To Glue or Not to Glue

It's not necessary at all. It would increase the life of the board. The vibrations will stress the leads and the solder joints. I would think more so on heavy machinery like a farm tractor.

What are the results of your vibration testing?

This will end any discussion regarding the robustness of your PCB.

Just curious: Is it listed by U.L.? RoHS compliant?

Well that is my struggle, to my knowledge there has never been any vibration testing or if there were it is not documented. I am relativly new to the company and recently tasked with being the quality department (we've never had a quality system before) and am struggling against an "I think I saw one fail once" mentality. We are RoHS compliant but currently not U.L. listed.

Commonly referred to as "staking" in the aerospace electronics industry it was common practice when leaded components were used in high vibration environments. With the advent of surface mount components it is not as critical due to the very small mass of the components.

There's not much to do but try some and see if they fail, since you have no vibration data.

Not sure where ozzb is going with his comments, but staking components to the circuit boards to reduce vibration induced failures is common. It will add to the cost of the assembly.

Cheers.

Most of our caps are of the surface mount variety, the polyfuses aren't. I guess the ME in me says the lack of mass would make the likelyhood of failure from vibration with the polyfuses slight.

Lynlynch you ever sit a tractor and plow a field? Most of the boards that I have worked on or replaced on heavy machinery has been encapsulated. I am not say that it has to be done. What would be his best argument would be maintenance records on the equipment. What has been the cost to the company for the board failures due to vibration under warranty?

ozzb,

I have sat on a tractor and plowed a field. Not for a while, though.

I agree that with no data what you suggest is the proper course.

Cheers.

Hey there ozzb.

Your comment "...you ever sit a tractor and plow a field" is sort of what was on my mind in comment #2. In the summer of '68 or so I spent about 11 years on the seat of a Ford Farmall. Plain and simple, the equipment of that era was a tooth-rattling, pile-driving (if I were a little older) bone shaker. For 12 or 14 hours a day, anything attached to that tractor was shaken HARD and STEADY.

Equipment today? Dunno, but I believe it is a lot more comfortable to operate the newer machines. Does the chassis still vibrate as badly? Dunno, but I think I would probably figure it so, just to be safe.

Should OP adhere components to boards? Dunno... if production is small, be prepared to replace some... if the consequence of failure is acceptable (as you said before). If consequence of failure is unacceptable, then an overly robust part (even if an increase of cost) is probably prudent (as lyn said).

I have been involved in farming on and off since 1969. I started farming for myself 5 years ago. Yes, the cab of today's equipment is very comfortable, almost like riding in a Cadillac. The ride of the chassis has not changed much though. 99% of the mounted boards and sensors are in fact encapsulated.

Just a note to the OP - Been in many, many situations, out in the field when a sensor or board has failed and needed replacement. As a rule of thumb when this happens, the operator of said equipment is usually quite ticked off due to lost time and lost money.

I can only assume that you do not supply to a USA OEM because because most if not all require encapsulation per their specifications for chassis mounted boards/sensors not only for vibration but for dust/moisture protection also.

Some farmers will try an off brand no name replacement part to save some money (in this economy especially for a farmer, who wouldn't) but generally if the no-name replacement part fails prematurely, they will not buy from the no name manufacturer again and will pay the few extra bucks for OEM replacements.

Bottom line, whether you are an OEM manufacturers sub or are manufacturing after market parts/accessories, build quality parts and you will build satisfied customer repeat business! Build junk and word will travel at the speed of light!

Good day to all - KJK/USA

Your idea of gluing the heavier and/or higher parts (relative to the print distance) is not bad.

The print connection is the electrical (soldered) solution, but has nothing to do with the mechanic load and stress on the wires.

The wires of heavier components are never designed to deal with these mechanic issues.

With the use of some heavy condensers you should even consider to make the condenser hold a (smaller) print and attach the condenser in a clamp.

If you can afford it, especially for your application, to glue (e.g) with a heat stick glue, I advise to do so. A diesel engine can produce severe vibrations.

Investigate covering your boards with an epoxy type of potting material.

This may make it better shock resistant and also prevent the effect of dampness containing corrosive elements used in farming from reaching your components or solder joints.

You may have to experiment with different types to find the best solution.

Thank you all for your responses, very good all.

I will carry out vibration testing. We have always glued anything that we think may be a problem so I have no historical data to research, so testing is my only choice.

Glue the larger parts to secure them from vibration/shock damage. Full potting is good for certain applications, but it can be expensive. We used a thin layer of UV cured conformal coating to protect our boards from moisture, dust, and insects. Very tough material once cured. Circuit board repair and spot re-coating are possible. Your process costs would dictate whether or not defects/returns should be repaired or just scrapped.

Why not a simple answer?

In PCB mfr'ing it is common to use an epoxy to affix larger mass components (fat thru-hole capacitors, heatsinks and other) onto the board. Not just for the sake of subjective environment but also for shipping and possible mishandling by the user. I've had very good success with this approach.

I would guess it would only take one mfr. production line employee to do this up to about 100 boards in maybe one hour.

Hope the OP deems this as being in a 'nutshell', and serious.

Yup, just pot it (epoxy) and be done with it.

You will also help prevent reverse engineering of your product by the competition. You will save money and increase profits, because you will eliminate repairs to pcb's and increase profits by selling whole new boards, and not repairing old ones, which is a pain in the a$$.

put the boards through testing.. then encapsulate in epoxy. It will protect against moisture, vibration, static, and even protect your customers... you won't regret it.

Chris

Do you rework production failures? Or just scrap it and get a new one?

If the first, use RTV staking. If it is work or scrap, epoxy the parts or pot the whole thing in RTV. Neither is easily re-workable.

Just got into an automotive project - the power supply PCB is submerged into EPOXY by

23mm and the epoxy covers all: mechanical parts, relays totally mainly to act against bad vibs and the build up of humidity on the wires which could leed to corrosion.

But all connectors (for cables and spade type fuses) have to be tight fit plastic and look upward to make sure that the contacts are not exposed to the EPOXY.

Seems to me simple than the the GM I have seen years ago with a dispenser following all IC pins in a complex pattern.

Regarding repair work - in automotive the final product is tested after the Poting and thrown away if defective - a) too complicated & too costly to repair b) in most cases a repair not considered due to safety and legal reasons (What if the real root cause was not found and the repair just covered up for a short period of time, or the board has defects implemented during BGA removal and fails again?) - so better send the part to the bin.