Root Cause Determination of Epoxy Encapsulated Electronics Modules

[squonkdb]

Sorry if a similar question has been asked and answered...

We have received an electronic device that we manufacture back from a customer stating that the device did not work properly. We confirmed their non-conformance. The electronic portion of the device is coated in an epoxy compound which makes analysis to determine if a component failed impossible.

The device has to pass four rounds of functional testing prior to shipping, so the only thing that I'm sure of is that the unit was functioning properly when it was shipped.

How do I determine root cause and corrective action in this situation? The cause of the non-conformance could be anything from a failed component in the device to a customer handling problem.

There is not a history of similar failures of this type for this product, however we did change the mechanical side of the design recently.

Thank you in advance for your input.

 

[Popeos]

Re: Root Cause Determination

Hi squonkdb

Note that you are looking for the cause of the problem. It might be a root cause straight ahead, butit is likely to be a technical cause first, and then you will try to understand why the failure mode occured and you'll cath the root cause

First you need to check whether this component actually went through all the expected testing (traceability of tests ?) or whether it is a reworked or not ok part that escaped to the delivery dock.

If you have evidence that it left ok, you should take a look at your production logs for the day it was produced : was there any special event, any deviation to usual parameters.

You need to ensure that the recent change you are talking about was validated in an appropriate way, maybe including your customer approval

Then if you feel more confident on your side, you may go and ask more details to your customer about how the part was handled, and even ask the possibility to audit it.

But in the end you might have to look through the epoxy, with x-rays or removing it

Hope it helps

Best regards

Popeos

 

[somashekar]

Re: Root Cause Determination of Epoxy Encapsolated Electronics Modules

The device has to pass four rounds of functional testing prior to shipping, so the only thing that I'm sure of is that the unit was functioning properly when it was shipped.

Did you get this complaint on a newly supplied device, which perhaps failed within its first few hours of working ?
Question: Do you perform soak or burn-in on load for certain hours after which you do your four rounds of functional testing ?
This could be a cause of infant mortality of the epoxy coated module (or a failure of a component within the epoxy)
Have you checked with your epoxy coated module manufacturer, if he has any x-ray analysis facility and if he can extend investigation on the failed module ?

 

[Bev D]

My first caveat woudl be that without more specifics we can only ask generla questions and provide several approaches that might - or might not - work for you.


Some questions that might help:
Are you sure that the electrical portion of hte device is at fault? How do you know that?
Why does the epoxy coating make failure analysis impossible?
How long was the unit in service before the failure?
Is there anything unique about the failure that can narrow down what part of the device's function failed?

 

[Popeos]

"The device has to pass four rounds of functional testing prior to shipping"

It has to take or it has to pass ?

and has it passed ? Do you have evidence of it ?

I have seen many pokayoke "bypassed" so your sentence is not precise enough without evidence

 

[squonkdb]

Thank you for the replies!

To answer some of the questions that were asked:

I'm not familiar with the x-ray analysis, but this may not be an option at this point given the low failure rate and relative low cost of the unit.

The unit has passed all functional tests and records of the tests are available.

A burn-in test was not performed, but also not required. We are currently performing the burn-in on all units we currently have in stock.

The failure itself is an electrical failure with regards to how the unit functions, but it is possible the failure was caused by something other than an electrical component failure.

There is nothing unique about the type of failure in that it could have several causes.

Right now, we are trying to repeat the failure in the burn-in test. We will attempt to get additional failure information from the customer, but this may be easier said than done since there are several layers between us and the actual user of the product.

 

[Mikishots]

Sorry if a similar question has been asked and answered...

We have received an electronic device that we manufacture back from a customer stating that the device did not work properly. We confirmed their non-conformance. The electronic portion of the device is coated in an epoxy compound which makes analysis to determine if a component failed impossible.

The device has to pass four rounds of functional testing prior to shipping, so the only thing that I'm sure of is that the unit was functioning properly when it was shipped.

How do I determine root cause and corrective action in this situation? The cause of the non-conformance could be anything from a failed component in the device to a customer handling problem.

There is not a history of similar failures of this type for this product, however we did change the mechanical side of the design recently.

Thank you in advance for your input.

I have encountered this problem before, and I can tell you what it was (understanding that it is for consideration and not presented as the failure mode).

Check the CTE of the epoxy material and the environment that the device may have been subjected to after it left your location.

 

[squonkdb]

Mikishots-Thanks for the reply.

We're trying to nail down the way the part was handled and tested in the field as well as the operating environment.

We've made this part for decades and don't have any record of a part returned from the field for an electronic failure for the last ten years (extent of our return records). While I'm willing to explore any possibility at this point, I think thermal expansion of the epoxy may be a long shot given this part's history. I'll probably run a few units through high temp function to be on the safe side. High temp is only required as part of qualification of the part.

One thing I did notice after taking a closer second look at the part, there was a small dent in the case near where an electrical component is attached. Whether the unit was dropped or forced into place when mounting would be helpful information, but getting this info from the customer may be wishful thinking.

I've run about 20 units through the burn-in test at this point and each unit has passed.

 

[Mikishots]

Mikishots-Thanks for the reply.

We're trying to nail down the way the part was handled and tested in the field as well as the operating environment.

We've made this part for decades and don't have any record of a part returned from the field for an electronic failure for the last ten years (extent of our return records). While I'm willing to explore any possibility at this point, I think thermal expansion of the epoxy may be a long shot given this part's history. I'll probably run a few units through high temp function to be on the safe side. High temp is only required as part of qualification of the part.

One thing I did notice after taking a closer second look at the part, there was a small dent in the case near where an electrical component is attached. Whether the unit was dropped or forced into place when mounting would be helpful information, but getting this info from the customer may be wishful thinking.

I've run about 20 units through the burn-in test at this point and each unit has passed.

Agreed, but our situation appeared to follow this reasoning at first - until we discovered that the formulation of the epoxy had changed slightly. The manufacturer had no requirement to notify their users.