Complete mechanical drawing for FCBGA1380 (Bay Trail / Atom Z3700)??

[eAgan]

The datasheet (Vol. 1) for Atom Z37xx / Z27xx family includes a package mechanical drawing for FCBGA1380 on page 303.

This package drawing is incomplete, however. No information is given about the pitch of the balls in the die shadow, and I am not convinced one can reliably calculate the pitch of the other balls from what little information is provided.

[Gabriel_T_Intel]

Hello Eric.

Thank you very much for contacting the Intel® Embedded Community.

We are checking this file and will get back to you soon.

Best Regards.

Gabriel

[Josue_C_Intel]

Hello Eric,

Please refer to https://www-ssl.intel.com/content/www/us/en/processors/atom/atom-z36xxx-z37xxx-datasheet-vol-1.html Intel® Atom™ Processor Z3600 and Z3700 Series: Datasheet, section 30.1.3 Package Diagrams, Table 1, symbol H, this is the minimum pitch for balls anywhere pattern.

Hope this is useful to you. Please let me know if you have more questions.

Best Regards.

Josue.

[eAgan]

Hi Josue,

Thank you for your reply. The diagram you mention is the one I am having trouble understanding.

Could you please explain how to calculate the horizontal pitch of the outer and inner balls from this figure?

Maybe I am missing something. The vertical (in the diagram) distance between the rows is not specified. If I assume the outer balls are the same distance apart horizontally as they are vertically (so, they are at a 45 deg angle to each other), then a 0.4mm distance (assuming this is nominal, which it probably is not, since it is marked minimum), would correspond to a distance of 0.4^2 = 2p^2 which becomes 0.4/sqrt(2) = 0.28284271247461900976033774484194... between the balls in adjacent rows, or twice that for balls in the same row...

I would expect the pitch to be a round number, however.

Furthermore, it seems the G1 and G2 figures cannot be correct if the above calculated pitch is correct?

The package size (B1 and B2) are given as 17 +/- 0.05, and G1 and G2 both are 0.502.

If the pitch really is 2*0.4/sqrt(2) and, from the (hard to read...) Figures 44 and 45, we know that the row with the most balls is 29 across, then we have a distance 28 * 2 * 0.4 / sqrt(2) = approx. 15.84mm. Compare this to 17 - 0.05 - 0.502 = 16.448mm. What am I missing?

Finally, how does one calculate the pitch of the innermost balls, in the die shadow area, which are specified as being a different size to the outer balls (for which H is specified)?

Please advise.

[Josue_C_Intel]

Hello Eric,

Can you please provide us the exact model that you are going to use in your design? S-Spec number

I need more specific details about your processor, do you have any other specifications?

Best Regards.

Josue.

[eAgan]

Hi Josue,

We're interested in using the Atom Z3770 (SR1M3).

Let me know if you need any other information.

Thanks,

Eric

[Josue_C_Intel]

Hi Eric

The distances requested in mils (1 mil = 0.0254mm) are:

For the Inner zone

22.67 * 0.0254mm = 0.5758mm

37.99 * 0.0254mm = 0.9649mm

21.25 * 0.0254mm = 0.5398mm

For the outer zone

15.69 * 0.0254 mm = 0.3985mm

22.19 * 0.0254mm = 0.5636mm

22.47 * 0.0254mm = 0.5707mm

Hope this is useful to you. Please let me know if you have more questions.

Best Regards.

Josue.

[eAgan]

Thanks Josue, that is useful.

Would it be possible to get the footprint in Gerber or DXF or Mentor Graphics PADS format?

Or at least a cropped PDF of the top layer of a known-good board using this footprint so I can load that into my CAD tool as an overlay?

The reason I ask is these numbers don't seem like a nice round figure in mm or mils, and I'm concerned that if I step & repeat the pads at these distances it will accumulate rounding error as it goes along.

Thanks very much,

Eric

[Natalie_Z_Intel]

Hi, Eric! Here is another URL that may be helpful. https://www-ssl.intel.com/content/www/us/en/processors/atom/atom-technical-resources.html https://www-ssl.intel.com/content/www/us/en/processors/atom/atom-technical-resources.html Also, you may soon be contacted by an Intel representative to discuss your request for a Privileged EDC account which would allow access to some Intel Confidential content.

[eAgan]

Thank you, however the documents in the given link do not contain a complete package drawing for this part.

[Natalie_Z_Intel · Accepted solution]

Hi, Eric. Document # 509619 has been delivered to you via email it has not yet been published to the EDC. Hope this helps! Please note that it is classified as Intel Confidential. Thanks! LynnZ

Update: the URL to document # 509619 is https://www-ssl.intel.com/content/www/us/en/secure/embedded/nda/products/bay-trail/z3700-series-rvp-fab3-crb.html?wapkw=509619 https://www-ssl.intel.com/content/www/us/en/secure/embedded/nda/products/bay-trail/z3700-series-rvp-fab3-crb.html?wapkw=…

[eAgan]

Thanks!

In case anyone else runs into this, I would like to point out that while the footprint in Volume 1 of the datasheet is incomplete (as of Revision 003), it is not as incomplete as I initially thought. I had not run in to staggered matrix BGAs before and it was not immediately apparent to me that the diagonal pitch (H) and that drawing are indeed enough to determine the position of all the outer balls. IPC-7351B makes this rather clear in its section on staggered matrix BGAs. In short, the A1 position is given by the fact that the number of rows and columns is odd, which implies that there is a ball in the exact center of the package. You can work out the rest of the outer balls from there with a little math.

What the datasheet omits is the pitch for the inner balls. Since I was only able to know that because of the document sent to me (which is protected by cNDA), I don't know if I should say what it is here. But I wanted to leave the note above in case anyone finds this who is also confused by the staggered matrix BGA footprint.

[Natalie_Z_Intel]

Thank you for the feedback. I will pass this information along to the document owner. LynnZ