Circuit Board Incoming Inpection Criteria

[jscholen]

Part of our Medical Device includes an electronic unit with multiple custom circuit boards.

Here's my problem: I believe that each board should have an incoming inspection criteria for functional requirement, however, our R/D Operations group says that it is a part of a closed looped system and will be tested as a whole system, so in essence our contract manufacturer just stuffs the board and we test it within a system not as an individual component.

I don't have an electronics background, but I would think that each board would have a functional output requirement that would need to be met prior to introducing into production?!?!

Does anyone have an arguement for or against their justification?

Thanks!

[Ajit Basrur]

In Reply to Parent Post by jscholen

Part of our Medical Device includes an electronic unit with multiple custom circuit boards.

Here's my problem: I believe that each board should have an incoming inspection criteria for functional requirement, however, our R/D Operations group says that it is a part of a closed looped system and will be tested as a whole system, so in essence our contract manufacturer just stuffs the board and we test it within a system not as an individual component.

I don't have an electronics background, but I would think that each board would have a functional output requirement that would need to be met prior to introducing into production?!?!

Does anyone have an arguement for or against their justification?

Thanks!

Sorry, I am not an expert in this area but can give a lead - GUIDE TO INSPECTIONS OF ELECTROMAGNETIC COMPATIBILITY ASPECTS OF MEDICAL DEVICE QUALITY SYSTEMS

[Doug]

In one of my previous professions, I managed quality inspection of printed circuit boards. That was many years ago. However, I know there are inspection criteria (board thickness, plating thickness, circuit continuity, component layout, hole sizes and I'm sure many more). Good luck in your search.

[Bev D]

Functional testing is capital intensive and requires soem heavy software development. If functional testing is to be done it shoudl be done at the board manufacturer. The difficulty with this is that often the functional failures are not due to board workmanship standards but more frequently due to component supplier quality and design marginalities so many board manufacturers dont' want to be - nor should they be - put in this place if they have no input to the design of the board.

For similar reasons I don't believe in IQA functional testing either. Unfortunately, teh response from your R&D guys as you related it leads me to fear that they don't fully understand the functional reliability of the board in relation to the system. This makes IQA functional testing almost totally useless. The focus would best be placed on understanding (any) failures to integrate into the system first. then develop true specification requiremetns for the board function, then hold the appropriate functional organization responsible for meeting those requriements.

[Jennifer Kirley]

The argument is usually one of a business case. Here are the general questions that drive the answer in a business case like this one:

Have the incoming PCBs been reliably conforming to requirements - do they always perform well when installed?

If yes, then supplier control may be all that is needed. If not, then:

What is wrong? (Pareto)
What kind of inspection is needed - do you need the gamut of IPC-A-600?
What is needed to correct the resulting problem assembly? Materials, time
What happens with the delays involved with replacing components that are bad?
Is the cost of all this repair work more, or less than the cost of receipt inspecting the PCBs?
What would a good, robust supplier control program cost so the above activities could be avoided?

[Bob the QE]

I worked in a CB assembly shop in which we had a closed loop system. The easiest way to try and explain it is in sections so I will try:
Incoming PCB's ( 12-16 layers .006 -008 lines)- we used only approved suppliers who were evaluated for process controls to fit a closed loop system. This left many small shops out because we needed tight process controls such as printing, etching, layup and especially testing. Once the boards were built they were 100% electricaly tested for point to point contact. Once they entered our process they were again tested and assigned a serial number with a bar code. Process documents from the supplier were reviewed prior to accepting the lot.
Boards were put into the production process and went through a thru hole assembly, manual inspection and then wave soldered, cleaned, visual inspection manually, and tested for the limited type of performances consistant to the components assembled on it. Each failure was analyzed as to why. There were several classifications of failures and each had its own esculation plan. The next process was surface mount assembly and hot gas soldering. it then went through a optical inspection station and on to testing then to burn-in and then from there to final test. Again each failure was analyzed by a process engineer and failures were catagorized and each one had their own esculation/quality level awareness plan.
Our yields were typically around 92-94% for raw PCB's, 80-85% thru hole 90-93% surface mount and 97% or so for final test and 98-98.5% after burn in.
When I left the org. they were gearing up to implement 6 sigma designing. They were then sold and moved off shore....go figure.

[Jim Shelor]

Many good points have been made here. I am an elcetrical engineer and I agree with the R/D Operations group. A test with the entire system is good and it is cost effective. As long as the closed loop system is tested thoroughly, (i.e., all functions) if there is a failure in your component you will most likely find it during the system test. There is a low probability that a failure you would find in a component test will not show up in a thorough system test. You would likely spend a lot of money doing unnecessary testing.

Sincere regards,

Jim Shelor

[Bob the QE]

I too agree with the R/D guys. Electronic componet quality CAN BE a solid base to build your testing stratagies around. CB manufacturing has also seen a remarkable amount of process improvements. Here is where I may differ a little, these two components are only as good as the manufacturer, the real manufacturer not the one on the label. You must have a strong supplier evaluation team that can help control component quality on a global level. There is (and I know from first hand experience) quite a bit of repackaging, bootlegging and right out fraud going on in these industries outside the US. Countries need not be named but I have been involved in such evaluations and determinations. Once a great closed loop process is infected with lower quality products, the process indicators can start a chain of misleading corrective actions and very costly delayes. I am realy glad I left that industry considering todays manufacturing climate.

Good luck