10
\$\begingroup\$

I came across this device just now, and I can't think of any reason for it to have the window that it has.

A 12-pin SO package of some sort, with a glass window on its underside.

This isn't just a cutaway display image, either; the datasheet specifically mentions that you should avoid getting glue under it to avoid cracking the glass:

A section of the datasheet, saying that you should avoid using glue lest it crack the glass.

So, what is the purpose of this window? Surely it increases the cost of manufacture, so it wouldn't be there without some good purpose, right?

I can't see it being for trimming, either, as what seems to be exactly the same part is also offered in a different package that doesn't have a window.

\$\endgroup\$
  • 1
    \$\begingroup\$ Probably some type of trimming of the frequency has to be done in manufacturing after packaging. Once the trimming is done the window goes on. Or possibly the part can be laser trimmed through the window, though I'm not sure if that's possible. \$\endgroup\$ – John D Nov 25 '18 at 2:16
  • \$\begingroup\$ @JohnD I looked into that after you commented and found that the same part is sold in packages without such a window as well, so while it was a "wait why didn't I think of that" level of obvious answer, it doesn't seem to be the right one. Details appended to the question. \$\endgroup\$ – Hearth Nov 25 '18 at 2:26
  • 1
    \$\begingroup\$ Another guess: the window package is low-profile (1.2mm) others are thicker. I'm wondering how a lead-frame can be encapsulated in pure epoxy and still leave wiggle-room for the tuning-fork? Maybe a pure-epoxy mold requires more height than a thin-glass cap to assure clearance. \$\endgroup\$ – glen_geek Nov 25 '18 at 3:11
  • \$\begingroup\$ Is it possible that Digikey shows a windowed version only to illustrate what is inside the package? Rather than to suggest you can get windowed parts? The datasheet they link doesn't (to my reading, anyway) show a windowed option. \$\endgroup\$ – jonk Nov 25 '18 at 4:19
  • 1
    \$\begingroup\$ The tutorial from Epson digikey.com/en/ptm/e/epson/spxo-and-vcxo/tutorial implies that the crystal is encapsulated in a ceramic package and that the chip is either inside this package or the package and chip are embedded in a plastic package. I am guessing that the lowest profile is obtained when one side of the ceramic package is exposed and that this glass is part of the ceramic package. Perhaps it isn't practical to encapsulate a crystal in plastic alone? \$\endgroup\$ – Steve Hubbard Nov 26 '18 at 3:31
7
\$\begingroup\$

@SteveHubbard has it right. The crystal has to be hermetically sealed preventing gas migration into the chamber, so plastic alone will not work. Glass is an excellent material for creating gas-tight hermetic seals with through-conductors. The crystal chamber will be filled with a dry inert gas or a vacuum.

\$\endgroup\$
3
\$\begingroup\$

The tutorial from Epson digikey.com/en/ptm/e/epson/spxo-and-vcxo/tutorial implies that the crystal is encapsulated in a ceramic package and that the chip is either inside this package or the package and chip are embedded in a plastic package. I am guessing that the lowest profile is obtained when one side of the ceramic package is exposed and that this glass is part of the ceramic package. Perhaps it isn't practical to encapsulate a crystal in plastic alone?

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.