Eltek,
Provide more detailed information of stackup. What is the overall, or after lam thickness
requirement? What style and how many plies of prepreg are you using in-between each core?
ASSUMING the after lam thickness is near .062," then you don't have many options. You will
probably need single ply 2113, or 2313 (possibly 2116) prepreg in between each core (and .0039
cap). The amount of warpage will depend a lot on the layer location and surface area & balance of
the circuitry (especially the plane layers). For your .0039" cap core, use 2-sided .0039 3/8 / 3/8 and
pre-bake all of your cores as received in lamination fixtures, under modest pressure (~25 psi) and
320 Deg F (for 2 hours at temp), to relieve stress. There is additional risk here of debris
becoming trapped between panels causing pits/dents and/or epoxy spots. Do everything you can to
keep the panels free of ANY kind of debris.
Etch off copper from 1 side of your .0039 core AFTER the pre-bake, and run it through your oxide
process with the other imaged cores. I would recommend laminating no more than 4 panels per
fixture (1 per fixture is best, if this is a very low volume, low frequency design, and you can afford
tying up your press resources that long). Use heat lagging material between panels and cauls. Use
enough to slow the heat rise way down to 3 - 4 Deg F per min through Tg. Find a way to slow the
cool down rate to 3 - 4 Deg F / min through Tg also. (We use an inline water heater to heat up the
platens of our Cool Down Press for this purpose). Remember to use a compatible separator plate
between panel and heat lagging material to avoid texture transfer to the copper surface. Even with all
this effort to minimize warp, you will probably encounter some anyway due to the design of the
circuitry relative to copper area, x-y distribution, x-y balance, and z-axis location/balance, and the
different cte properties of the .0039 core compared to the .016 cores.
I would do everything possible to convince the design group to re-think their odd-layer count. There
are many construction options, such as duplicating a layer to get 7 functional, 8 real for balancing.
Keeping all the cores of the same thickness, and using prepreg to adjust the core to core dielectric
spacing will minimize warp also. Another good practice is to limit the prepreg to one type (e.g.
2116) and make the number of plies per bond section symmetric about the z-axis center line.
I hope this helps. Let me know the results, if you follow any of these recommendations. Good luck!
Fred J. Paul, Staff Engineer, PCB Operations
FLUKE CORPORATION
P.O. 9090, MS 55, Everett, WA, 98206-9090
direct: 425 356 5734 fax: 425 356 6070 email: [log in to unmask]
>>> VISIT OUR WEB SITE AT WWW.FLUKE.COM <<<
At 13:00 24/03/98 +0000, you wrote:
>We currently have a board requirement for a 7 layer board.
>
>The design limits the modification to make it a balanced
>even numbered layup and to rearrange thicknesses.
>
>The cores are
>
>single sided cap with 3/8 oz. CU.
>.0039 core
>
>followed by
>.016 core 3/8 / 3/8 CU
>.016 core 3/8 / 3/8 CU
>.016 core 3/8 / 3/8 CU
>
>Question: Does anybody have any ideas to help reduce the
>warpage that will occur on this type of layup?
>
>ROR, Lagging, Cool Rates, etc.....
>
> thanks
>
################################################################
TechNet E-Mail Forum provided as a free service by IPC using LISTSERV 1.8c
################################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with following text in the body:
To subscribe: SUBSCRIBE TechNet <your full name>
To unsubscribe: SIGNOFF TechNet
################################################################
Please visit IPC web site (http://jefry.ipc.org/forum.htm) for additional information.
For the technical support contact Dmitriy Sklyar at [log in to unmask] or 847-509-9700 ext.311
################################################################
|
