Assembly Followup


renegadechic

Active Member
Joined
Sep 10, 2008
Messages
787
I figured i would start a thread since the update was last posted right in the middle of another thread here http://www.gp32x.de/board/index.php?showt...mp;#entry670836

I wanted to see if MWeston or any of the other devs had any kind of progress update on this. Have the failure rates gone down and are you slowly moving towards larger production batches yet? I found it a pretty useful and interesting post and it was good getting an insight into the manufacturing side of things!
 
Last edited by a moderator:
RenegadeChic said:
I figured i would start a thread since the update was last posted right in the middle of another thread here http://www.gp32x.de/board/index.php?showt...mp;#entry670836

I wanted to see if MWeston or any of the other devs had any kind of progress update on this. Have the failure rates gone down and are you slowly moving towards larger production batches yet? I found it a pretty useful and interesting post and it was good getting an insight into the manufacturing side of things!



This doesn't answer your question but look at the technology you're dealing with and TI's advice on who can and how to deal with it:

http://focus.ti.com.cn/cn/lit/an/spraav1/spraav1.pdf

Basically, the PoP (Package On Package) means you're soldering a chip to the PCB and then another chip to the chip, with 500 connectors at 0.4mm intervals... that's no mean feat and the PDF is full of warnings about how simple, microscopic things can make the connections useless.

I don't know anything about PCB assembly past what I can solder with a 8mm soldering tip, but I found it interesting.
 
Last edited by a moderator:
This Chip is really tricky - just read through the manual and man is it one tricky mother!

With such a tricky PCB design and manufacturing process, you kind of wonder if they should of used the same Contract Electronics supplier (the company which basically gets all the bits and sticks them together) as the BeagleBoard - I suppose that would of brought up other issues with the BeagleBoard versus OpenPandora

But man can you understand what a tricky process it is using the OMAP

I worked running shop floors for various Contract Electronics Suppliers - using Surface Mount processing, I can see why MWeston was spot on about ensuring the process is refined and failure rates reduced as much as possible
 
linked document said:
Design of circuit boards for fine-pitch BGA packages at 0.4mm and smaller is more of an art than a science
What! This is some seriously amazing stuff... Thanks for the link, I love reading about this! B)
 
Last edited by a moderator:
If one bga wasn't enough on a board this one has 2 stacked on top of each other. Does anyone know if they are using 1 ot 2 passes in the reflow oven? on spec sheet it says that you can do both. On the plus side you don't have to run any traces for the address lines, r/wr, etc, to connect an external memory chip.
 
Back
Top