good specs for a "break off" debug + testpoint tab

[julianhigginson]

Has anyone got any basic spec or approach or pointers for a breakaway debug header + testpoints tab?

I've finally met a design size requirement that's not going to work any other way.

PCB is nominal 0.8mm 4 layer with HDI aspects (laser drilled blind vias, and vias in pad)

here's a screenshot of my first attempt..
FYI, the 1mm routing path break gives the tab a min distance of 3.6mm, and that's 4 x 0.5mm non plated holes at 1mm pitch between them on the original board edge.
I'm a little worried it's going to cause problems as-is.

[Gribo]

If you plan on using pogo pins in production, you want to keep your pitch at 0.05" at minimum. 4 test points at 1mm pitch will have to use a staggered arrangement.
Also, have a look at the Tag Connect footprint, you might be able to fit it somewhere.
I would also rearrange the traces not to be on top of each other, you might have shorts after de-panelizing the PCB.

[Ian.M]

Also neck down surface traces where they pass through the mousebites to control where the copper breaks and reduce the risk of peeling up the end of a trace.

[julianhigginson]

yeah I have a tag connect cable already.. used it once where I had a board that didn't have the space for a header, thanks to a low profile custom case... But I prefer cortex-debug header normally it's a lot more straightforward.

Also, yeah testpoints for production will be spaced properly for bed of nails... I can do that. This version of the board is gong to be more for development/prototyping, but I will try to do test points for Pogo pins  on the breakaway tab just in case it's perfect and can become full production.. :-) 

Thanks for the idea of narrowing down the traces at the intended break.. I was worrying about the copper not breaking, and ripping off the board. though would be cutting the mousebite, so probably not likely unless someone tried doing it by hand.

[julianhigginson]

I was avoiding internal traces, and have only routed top and bottom across the mousebites because i was more worried about internal traces shorting.. is surface track tearing generally seen to be more of a worry than shorting of cut mousebites?

I could relatively easily change it to routing on the two internal layers if that is seen as better, in actual experience.