Hi,

If there is no space for normal test points, or if one has forgotten to add a specific test point, an alternative could be to use a component pad
as the test point. I particularly think of probing the soldered side of a capacitor/resistor/ferrite bead 0805/0603 pad.

There are some potential issues I can think of:
1. Capacitor cracking due to too high force while the pogo pin presses against the component.
2. Solder flux residue which may prevent good contact.

I would guess the safest parts to allow for test probes are 0805 (and bigger) resistors and ferrite beads...

What first comes to mind is using a flat, larger, tip with small sharp teeth (not the conical tip style) pogo pin in order to evenly distribute the force.

Has anyone done this, and what is your experience?

Place the SMD pad of your choice directly on the net and assign both pads to the same net and use something like these:
https://www.digikey.co.nz/en/product-highlight/k/keystone/surface-mount-test-points

These SMD test points and most like them bridge both pads and therefore maintain connectivity and if these are only for prototypes the pads can either be removed in the next PCB version or a zero ohm SMD resistor be added where the test point was.

tautech: This not what I am talking about.
I would like to probe (in a test jig) one side of a 0805 resistor on a production PCB since I forgot to add a test point on this net.

tautech: This not what I am talking about.
I would like to probe (in a test jig) one side of a 0805 resistor on a production PCB since I forgot to add a test point on this net.

Bugger, I hoped it was easy. 

AFAIK if SMT parts are probed directly, then usually a needle is pushed under angle into the solder joint, not directly onto components. It's certainly a high risk cracking ceramic components if probed directly. And it's not like exposed terminals on top of the components are big enough that you won't miss them with a test needle. And this is done with flying probes, not test jig.

I am not speaking from experience, but my thoughts are that:
1. Parts crack because of flexing of the board; so it doesn't matter whether you probe the capacitor pad or a pad near the capacitor; the stress on the cap would be the same.  In fact, probing a test point could flex the board more, depending on where the supports are.
2.  Would a single pogo pin exert enough force to flex the board significantly?
3.  You can support the board from behind to minimize flex.

I am not speaking from experience, but my thoughts are that:
1. Parts crack because of flexing of the board; so it doesn't matter whether you probe the capacitor pad or a pad near the capacitor; the stress on the cap would be the same.  In fact, probing a test point could flex the board

The only way to probe directly is using sharp needles. Force needs to be big enough to punch through flux residues. Then what do you think happens if you punch needle directly into the component? You can make a crack in ceramic.

If you're testing boards I assume they've already been washed. I've made test jigs that did this same basic thing. Require some patience and precision. Also a human to keep board in place during test program. Can use bare fr4 super glued to face of jig as a guide which actually works alright.

Then what do you think happens if you punch needle directly into the component? You can make a crack in ceramic.

I maintain that the only way you are likely to crack a capacitor is by flexing the board at the site of the capacitor.  Not likely with direct pressure on the capacitor - especially from a pogo pin.

Quote from: wraper on 26 Mar, 2018 23:39

Then what do you think happens if you punch needle directly into the component? You can make a crack in ceramic.

I maintain that the only way you are likely to crack a capacitor is by flexing the board at the site of the capacitor.

I guess you don't deal with with MLCC that much. MLCC can easily crack by just hand soldering them.

Not likely with direct pressure on the capacitor - especially from a pogo pin.

Round pogo pins are pretty much useless for this purpose. Sharp one can damage the part.

Quote from: wraper on 26 Mar, 2018 23:39

Then what do you think happens if you punch needle directly into the component? You can make a crack in ceramic.

I maintain that the only way you are likely to crack a capacitor is by flexing the board at the site of the capacitor.  Not likely with direct pressure on the capacitor - especially from a pogo pin.

SMT ceramic caps break if you look at them cross-eyed. A pogo pin will break it.

How big is the pad?  If it's a nice big hand-solderable area, there's probably enough for a sharp pogo pin without too much trouble.  Minimal pad area, not so much..

Tim

The pad is a normal 0805, so the pogo has to hit the top of the component. It could be argued that using capacitors may not be a good idea.
However, ferrite beads and resistors may perhaps not take any harm. And, the solder flux may even not be present on the top of the component, as the
solder normally don't migrate all the way to the top during reflow, hence a sharp pogo may not be necessary.

tautech: This not what I am talking about.
I would like to probe (in a test jig) one side of a 0805 resistor on a production PCB since I forgot to add a test point on this net.

I would not exert force on a soldered smd component. And as you said, flux residues (this can be mitigated with proper board cleaning after assembly) and variations in solder height will make it hard to get good contact in a repeatable way. Additionally, solder joints are not flat, which will add to the difficulty.

You can try with very low-force, pointy pogo pins. But it still may be hard to scale this to a production level.

I guess you don't deal with with MLCC that much. MLCC can easily crack by just hand soldering them.

I have been probing MLCC caps with scope probes for decades - often with considerable force and not once EVER has one cracked.  And thermal stress from soldering is a completely different issue.

The pad is a normal 0805, so the pogo has to hit the top of the component. It could be argued that using capacitors may not be a good idea.
However, ferrite beads and resistors may perhaps not take any harm. And, the solder flux may even not be present on the top of the component, as the
solder normally don't migrate all the way to the top during reflow, hence a sharp pogo may not be necessary.

When I did it we used 0603 and 0805. We used sharp pogo pins which didn't require a ton of force but we also used a small clamp to hold boards during testing program. This was used for years for production boards and the jig was never the issue. The boards were always washed but two 3 pin cables were soldered after the wash and those weren't cleaned or used in testing. These were audio boards.

There is also another solution when applicable. Very often there are DNP components in a design, the pads for those components could be used
as test points. The only issue here is that there definitely is flux residue present. One could use pogo pins which are designed for penetrating the flux.
Another solution is to make some sort of masking on the stencil in order to prevent solder paste to deposit on the pads.

I use GC 10 paste.

The first idea I think of is placing some kapton tape beneath the stencil. Stencil rinsing and uneven height below the stencil are potential issues though...
I wonder if there is a known working solution for masking apertures on stencils...