While designing an FPGA Based control system for a renowned Industrial Control System Equipment Manufacturer, we came across a "showstopping issue" after the populated board came back. The PCBA was a fairly complex controller using a Cyclone III FPGA from Altera. Due to some reason the layout engineer missed the fact that the FPGA has a slug underneath that has to be connected to ground. Well bygones remain bygones and here we were with the 144 pin TQFP FPGA soldered on the board giving us no chance to sneak underneath to get access to the ground plane.
There you go - one more interesting day at MAPYN !!
Just when we had almost concluded that another turn around was inevitable before we can actually "bring up" the board, one engineer who was looking at the layout came running down the stairs.
"What if we drill our way till we get access to the slug ?" he asked. I looked at him with a look that basically said "What are you smoking this morning ??"
But then some rounds of discussion and some time infront of the tool boxes and the PCBA, started giving me some confidence that maybe this is actually possible. (Looking back , maybe it was me who was smoking something)
Pressure can sometimes make you accept such dangerous suggestions , doesn't it ?
Well so here we were in front of the layout measuring the exact area where we could get access to the slug and marking that area on the board. Some components on the bottom side were kind enough to give us way to this drilling operation. (So the layout guy did something right , didnt he? ... ouch )
Now here was the plan. Slowly start the drilling operation by counting how many layers of the 6 layer board have we crossed (counting copper and substrate both of course) and finally stop when the right number arrives.
It was not actually as hard as I thought it would be. Thanks to the Black and Decker toolkit that we have at MAPYN. After 45-60 minutes of careful drilling and brushing away the filings there was a point when I actually saw the silver color, and the drill bit felt as if it was hitting something harder than what we were drilling against for the past hour.
Yeah !! This was it. We have hit gold... no sorry silver .. no sorry slug (It was actually a pleasure this time that I did not see gold .. because had I seen it that would have meant that I am staring at the bonding leads and that would not have been a pretty sight at all). So much for not being a gold digger !
Anyways. There I was with the slug in front of me .. what next ?
Now the next step was to take some solder paste and carefully apply it to the exposed slug. That was a tricky part because the drilling exercise had exposed the internal planes and the fear was that the solder would short the internal layers and that would be the end of story.
So carefully we applied the paste, took the rework station hot air gun and starting carefully blowing hot air (not literally...) on the newly created well.
As soon as I saw the paste melt I quickly grabbed the small 30 AWG wire that I had in my other hand and inserted the stripped end in the well and removed the air blast.
Success was not a one-shot thing here, but patience and persistence bore fruit and there we were with the other end of the wire being soldered to the ground connection on the board.
Felt good to finally do real soldering (I will no more complain if I have to solder 0402 parts.. at least not after this task).
Byteblaster connected and bingo ... the binary file was in the Flash and board was up. We almost danced as if the project was over .. only to realize after a couple of minutes that it had actually just ... started.