Printed Circuit Board Assembly & PCB Design SMT Electronics Assembly Manufacturing Forum

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Views: 10122

Accuracy - Aug 26, 2011 by Otis  

Accuracy - Aug 26, 2011 by jdengler  

Accuracy - Aug 26, 2011 by mac5  

Accuracy - Aug 27, 2011 by KaHrpr  

Accuracy - Aug 29, 2011 by bobpan  

Accuracy - Aug 29, 2011 by robgd3  

Accuracy - Aug 29, 2011 by Patrick Bruneel  


Accuracy | 26 August, 2011

I'm looking for some guidance here and would appreciate input from some seasoned professionals. I am the Op. Mgr. for a medium sized electronics contract manufacture, but my background is almost exclusively in the machine tool industry and I would like to get an "outside" opinion on an issue that we are having. We have a two JUKI SMT line. Both machines are well maintained and in good overall condition. We are a job shop, so, we get a fair number of new jobs to set up. My programmer does the programs manually (using customer supplied X - Y location data, when available)then insists that it is necessary to step through each location visually on the machine, using a bare board, to verify the locations. This can be a very time consuming process, especially when you have six or seven hundred individual placements on a board. My question is - is this standard in the industry? Does everyone do this? I appreciate your input..........

reply »


Accuracy | 26 August, 2011

If he is typing in numbers I would say this is necessary. If you are importing these numbers and then generating the placement file this is overkill. You do need to do a first article inspection/test on the first board to verify the program.


reply »


Accuracy | 26 August, 2011

On any new project we use double sided tape and place all the components to check for placement accuracy and polarity prior to pasting a board.

We salvage any large/high dollar parts and toss the passives.

reply »


Accuracy | 27 August, 2011

Basically it pretty standard to run through the board. For programmers /manufacturing engineers it’s a catch 22. If you check everything management is complaining it takes too long. If you don’t check management is complaining because of rework and you did not check. Hence the catch 22 you’re damned if you do, you’re damned if you don’t. That’s our world.

reply »


Accuracy | 29 August, 2011

Like jdengler says.....if he is manually teaching the placements and fiducials....he has to run through all the placements......if you are using cad data and fiducials to generate the program....then should check that the first few placements are on....then run the machine.

Good luck

reply »


Accuracy | 29 August, 2011

We program off-line, and still step through the first board...mostly to verify polarities. As has been mentioned above, the choices are check them before they build, or fix them after the build...which will include the ME time at the machine, plus the rework time on the bench, plus the QC time to catch the error.

I'm not sure how the Juki's work for stepping through, but, I wouldn't expect it to take that much time to step through a board...unless they're verifying p/n's as well as placement and polarity. One thing you could have them do while stepping through is actually build a the same time. I assume the machine can do a single placement while stepping through the parts; have them place the part while verifying. It will take some time, but, at the end of that time, you'll have a first piece built, with a pretty high confidence in the quality of the board.

Cheers, ..rob

reply »


Accuracy | 29 August, 2011

Excellent advice!!!!!!

reply »

NexJet System - Flip Chip Underfill

SMT Nozzles and Consumables