Connect with us

0.031" Thick FR4 PCB

Discussion in 'Electronic Design' started by Darol Klawetter, Nov 6, 2012.

Scroll to continue with content
  1. I'm considering using a 0.031" thick FR4 PCB for my next design. What has been your experience with 0.031" PCBs? Have you had durability problems? Did it complicate manufacturing? My PCB will be roughly 3" x 3".


    Darol Klawetter
  2. Have done 0.02" (2-layer) with no problems.

    If your 0.031" board has more than 4 layers it might be an issue.
  3. I'm planning on 6 layers. Stackup is as followins

    1 oz Cu
    1 Sheet 106, 1 Sheet 1080
    1 oz Cu
    0.005" Core
    1 oz Cu
    1 Sheet 106, 1 Sheet 1080
    1 oz Cu
    0.005" Core
    1 oz Cu
    1 Sheet 106, 1 Sheet 1080
    1 oz Cu
  4. Jamie

    Jamie Guest

    That's a nice size but one time we got carried away and forgot to
    check the rail specs a extruded aluminum chassis's we had already
    purchased, had to machine them a little to get the boards in.

    If you don't have any problem there then thickness shouldn't a

    We visited a tyco facility once, watched them do some 64 layer
    boards but I guess that isn't the limit, but it sure was interesting to
    say the least.

  5. Guest

    we do 10 layers in ~1.5mm (~0.06")

    board thickness put a limit on via and hole sizes, thicker board and
    the holes need to be bigger

  6. Guest

    Yep. That's the reason we're starting to go to less than 1.5mm (1mm,
    IIRC) boards; can't do the vias on .5mm BGAs. Don't know that we're
    going to actually do any products with .5mm BGAs but we need them for
  7. Guest

    ccould do blind/burried vias, the problem is plating deep a shallow
    hole, vias that only go through part of the stack shouldn't have that

    or maybe via in pad will work

  8. Robert Baer

    Robert Baer Guest

    I have used PCB material down to 14 mils with no problems.
    Then again,what i used was equivalent to Getek (Megtron).
    To 200C no sweat.
  9. Jamie

    Jamie Guest

    We ordered some FR5/G11 recently and it was a task to get it
    here in the states in small quantity. But we did manage to get
    3/16" thick and sure works well for HV plus structural

  10. Tim Williams

    Tim Williams Guest

    Some day I might build a particularly high performance planar transformer,
    using a PCB stack thicker than it is wide.

    A number of manufacturers are offering "heavy copper" these days (> 10
    oz.), pretty cool.

  11. Guest

    We do some of that, but it's expensive and I don't know if the process
    is qualified for production. OTOH, thin boards have their own set of
    No, that doesn't solve the l/w limit on vias.
  12. Guest

    I've seen 1 and 2 mil FR4. It makes a pretty nice flex circuit,
    better than kapton in many ways.
  13. Guest

    it doesn't solve the l/w limit, but it may be possible to route
    with the bigger vias when you have the vias in the pad and don't
    need to fit them between the pads

  14. Boris Mohar

    Boris Mohar Guest

    The only issue I had was with small panelized boards. The panel was scored
    but the gorilla that separated them managed to induce invisible cracks in
    ceramic caps near the edges. On the next run I re scored the boards and
    separated them without flexing.
  15. Guest

    At the PPoE they had problems with traces cracking after panel
    separation. The solution was deeper scoring and larger mouse bites.
    Make sure there is a decent setback for any planes or signal traces,
    as well.
  16. Fred Bartoli

    Fred Bartoli Guest

    a écrit :
    I have a small board (8mm dia) that's top&bottom stuffed with 0402,
    0201, uBGA,... and we also needed some accuracy on the max board diameter.

    Usual separation was... scary, at least, so we ended to laser cut them:
    fast, accurate, no strain. What's more to be desired? Ah, yes, low cost...
  17. Tim Williams

    Tim Williams Guest

    Good practice anyway to place chip components perpendicular to nearby
    sources of strain, in this case, parallel to the board edge. That's what
    the manufacturers recommend. May not be enough though, given the presence
    of said gorillas.

    I like routed/milled boards better anyway: you don't get fingers full of
    itchy fibers from handling them :)

  18. Guest

    It's tough to do with components on both sides of the board, on both
    sides of the score. There are tools to do this but they were too
    cheap (but wasted *tons* in other areas). The problem is that the
    recommended score and mouse bites are marginal, and of course everyone
    crowds the edges (raises hand). "I *know* that ten pounds will fit in
    that 5# bag."
    I would think you'd need a different tool for each board. Several, in
    fact. Once we got the recipe down, there weren't a lot of problems.
    The problems were up-front, when the schedules are the tightest, of
    course (get the thing to market so the boss can afford to pay us).
  19. Guest

    In the ECL mainframe days, 80ohms was the normal trace impedance (some
    used 50 ohms for clocks). The higher impedance reduced power, at the
    cost of speed, of course (which is why the clocks were often 50 ohm).
    When I designed the testers, I usually chose 90 ohms; lower current
    yet and it's close to ribbon cable (~100 ohms).
  20. Guest

    the problem is in the other direction, when you get very
    thin dielectric you can't make a trace that is thin enough
    to get up to 50R

Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day