EEVblog Electronics Community Forum
General => General Technical Chat => Topic started by: jmsc_02 on July 07, 2019, 11:50:39 pm
-
Hello all,
I have questions about the board reliability after a rework/repair is ended, so i will show my thoughts about this processes and after that i would like to ask some questions.
Here it is the scenario:
I have had jobs where I repaired boards catalogued as industrial, so the repair operation usually worth it mainly to the board cost/availability.
In this kind of jobs I was not too much concerned about the stress created in the repaired zone. I only followed the trainings of my employeers about how to do a specific job.
Time later, I started to document by myself getting information from different organizations like IPC and read books of assembly processes. As a note, in my researching I discovered this place.
And my general opinion after read lot of related stuff, see a lot of rework/repair videos and work with several "handsolder" colleagues is: seems like almost everybody has not care about the thermal/chemical stresses that happens in a board and neighbourhood components when the solder iron touchs the pcb.
Example: usually i read from components datasheets stuffs like "to preserve the integrity and performance of X component you must follow a determined heat curve and you must not surpass X temperature during Y time" but seeing different videos where repairs are performed I usually see how the time emplyed to solder a component is surpassed several times the manufacturer recommendations and performed in a way that (i think) is not recommended by the manufacturer (hand soldering for small IC smt components where you need to apply several peaks of temperature to solder the IC to the pads instead of one heat peak soldering the entire IC in a controlled environment like an oven)
So after this long story short, my general opinion about this kind of process is they are necessary if you have not any other option. In other words: Everybody must avoid this kind of processes like people avoids a surgical intervention if they have other options available.
But now I have a new variable in this subject: preventive maintenance for long term boards: Where I work there is a scheduled plan for every board that basically consists in the capacitor/battery replacement after several years of operation in the field independently of the component status, and the main reason to perform this task is economical: A equipment failure in the field has more associated costs than a preventive repair. (The secondary reason is to protect the equipment against leaky batteries/caps). Also the equipments are so expensive that a replacement is annon-sense at first instance, but this could be an answer to a bad cost planification.
Initially I saw the preventive maintenance as a clever idea but I also see is against the "if the system work then do not touch" principle so...
Is the preventive maintenance on electronics a non-sense?
Is the thermal stress a myth to the daily job? (I believe not)
how can i measure the reliability of a repaired board where i put a specific place at 300 celsius for 20 seconds? How can I measure the new reliability of the reworked/repaired tracks/neighbourhood components after exposing it to fluxes and high temperatures during undetermined time? (I know that there is no clear answer about this questions)
As a final note: it is Known that OEMs refuse to repair their electronics under warranty period and they only do replacements. is this decission taken only by economical reasons (man hour vs new board costs)? My thought about this is there are another factors like pcba reliability after a repair.
So what are your thoughts/experiences about this kind of process? Have you a report/paper/research where the rework/repair process evolves to a performance downgrade? Is a repaired board like a new if you follow the play rules?
Note: My official documentation about this processes are IPC standards like 7721/7711/610 and a couple of ieee papers, mainly https://ieeexplore.ieee.org/document/5898628
Note2: I love repair stuff.
Regards
Edit rev history:
1 Erratas
2 Added ieee reference https://ieeexplore.ieee.org/document/5898628
3 Added general question: "is a repaired board like a new if you follow the play rules?"
4 errata
5 add surgical intervention comparative
-
Manufacturers of consumer goods have almost nothing to gain by having random techs out in the field doing component level troubleshooting and replacement. Swapping entire assemblies means that any old idiot who knows to hold the plastic end of the screwdriver can accomplish the repair. I think this is almost 100% driven by skills and certainty of outcome.
With regard to your situation, I wouldn't worry about the reliability being degraded by replacement of through hole electrolytics by any reasonably competent technician, especially if the board was designed for such service (with generous pads and clearances).
-
That's a good answer for OEMs with consumer grade products, but why an OEM prefers a replacement on a $10000 equipment instead of repair and refurbish I think is more than an economical reason.
You introduced an interesting variable to the equation: the skills of the technician which manipulates the board. I try to avoid this variable by assuming that my hypothetical technician has enough ability to repair the board as perfect as the process can be executed so this way we can center us in the effects of the heat&chemicals involved in the process.
Edit rev history:
1 reformulate answer.
2 Added second paragraph
3 fix erratas and add rev history
-
Most industrial customers are probably more sensitive to failure rates and unplanned downtime than they are to a sub-assembly replacement cost. If I think about our largest factory, if you take it entirely offline for the day, we're going to lose a handful of millions of dollars of shipped revenue that day. Some of it we can make up tomorrow or over a weekend, and it's hard to take us entirely offline with one issue, but in light of that, I don't really care if your whiz-bang tech can charge me $500 in labor and $10 in parts to get me going again or $500 in labor and $10K in parts to get me going again, so long as he gets me going again and I don't have a substantial risk of recurrence right away.
For crops or other items where you might not be able to recover by just working harder tomorrow, it's maybe even more sensitive to time and certainty and even less to price.