ESD problem, programmed data error. Needs reprograming and then board works perfect

[nitin1518]

Hello experts!

I am working on a ESD project in company as a in-plant trainee.

Problem:
- Using SU 3000 BGA FBGA64: Chip is programmed, verified and if found OK, its integrated on the pcb board with other components
- 50 out of 10000 boards have the flash error (program verified OK on programming desk but comes out faulty once installed in the PCB)
- Those 50 board chip's when reprogrammed, the board works fine

* The operators are wearing ESD bands and other precaution necessary

Actions taken till now:
- Grounded the production line for extra precaution
- Labeled the operators to identify if the problem is from a specific operator: Problem occurring average is similar to to all the operators
- Consulted with experts and everyone stated it's ESD related: Cannot be removed 100% but surely can be reduced a lot

Can you please suggest the actions that can be taken to solve the problem? Or what else could be the reason?

I know it is a very specif problem and could be resulting because of various reasons. But please suggest what ever you can so that we can work and try them out.

Please feel free to ask any other details if needed.

Thank you,
Regards.
N

 

[nachiket103]

Dear Nitin,
With taking care of ESD externally, also add some ESD protections in ur circuit.
U can add 4.7nF/200V or 47nF/200V 0603/0805 capacitor at the input of Power supply.
Also place it on the Vcc line of BGA chip.
U can also use dedicated ESD protectors from Wurth Electronics. there devices have very good response.
I hope this may solve ur problems.

 

[nitin1518]

We are not the chip manufacturer. So no question of on-ship modification.

I am looking into ESD and other related issue effecting the programming on production line.

- Socket pin contact degradation
- Electrostatic problem

But as the program is been programmed successfully (verified on gang programming board). But once the chip is integrated with the PCB board, later at the final test, it has flash problem. We have to reprogram the chip (onboard) and then it works perfectly.

The faulty boards have program data been altered at some address. So wondering if this is due to the ESD or it has something else effecting the case.

Please let me know if you have any experience/knowledge dealing with the same.

 

[nachiket103]

Dear Nitin,
I am suggesting the modifications in ur circuit.
Basically flash corruption problems occur mainly because of ESD.
As ESD, being a high voltage and if not suppressed to an adequate level will cause flash corruption as flash memory is written by applying higher voltage levels.
I would like to suggest u to pls. go through circuit schematic once again and add necessary ESD protections if its not there.

 

[nitin1518]

Hello Nachiket,

I do understand what you are trying to say. But as we are not the chip manufacturer it's not a practical solution to modify the chip or change the chip.
More, this problem is occurring on 2 different chips programmed during production. So I as the number is not that big, I am considering it to be something related to ESD prevention on production line. If their was major problem in the chip related to ESD the number of rejection would be much larger.

Let me know if you have any suggestion related to this issue other then modification in the CHIP itself.

Thank you,
Regards,
NS