Always one root cause?

[Marc]

Back in the 1980's I was working with complex aviation electronics. I can not find the old file but it shows a component failure that we were seeing consistently. It took a lot of testing and IR temperature 'photos' before we identified the actual root cause. An upstream component was causing intermittent voltage spikes which killed a downstream component. All these years later I can not remember why that component was failing, but I remember we did find the cause. Complex systems are problematic to identify a definite root cause in.

 

[kbg1000]

Why is working in two shifts a root cause? Why is low light over inspection a possible root cause? Brighter light would only improve detection and not eliminate the problem.

Bad inspections do not cause rejects unless the inspection process introduces defects. Building product incorrectly cause defects.

Hi Al, actually, if the product being inspected is easily damaged, "bad inspection" can be a root cause of defects. Printed circuit board assemblies made with surface mount components can be damaged by improper handling during inspection, causing new defects that weren't there previously. Components can be cracked or even knocked off the board.

 

[Al Rosen]

https://elsmar.com/elsmarqualityforum/attachments/ch4-identifying-causes-cochran-pdf.10188/

Printed circuit board assemblies made with surface mount components can be damaged by improper handling during inspection, causing new defects that weren't there previously. Components can be cracked or even knocked off the board.

I mentioned introducing defects during inspection. Improper handling can occur during anywhere in the process.

 

[outdoorsNW]

I second that for PCBAs, inspection can be a high risk area for handling damage. The handling damage risk is higher than most other PCBA processes.

During inspection, the board needs to be rotated and tilted a lot. We have a good AOI machine with multiple angled cameras (9 total cameras), but it can not see everything, requiring the use of microscopes to see what the machine can't. Microscope inspection may require tilting the board close to 90 degrees to check if there is a lifted pad, resulting in the board hitting the microscope if the inspector is not careful and aware of where the board is as they look through the microscope. You almost need three eyes, two for the microscope and one to keep an eye on how close the board is to the microscope.

Minor rework of solder joints may occur as part of the inspection process.

 

[Bev D]

some visual inspection certainly can damage parts as can some types of testing and measurement. Unfortunately we don't yet know what qualprod's "inspection" actually is or what is being "inspected". Even if it damage due to "inspector" handling, their problem statement is ambiguous and the 'causes' that are listed are highly suspect. qualprod has posted no data, no evidence, no explanation of how they proved that these ambiguous causes (two shifts and competency) are in fact 'root causes'...

 

[qualprod]

Why is working in two shifts a root cause? Why is low light over inspection a possible root cause? Brighter light would only improve detection and not eliminate the problem.

Bad inspections do not cause rejects unless the inspection process introduces defects. Building product incorrectly cause defects.

Thanks Al

It was determined that inspector worked too hard and to often was tired because was doubling shift.
Other point that with low light it was not possible to do good revisions on the product.
Because the people in manufacturing produce with certain visual criteria (is a print shop) and who approves the production
is the inspector by using a colorimeter and a pantone and is the only who can approve the product.
Hope is clear

 

[AMIT BALLAL]

Of course, there can be 1 or multiple root cause depending upon the problem. But I don't think fishbone is the right tool to do root cause analysis. It does help in identifying primary / secondary causes (after validation). But further why-why analysis needs to be done to identify the root cause.

 

[qualprod]

Of course, there can be 1 or multiple root cause depending upon the problem. But I don't think fishbone is the right tool to do root cause analysis. It does help in identifying primary / secondary causes (after validation). But further why-why analysis needs to be done to identify the root cause.

You say a combination?
First , fishbone, after identification of main bones, go further with 5 why?
Thanks
PD, I don't think it's enough with just 5 why.

 

[AMIT BALLAL]

You say a combination?
First , fishbone, after identification of main bones, go further with 5 why?
Thanks
PD, I don't think it's enough with just 5 why.

Yes, do why-why analysis once you find out the causes in fishbone analysis.
Most of the time results can be achieved with maximum 5-whys, in rare cases you might need use more than 5 whys. That's why I prefer using the word why-why instead of 5-why.

 

[Bev D]

Were these defects damage that the inspector created or were they ‘escapes’ that the inspector failed to ‘catch’?

So you have data that proves that the inspector who was ‘tired’ because they worked double shifts was the primary source of the ‘defects’? What was their defect rate compared to other inspectors? Did this inspector’s defect rate go down when they only worked 1 shift?

How did you determine that poor lighting was a secondary cause of defects?


Forget the talk about fishbone diagrams and why-why. These are only guessing and questions that try to lead you in the right direction. It is the data that answers the questions and proves the conclusions or claims of causality. Can you answer the above questions?