[dorkbotpdx-blabber] 0201 parts in the toaster....

[Erik Walthinsen]

Somehow I managed to mis-order some parts a while back, and don't have
anything of the right size at the moment. I ended up getting a strip of
0201 beads rather than 0402...

I decided 'what the heck' and tossed one on the board. It worked...!

The "big" black is an SOT-23A LDO, and next to that is an 0402 capacitor
and an 0201(!) bead, un-reworked out of the toaster.

I'm getting a Silhouette Cameo tomorrow, and I'm curious whether it can
cut stencils that small. It should, since Travis cut us a stencil with
0.6 x 0.3mm holes for an LGA chip, and the IPC land size for 0201 is
0.42 x 0.44mm. Whether I can squeegee that is another matter.

[Rich Hoge]

You will probably get more tombstoned devices trying to use the 0201 on the 0402 pads, especially of your placement is shifted too far either direction.  Depending on how many 0201 devices you have to populate, you might be better off just printing the pads, then manually soldering.

_______________________________________________
dorkbotpdx-blabber mailing list
dorkbotpd...@dorkbot.org
http://music.columbia.edu/mailman/listinfo/dorkbotpdx-blabber

[Erik Walthinsen]

On 08/02/2012 11:32 AM, Rich Hoge wrote:
> You will probably get more tombstoned devices trying to use the 0201 on
> the 0402 pads, especially of your placement is shifted too far either
> direction.

Only 1 of 4 I've done so far has actually tombstoned. Honestly I was
expecting them all to....

> Depending on how many 0201 devices you have to populate, you
> might be better off just printing the pads, then manually soldering.

I'm planning on being better off by not using 0201 ;-)

[Monty Goodson]

In my previous job a design team was forced to resort to board level circuitry after upper management declared "no new ASICs!" for the project. That was a bit of a concern as the circuitry needed to process 16bit words at multiple GHz. The result was a very dense mass of small RF transistors and ECL logic along with 1000's of 0201's. The 0201's were new for the in-house board assembly, and they had major concerns about yield. In the end, they saw almost no tombstoning and surprisingly good yields. Granted this was professional assembly, and they paid a lot of attention to the process parameters.

I was curious as to whether there might be something about 0201's to make them less prone to tombstone than larger sizes. A quick google search found this nice study of the effects of pad geometries and other factors on 0402 and 0201 tombstoning:
http://www.xs4all.nl/~tersted/PDF_files/Plexus/tombstoning.pdf

0201's had more defects, with grossly oversized pads being the dominating factor.
0402 defect rates were surprisingly insensitive to pad geometry

Thanks for sharing your results Erik,

Monty

[Rich Hoge]

On well designed pads, with good printing and utilizing SPC, 0201 are not prone to defects at any higher rates than other SMD discrete devices. Pad design geometry and paste print openings in the stencil can be optimized for robust manufacturing. Tombstoning generally happens when there is a large metal imbalance between the two sides of the SMD device. 

My comments in relation to Erik's project was based on having too large of pads and possibly basing the device to on side or the other.

Rich