paste mask and SMT test points

[david_c_partridge]

I just produced a paste mask file for a design I'm looking to get produced, and was surprised to see that test points get solder paste! 

I would have expected not to paste these, as doing so will create a small solder pillow at the TP location, rather than a small saucer (so to speak) with the solder mask acting to make the edges (and allowing a probe to sink to the bottom of the TP, rather than skid off the top).

Is this intended or was it just not considered?

Thanks

Dave

[yyao]

Hi Dave,

Thanks for your question. I believe that the test point that we had in mind is more similar to:

http://news.thomasnet.com/images/large/819/819105.jpg

http://media.digikey.com/photos/Keystone%20Elect%20Photos/5015.jpg

Thus, it explains why the test point is implemented as a through hole or SMT pad. When I use test points, I don't bother soldering these lugs on (except for ground). I find that it is easy enough to place the pointy tip of my oscilloscope lead in the test point hole. Is the practice of not having solder paste on test points common? It seems intuitive to me that you shouldn't leave exposed copper unless it is absolutely necessary as it oxidizes quite readily.

We would be interested to have your opinion.

[david_c_partridge]

That's exactly how I use EWB SMT test points.  That's to say I probe the point on the PCB directly.   For that reason, I 'd rather assumed that the idea of the EWB SMD test point was just a hole in the solder mask.

I too very rarely have anything soldered to any sort of test point, except as you do for ground TPs, and for that I use some real 0603 test points similar to your second picture, and I have a quite different part in my personal parts list for that purpose that uses an 0603 footprint with both pads connected together.

SMD Circuit Probe Pad 0603

OTOH, if the design intent was for them to be used with things soldered to them then I guess putting solder paste on 'em makes good sense.

As for bare copper I don't think I have any of that around on my PCBs  - HASL or electroless gold maybe, but bare copper - shudder.

PS I tried the blunt end of the 'scope probe but it didn't work for some reason :-)!

[yyao]

OK, so in your opinion then, what should be the right combination of solder paste/mask and copper for a THT or SMT test point?

So far, the only case that we've seen the need for exposed copper is for fiducial points on fine pitch SMT components. The different colour of the exposed copper helps the pick and place machine camera easily identify the location and orientation of the component. Since the fiducial point is electrically inert, there's no harm in letting it build a healthy layer of patina.