Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

How to manually solder a QFN chip with a hot air gun?

  1. Aug 12, 2009 #1
    Hi there,

    I have a tiny 48-QFN chip that I have to solder manually. The pin width of the chip is .28mm and the clearance of the pins is .22mm. The pins are totally under the chip and from side I can only see little metal indicating the positions of the pins, which makes it impossible to solder with iron... There is also a huge ground pad under the chip.

    I set the air gun's temperature to 750F and blew the chip on the board for almost 2 min... Nothing happened. I guess the solder didn't even melt, probably due to the ground pad, or the ground plane... What can I do with it?

    Btw, where can I send the chips and the boards for soldering if I can't make it...?

    Thanks.
    -fiedel
     
  2. jcsd
  3. Aug 13, 2009 #2

    berkeman

    User Avatar

    Staff: Mentor

    Welcome to the PF. As with reflow soldering of SMT parts like this QFN IC, you need to put down solder paste on the pads, and also several paste dots on the ground pad on the PCB. Then place the QFN IC onto the paste, and heat with a hot air gun with the appropriate 4-sided air funnel attachment. The QFN should suck down into position as the solder paste melts.

    Reworking QFNs is not easy -- it takes our best rework person to do them in our lab. We can send larger rework/soldering jobs out to local assembly houses here in Silicon Valley. Where are you located physically?

    I can ask our senior rework person tomorrow for more details about how she does it, if you like.
     
  4. Aug 13, 2009 #3
    Thank you berkeman! Unfortunately I'm in New York and I haven't heard of any local assembly houses around. I quoted from some companies from their web sites, approx. $20 each for less than 100 pieces. Is it that expensive?

    The datasheet says the safe temperature for reflow soldering is less than 260 °C for about 10s. I don't know what temperature and time is good for reworking with a hot air gun, especially with the large ground pad under the device... Another big problem is that the chip is so tiny that a slightest touch could move it... I really want advices from your rework expert. :)
     
  5. Aug 13, 2009 #4

    berkeman

    User Avatar

    Staff: Mentor

    I will talk with her in the morning to see what advice I can post.

    What hot air rework station equipment are you using? What funnel are you using? Which solder paste are you using? What is the distribution of your solder paste dots on the ground pad? What temperature are you setting your hot air gun at?

    $20 each at 100 pieces seems in the ballpark to me, depending on your turnaround time. I can reality-check that with my rework person's manager tomorrow as well, although the price probably varies for Silicon Valley versus New York. It might be cheaper by a significant amount to ship the job out here.

    BTW, the latest chip that I've worked on is in a QFN-48 package -- any chance it's one of ours?
     
  6. Aug 13, 2009 #5
    The chip is TI CC2400 - is it this one? I'm soldering with WHA900 hot air gun from Weller. The nozzle is NQ40 (26.0mm x 26.0mm). It's much bigger than the chip, but I only have this one at hand... The solder is Kester 275 which is lead free with melting point 422 ~ 428°F. I distributed solder evenly on the ground pad. For the hot air gun I tried 450°F, 550°F, 650°F and 750°F seperately.
     
  7. Aug 13, 2009 #6

    negitron

    User Avatar
    Science Advisor

  8. Aug 13, 2009 #7

    berkeman

    User Avatar

    Staff: Mentor

  9. Aug 13, 2009 #8
    Thank you negitron. Swipe method is great for QFP chips, but I'm not sure if it works for this QFN chip because it dones't have long pads that can be reached by iron tips.
     
  10. Aug 13, 2009 #9

    negitron

    User Avatar
    Science Advisor

    They do IF the PCB manufacturer used the recommended footprint, which extends the pad out past the device body just a hair--enough to get a scope probe or a soldering tip onto it. If that's not possible, what I do is presolder the PCB pads and apply a dot of flux to each pad on the device and use a fine-tip air pencil to heat the joints in small groups of one to three pins.

    If you must use your heat gun, fold a square dam from sheet metal (any kind will do) so it's about 1/8" larger than the device and about 1/2" high and use solder paste as berkeman outlines. The dam will keep hot air from lifting nearby components and will also keep the heat in so you don't have to heat the device as much as if it were in the open.
     
  11. Aug 13, 2009 #10

    negitron

    User Avatar
    Science Advisor

    It's available on YouTube, as well, but I can't get to it from work since video sites are blocked due to bandwidth issues.
     
  12. Aug 13, 2009 #11

    berkeman

    User Avatar

    Staff: Mentor

    Okay, I spoke with our rework specialist. As negitron says, the PCB pads probably extend out past the QFN IC package slightly (they do for our own chip and its recommended PCB footprint). So with a fine-tip soldering iron (we use Metcal irons exclusively) and fine SMT solder (Kester 24-6337-0001, 0.01" diameter), you are able to solder the QFN pins to the pads.

    But you also have to solder the center ground pad, and you have to be careful to keep the pins in alignment while soldering the center pad. Our specialist says that she first puts down a few small solder dots on the center pad, but is careful to keep them low-profile so that the QFN can sit on top of them without being held up off of the PCB pads very much. She sets the QFN down in position, and carefully solders a pin per side to hold the IC in place. Then she uses hot air on the part, while holding it down and in place with tweezers, to reflow the center pad. The IC will squat down when the center pad (and the 4 side pins) reflows.

    Then she goes back with the fine solder and iron, and solders each one of the other pins. All of the pin soldering is done under a binocular microscope.

    Hope that helps. It definitely takes some practice, and steady hands. Skip the coffee!
     
  13. Aug 13, 2009 #12

    Integral

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Berke did not mention the binocular microscope that is nearly necessary for hand soldering and rework at that scale.
     
  14. Aug 13, 2009 #13

    negitron

    User Avatar
    Science Advisor

    If you have trouble with it after all the advice given here I'll extend an offer to mount the chip for you, provided you pay shipping both ways. PM or email me.
     
  15. Aug 13, 2009 #14
    - OR -

    You can do that unholiest of things - use leaded solder. I'm not as scrupulous as some when it comes to rework, and the single best aid I've found is to chunk the RoHS solder.

    Apply a bit of flux on a stretch of solder wick and use it to wick away the unleaded solder. Then, apply tin lead solder to the leads and wick that away (to be sure of getting the contaminates from the unleaded out).

    Now, I put down some fresh (keep it refrigerated) tin-lead solder paste. Just on the heat spreader and the corner pins. Less is more, because you don't want the chip floating on it. You just want it to suck up to the board and align the chip.

    Carefully align the chip and heat it. You may be surprised at the difference that only 50C makes! Now comes the not so fun part - applying a tad of solder to each leg with a nice bright solder tip. Flux is a definite help here.

    This works for me for 1-3 units. With 100, you may start to have second thoughts after a few.

    Best of Luck,

    Mike
     
  16. Aug 13, 2009 #15
    Thank you for such detailed instruction! I'd go and get some fine-things and give a try. I can see the pads extend out very very slightly, and that's how I can allign the chip. Soldering the pins with iron will be very difficult, I guess so, because the chip is so sensitive to a non-vertical force. Anyway, hope it works to me.
     
  17. Aug 13, 2009 #16
    Wow, I have a 2x magnifier...
     
  18. Aug 13, 2009 #17
    Thank you negitron for your kind offer. I will try the methods. I really hope I don't have to bother you.:smile:
     
  19. Aug 13, 2009 #18
    This is very attractive idea. I'll definitely try it. Thank you Mike!
     
  20. Aug 14, 2009 #19

    berkeman

    User Avatar

    Staff: Mentor

    I agree. That's a very nice and generous offer, negitron. Thanks. Another awesome example of how the PF can work so well for all of us.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: How to manually solder a QFN chip with a hot air gun?
Loading...