snap-off pcb sections - mouse bites vs solid tabs.

[phil from seattle]

I've got a design that I want to be able to make a PCB with 1/2 of the board that can be snapped off. Similar to Sparkfun's old Proto-Snap stuff or the STMicro's Nucleo-64 boards (where the programmer can be snapped off). I see a lot of people talking about "mouse bites" (a series of really close 0.5mm or smaller holes) in the connecting tabs but the Nucleo-64 boards have solid tabs. I can certainly make mouse bites but STMicro seems to be shipping a lot of boards with these solid tabs. See the image below.  I measured a couple that I have. The slots are 1mm and the tabs range between 2.5mm and 3.75mm wide.  The PCB appears to be 1.6mm FR4, could be 4 layer. I am tempted to go with this style even though a lot of people seem to be pushing mouse bites.  Am I missing something?

I will neck down the traces that cross the snap line to avoid lifting them though the Nucleo boards do not appear to do that. I'll probably instruct people to score the line with a knife before snapping.

[retiredfeline]

Have you tried bending FR4 before? I wonder if they expect people to dremel it.

[goldserve]

There are probably traces that run over the section so it is expected to use something like a dremel to cut the debugger off.

[phil from seattle]

Hmmm, I've been assuming you can just snap it off. But, when I went to look for info on that.  Nothing.  They usually say to "cut it off" with absolutely no detail and show a scissors icon in their diagrams. No video of cutting it off. Nothing. Has no one ever done that? I guess I should just give it a try.

By the way, I'm pretty sure I can just scribe the tabs with a knife to cut the traces.

[retiredfeline]

Try it with a junk board. You'll find dremel or hacksaw is more appropriate than scissors or cutters.

[ataradov]

Hmmm, I've been assuming you can just snap it off.

There is absolutely no way you will snap off those ST boards. You will need tools.

Trying to snap off that much will require a lot of force and it is highly likely that PCB will delaminate much further than the connectros. C14, C15 and R27 are all goners if you try to do that without additional scoring or just using a dremel all the way though..

Mouse bites is the way to go, even if you need to route traces - do that between the mouse bites.

[mikeselectricstuff]

Definitely mousebites - you'd need a depanelling nibbler tool without.
  I use unplated 0.5mm holes 0.75mm apart

[phil from seattle]

Yeah, I gave the Nucleo a try and it definitely didn't want to break.  Mouse bites, here I come...

[Doctorandus_P]

C13 & R26 (Above middle tab) and C14, C15, R27, R29 (near right tab) are exceptionally badly placed, and an attempt to break (or even bend) this Nucleus board is could break any of these components.

[poorchava]

Mouse bites (perforations). From my experience it's  actually easy to overdo it to the point where they are difficult to separate, even when there is like 0.3mm of material between holes.

5 holes/4bridges of 0.2...0.3mm width every 60...80mm is generally fine for pnp assembly. Anyway that's the rule of thumb I've been following with good results.

[phil from seattle]

Thanks, I went with mouse bites - two tabs, one 2 hole and one 3 hole with .3mm between holes. [probably should have made both 3 hole, we'll see on that.] Used multiple thin traces and vias to anchor them to prevent peeling/lifting.  Should be getting the first turn of the boards back in a few days. Will post results.