Forum Discussion

Altera_Forum's avatar
Altera_Forum
Icon for Honored Contributor rankHonored Contributor
11 years ago

Cyclone III 144 QFP package forgot the ground pad on circuit board

Our circuit board guy forgot to put a ground pad on the board for the 144 pin package on the bottom of the chip. We are unable to detect it thru the JTAG port and assuming that is why. What may also be a problem is there are untented vias and traces (soldermasked though) underneath the pad of the chip. I wanted to find out if anyone else has made the same mistake and what they were able to do to repair it? Drill hole and run a wire? Pull the chip, run some copper trace tape underneath and reattach it? Any suggestions? or is it a lost cause? I think the fact that there are uncovered vias possibly shorting to the ground pad on the chip makes it total loss.

*************UPDATE NOW FIXED!!!!!!!!!!!!!!!!!!!!******** See attached photo and diagram of leadframe.

I gave up on drilling a hole underneath thru the circuitboard to access that ground pad, I ended up drilling too deep (into the chip) and plus it’s impossible to get solder into it on a wire and it would short against the middle layers in the hole that would be exposed.

So anyway I pulled the chip and measured with an ohm meter that that bottom ground pad is not the same as the ground pins, but does connect to the Tiny tiny little bit of lead frame at that is barely visible at each corner of the epoxy package when they mold it. (See diagram). So I was able to solder a wire on a new board to the corner of the FPGA and run it to ground. Its extremely fragile because there's barely enough to solder to.

I powered it up and bang all of a sudden I can connect with quartus programmer, it now detects the chip via JTAG!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

So I guess this proves that that pad has to be grounded, but it’s not really a permanent or even temporary solution, it’s too fragile. Maybe if I drill into the epoxy package to expose more of the corner of the leadframe, I will have a better connection that will make our prototypes more durable. I will try that.... Update.... Drilling with a tiny drillbit to expose more of the leadframe worked great! see this link for details.... http://www.alteraforum.com/forum/showthread.php?t=46791

http://www.alteraforum.com/forum/attachment.php?attachmentid=9654&stc=1

6 Replies

  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    11 years ago

    Yes, there are several reports at Altera forum about successful repair of prototype boards with missing exposed pad connection. Of course it depends on the design details.

  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    11 years ago

    Well it turns out I was able to probe the pad thru a via with a wire and it is already grounded! Perhaps that pad is grounded thru a ground pin or on the internal leadframe or something, maybe Altera wants it grounded for noise or power issues. I'm thinking that's not the reason it can't be detected via JTAG. I put in a support request to see if they know if that pad isn't grounded USB Blaster wont detect the chip on JTAG. I doubt I will get a meaningful response, the tech support people seem to be in india and just quote documentation, no real world experience with customers or contacts with people at altera to know this.

  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    11 years ago

    Shouldn't be very difficult to drill a hole and solder a wire to the pad, just make sure you don't drill into the chip :) Altera says you must ground the pad - from the CIII Handbook:

    --- Quote Start ---

    The E144 package has an exposed pad at the bottom of the package. This exposed pad is a ground pad that must be connected to the ground plane on your PCB. Use this exposed pad for electrical connectivity and not for thermal purposes.

    --- Quote End ---

    Also check the output of the 1.2V and 2.5V power regulators, just to make sure everything else is ok.
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    11 years ago

    --- Quote Start ---

    Well it turns out I was able to probe the pad thru a via with a wire and it is already grounded! Perhaps that pad is grounded thru a ground pin or on the internal leadframe or something, maybe Altera wants it grounded for noise or power issues. I'm thinking that's not the reason it can't be detected via JTAG. I put in a support request to see if they know if that pad isn't grounded USB Blaster wont detect the chip on JTAG. I doubt I will get a meaningful response, the tech support people seem to be in india and just quote documentation, no real world experience with customers or contacts with people at altera to know this.

    --- Quote End ---

    According Altera specifications, it's clear that the exposed pad must be connected. There are also various user reports about complete function failure (e.g. no JTAG operation) with unconnected exposed pad. Of course there may be additional issues in your design, but it clearly looks like the missing ground connection is a sufficient failure condition.

    In so far your support request seems to track mainly the academic problem why the function fails, although there's still some kind of ground connection. Did you also measure the resistance respectively voltage drop between both grounds, by the way? Quickly fixing the layout problem would be more target-aimed, I think.