Adding Test Points to PCB Layout Traces

[Steve Mavronis]

I was wondering if test points were added to a trace as part of a PCB layout design, should any considerations be made in regards to the test point pad diameter vs trace thickness?

For example if you have a 50 mil thick trace and a test point pin through it should the pad around it be say 100 mill or so in diameter?

Sorry for these off the wall wierd questions that probably are of no real concern! Maybe I should just think of it being no different than any component solder pad.

[petemoore]

  Ill effects may be cumulative is about the only caveat [a weak one at that], assuming you aren't running long leads to the 'elsewhere' chosen as a reading point and only need about a few pads.
  That said...
  I occasionally find need for [and was glad they were created] my slightly raised component lead, for referencing 'important' schematic node points on the PCB etc, also for  having a place the probe can find or the clip can attach.
  Perhaps a dot of color code ink or paint there, on the PCB.
  I can think of the times it was hard to get the probe on pin3 of the opamp and nowhere else, but this is fairly easily circumvented with layout forethought...especially if there happens to be a component that wears a very large tophat...there've been times I couldn't even see the probe point for sure to find it for sure, much less try to only touch it ! Makes it extremely difficult especially when measuring inter-board continuities or resistances.
 

[cpm]

i got a bunch of this, solder some solid core wires and use them to attach to breadboard for circuit testing and modding

they are cheap, and reach easily in between ic pins, and tight components...

[Steve Mavronis]

Yes these probes are good for that. I was thinking of test points in a trace connecting parts of a circuit where you may want to scope the signal. Like on a distortion/overdrive circuit: one between the input and the first component, one after the gain circuit stage just before the op amp input, one after the op amp output, and one after the diode clipping section before the output.

[defaced]

If you're doing professional boards, you can define areas (at least you can do this in eagle) to block the solder mask - whether it be on a trace or a ground pour or whatever.  This area will be tinned by the tinning process and available for a test point.  This is similar to what was on on the POD XT Live board i have.  Basically they're solid circular pads that have a label next to them. It made working on the board really nice.   

[caress]

I was wondering if test points were added to a trace as part of a PCB layout design, should any considerations be made in regards to the test point pad diameter vs trace thickness?

For example if you have a 50 mil thick trace and a test point pin through it should the pad around it be say 100 mill or so in diameter?

Sorry for these off the wall wierd questions that probably are of no real concern! Maybe I should just think of it being no different than any component solder pad.

no different unless you want to go down the route defaced suggested.

[PRR]

There should be TONS of clear space all around the TP, and nothing should smoke when I aim my 0.5" jitter-hand at your 0.1" pad.

0.1" is no easier than poking a pin directly.

If your design is so flawed (or in this case: intended for less experienced builders) that TPs are "essential", then I'd argue for stand-up pins, lugs, or loops to hook onto.

[Steve Mavronis]

I was considering stand up test pins not because of any design flaw, but for measuring performance and studying/documenting operating specifications much like you do with software simulations like LT spice.

[R.G.]

I was considering stand up test pins not because of any design flaw, but for measuring performance and studying/documenting operating specifications much like you do with software simulations like LT spice.

If you need test points for a production (i.e. over quanity 100) design, that's one thing. If you need test points for studying performance on a board you're building one or two of, it might be simpler to find the places where you want to probe, then solder cut-off resistor leads to the existing points in the circuit on the component side. The components offer some protection against bent-over leads, and you can find the test points on the board easily to clip onto with probes.

[PRR]

I didn't mean to offend.

After bitter experience with a very complicated gizmo built as-tight-as-possible, my feeling for experimental work is to lay-out as BIG as is rational. My next complicated gizmo was 2/3rd the stuff in twice the board space.

When experimenting, especially if (as often happens) it isn't doing quite what you think it should do, then ANY point may be the point you need to probe. I've traced odd glitches to very obscure corners, places where I had not expected a problem.

Building BIG is your best defense against needing to probe something in a tight spot.

In high-voltage/power work, hand-holding probes is bad luck, and a popped-off clip-lead may brush something nasty. Building big is good practice

OTOH, when it is pretty sure to work well, and you want to document or ponder, OR when you build stuff which needs calibration, then a few well-picked points may be all you need.

When you build a TV, and power-up, right away you want to trim the V and H circuits so the loop-control voltage is in the middle of its range (so it won't out of range and stress costly parts). On the assembly line, the worker drops a probe on a test-point and trims for green-zone on a meter.

R.G. suggests lead-scraps, and I've studded boards with them. More formally, bend scraps in a U and insert them in two holes: cheap, stiff, and a clip-lead tends to stay clipped. Alternatively, when you know you are just voltage-probing, stuff one end of a 1K resistor standing vertically and loop the free end. Now if your DMM happens to be on AMPS, or brushes ground, you have a 1K shunt instead of a dead-short, which might smoke a part.

Of course if the board is loose, and all/most the traces on one side, you memorize the parts, mirror your memory, flip it over, and probe from the back. Classic PCB probes great. 2-sided, 6-layer, and surface-mount may not probe nice.

[Steve Mavronis]

No offense taken. I guess my main rational for maybe having test points wasn't purely for troubleshooting problems. So far I'm trying to build from classic established circuits and wanted measurable result documentation of the signal at different stages as a way of benchmarking. I guess you could just as well do that with handheld probing but I was thinking hands free for scoping, voltage measuring, etc. These posts I make are most of the time thought provoking exercises anyway to get exactly the kind of feedback you've all been giving!

[Nasse]

In these "throw away and buy new, it´s cheaper" days it is good thing to make products more service friendly. Those probes linked are handy, and I got few and found them useful when breadboarding.

[Processaurus]

It's handy to solder resistors standing up, so at the opamps' output side of the resistor the lead makes a loop that you can clip a probe onto.  In general putting the longer, loop side of a standing up resistor anywhere you might need to look at on the scope is handy.  Also having a big loop of wire on the ground is essential.

I think making official testpoints are more appropriate for production debugging, or calibration, rather than experimenting, which is usually easiest on a breadboard.   I've noticed some builders here tend to saddle the horse before they're ready to ride, jumping into the process of layout and PCB making, before designs are tweaked right or debugged on a breadboard.  Modding designs on a prematurely made PCB is usually ugly and time consuming, and later the mods can be squirrelly to follow, especially on a two sided board.