• AVR Freaks

Hot!PIC32MM0XXXGPM064 Incorrect Footprint In Datasheet?

Author
acharnley
Super Member
  • Total Posts : 389
  • Reward points : 0
  • Joined: 2016/05/01 06:51:28
  • Location: 0
  • Status: offline
2019/11/13 05:01:50 (permalink)
0

PIC32MM0XXXGPM064 Incorrect Footprint In Datasheet?

HI All,

I created a prototype using the footprint measurements in the datasheet. The chips arrived a bit later and I've been attempting to reflow them onto the board. I'm usually rather good at it (if I say so myself) but I've not yet managed to make a good connection to allow ICSP to work (or there's a problem elsewhere but I think it's the reflow).

Anyhow you can see that the chip has a significantly smaller pad area then the datasheet and I'm wondering if it is causing the issue. A few times I've had shorts as the gap between pins and pad is minimal so alignment has to be ultra spot on. Have Microchip made an error?

Regards, Andrew

Attached Image(s)

#1

5 Replies Related Threads

    deelbug
    New Member
    • Total Posts : 5
    • Reward points : 0
    • Joined: 2019/04/19 06:42:57
    • Location: 0
    • Status: offline
    Re: PIC32MM0XXXGPM064 Incorrect Footprint In Datasheet? 2019/11/13 05:39:56 (permalink)
    5 (1)
    Looks like a footprint for the TQFP package whereas your chip is the QFN package. The footprints are different
    #2
    Howard Long
    Super Member
    • Total Posts : 726
    • Reward points : 0
    • Joined: 2005/04/04 08:50:32
    • Status: offline
    Re: PIC32MM0XXXGPM064 Incorrect Footprint In Datasheet? 2019/11/13 06:30:43 (permalink)
    0
    Are there any thermal vias on the thermal pad? They have a small wicking effect thus allowing you a little leeway on how much thermal pad paste you use. Without them, the solder's got nowhere to go but the pin pads.
     
    What is your reflow method? Are you using a paste stencil and an oven, or wetted pads and hot air for example?
     
    For development, you could just not apply paste to (or avoid wetting) the thermal pad as it's already tied to Vss.
     
    FWIW, for reworking these, I use hot air and wet the pads, I don't use paste or stencils. How much solder you use to wet the thermal pad is probably an experience thing, but the short answer is not too much and not too little.
     
    When applying hot air, the device should be gently nudged with tweezers under the microscope in a way that you can see it returning itself to the correct position without you forcing it. Then, drag solder the exposed castellations of the QFN with a 0.2mm bit.
     
    post edited by Howard Long - 2019/11/13 06:32:09
    #3
    mbrowning
    USNA79
    • Total Posts : 1564
    • Reward points : 0
    • Joined: 2005/03/16 14:32:56
    • Location: Melbourne, FL
    • Status: online
    Re: PIC32MM0XXXGPM064 Incorrect Footprint In Datasheet? 2019/11/13 08:28:28 (permalink)
    0
    The datasheet pg 342 says "64-Lead Plastic Quad Flat, No Lead Package (MR) – 9x9x0.9 mm Body [QFN]With 0.40 mm Contact Length and 7.70x7.70mm Exposed Pad". The recommended land pattern says Max 7.5mm for the "optional" center pad. The product ID figure shows the 64QFN to have the designation "MR", and there's a note!
     
    For the most current package drawings, please see the Microchip Packaging Specification located at http://www.microchip.com/packaging
     
    In the Packaging Spec the MR package says "64-Lead Very Thin Plastic Quad Flat, No Lead Package (MR) – 9x9x0.9 mm Body [VQFN]With 7.15 x 7.15 Exposed Pad [Also called QFN]"
    and the EP size is 7.15mm. with recommended "optional" land pattern max 7.25mm.
     
    Eyeballing your picture, I would guess the EP is closer to 5.5mm square, so there's definitely a discrepancy.
     
     
     
     

    Go Navy! Beat Army!
    #4
    acharnley
    Super Member
    • Total Posts : 389
    • Reward points : 0
    • Joined: 2016/05/01 06:51:28
    • Location: 0
    • Status: offline
    Re: PIC32MM0XXXGPM064 Incorrect Footprint In Datasheet? 2019/11/20 01:15:08 (permalink)
    0
    I've been on to Microchip, so far it seems the datasheet is right and the batch of chips is wrong.

    I reflowed about 8 of these, they are quite painful to do, but perfectly done as per microscope checking I still couldn't get them to work. I always ended up with a consistent 3.4ohm short.
     
    After exhaustive checking of the prototype board for errors I've concluded not only should the pad be smaller but it is not connected to GND (odd I know!). There is no mention in the datasheet about the pad being connected to GND but I made the incorrect assumption.
    post edited by acharnley - 2019/11/20 01:26:07
    #5
    acharnley
    Super Member
    • Total Posts : 389
    • Reward points : 0
    • Joined: 2016/05/01 06:51:28
    • Location: 0
    • Status: offline
    Re: PIC32MM0XXXGPM064 Incorrect Footprint In Datasheet? 2019/11/20 03:16:21 (permalink)
    0
    Howard Long
    Are there any thermal vias on the thermal pad? They have a small wicking effect thus allowing you a little leeway on how much thermal pad paste you use. Without them, the solder's got nowhere to go but the pin pads.
     
    What is your reflow method? Are you using a paste stencil and an oven, or wetted pads and hot air for example?
     
    For development, you could just not apply paste to (or avoid wetting) the thermal pad as it's already tied to Vss.
     
    FWIW, for reworking these, I use hot air and wet the pads, I don't use paste or stencils. How much solder you use to wet the thermal pad is probably an experience thing, but the short answer is not too much and not too little.
     
    When applying hot air, the device should be gently nudged with tweezers under the microscope in a way that you can see it returning itself to the correct position without you forcing it. Then, drag solder the exposed castellations of the QFN with a 0.2mm bit.
     



    I use an identical reflow process (heated bed + air) it seems, largely found by trial and error. You also assumed the pad is tied to Vss but this doesn't appear to be the case with this particular model!

    I've ran off another board, an annoying three week wait for it but I'm well used to it in prototype design!
     
     

    Attached Image(s)

    #6
    Jump to:
    © 2019 APG vNext Commercial Version 4.5