plate anyway

Being in the EU, these are lead-free HASL FR4 boards and all components are RoHS compliant. I checked a few datasheets but most don't specify the plating.
 
I have never seen devices return from the field this way, even after several years of operation. Despite the heavy corrosion and apart from the high power consumption the board worked just fine.

[mrpackethead]

Being in the EU, these are lead-free HASL FR4 boards and all components are RoHS compliant. I checked a few datasheets but most don't specify the plating.
I have never seen devices return from the field this way, even after several years of operation. Despite the heavy corrosion and apart from the high power consumption the board worked just fine.

They almsot certainly are not tinned then.

[Ice-Tea]

The thing is: the electronics are housed in an RVS tube sealed with a nylong plug that has two silicone o-ring seals, watertight connectors (M12) and a desiccant bag.

Nylon isn't great as a moisture barrier AFAIK. And are the o-ring seals greased (vasiline would do..)? Perhaps pack a moisture indicator next time a battery is replace?

[Smallsmt]

Another thing for lead free electronics you need to clean the pcb well to prevent tin whisker.

Meh. Tin wiskers were a problem because PCB manufacturers were using pure tin for plating.   no amount of cleaning solved that problem.     You'd be hard pressed now to find a plater who used pure tin, as they all dope the tin with tiny amounts of lead which stops the whiskers.  ANd if you are doign PbF seriously, you're probably not using tin plate anyway

It happened on quality lead free PCB from german manufacturer the reason was remaining solder flux on the boards plus air humidity allow the tin wisker growing.

[Styno]

Another shot of the victim compared to a fresh one, just for the lulz...

The thing is: the electronics are housed in an RVS tube sealed with a nylong plug that has two silicone o-ring seals, watertight connectors (M12) and a desiccant bag.

Nylon isn't great as a moisture barrier AFAIK. And are the o-ring seals greased (vasiline would do..)? Perhaps pack a moisture indicator next time a battery is replace?

I'm not entirely sure it's nylon but the plug is relatively thick (>5 mm). The o-rings were sealed with vaseline, but I suspect they are not re-greased as they feel pretty dry. I'm considering changing the design by adding a moisture sensor so the user is notified when the humidity level goes above -say- 30%.

The moisture indicator idea is nice too

[ptricks]

Conformal coating will work for the pcb if applied correctly, I have even used liquid tape before but that can harm some components.
Use a dielectric grease for the connectors, same stuff used for wire nut connections in wet locations, has an appearance of vaseline but not harmful to plastics so you can really pile it on the moisture prone areas. Vaseline can degrade some rubber types so be careful with o-ring usage.

[mrpackethead]

Another thing for lead free electronics you need to clean the pcb well to prevent tin whisker.

Meh. Tin wiskers were a problem because PCB manufacturers were using pure tin for plating.   no amount of cleaning solved that problem.     You'd be hard pressed now to find a plater who used pure tin, as they all dope the tin with tiny amounts of lead which stops the whiskers.  ANd if you are doign PbF seriously, you're probably not using tin plate anyway

It happened on quality lead free PCB from german manufacturer the reason was remaining solder flux on the boards plus air humidity allow the tin wisker growing.

 Go and learn why Tin whiskers grow.   The Flux had NOTHING to do with it.   Google is your friend.       But heres a good summary of the papers you coudl go and read on the topic.
http://thor.inemi.org/webdownload/newsroom/Presentations/SMTAI-04_tin_whiskers.pdf
Internal metalic stress is considered as the driving force that causes the growth of Sn whiskers.    You can accelerate   it by high temp and high humidity.   Your so called High QUality German Manufacgturers PCB must have been using old school plating techniquest where pure or nearly pure tin was being used.           Flux and solder have been mistakingly blamed by lots of people, but its well understood engineering what the problem is.     
Simple fix for stopping Tin Wisker growth.  Don't use Pure Tin Plated PCBs's   

[technotronix]

1. Avoid the humidity on your PCBs.
2. Use a protection of the PCB with a conformal coating.
3. Avoid your PCB to be in contact with metals that can produce corrosion.
4. Avoid have your electronic circuits in hard environmental conditions.

Learn more about corrosion on PCBs and how you can prevent it.