de-panelizing scored PCBs

You can buy cutters to do this. The cutter uses 2 sharpened wheels that seperate the boards.

Look what I found with Google:

--
Reply to nico@nctdevpuntnl (punt=.)
Bedrijven en winkels vindt U op www.adresboekje.nl

The appliance is called a pizza cutter (at least our mfg eng's call it that). For small volumes, I suggest tab-routing and use a grinder to flatten the surface.

If working by hand, you can (at some effort) depanel with minimal stress to the PCB by running an Exacto blade down each side of the score repeatedly until you've thinned the material in that location sufficiently to break easily. A Dremel also works but is MUCH harder to control; you almost certainly won't get nice clean straight edges that way!

You can 'recess' the tabbed section to avoid this.

Graham

I've done scoring like this with an X-acto that had the tip broken off, so that there was a square corner on the blade, then I shoved it along the score line, and dug material out rather than slicing it.

Cheers! Rich

If a board has to butt up against _anything_, it's a poor design in the first place. (unless it's part of the enclosure itself, in which case you'll have a whole nother set of tolerance issues. ;-) )

Cheers! Rich

This little guy is a pull cutter/scorer, which might be a little easier to control. I've used it on plastics; should work OK on FR-4. Cheap, and the working edge folds into the handle.

The link is just an example based on a google search. Got mine at the local un-franchised hardware store down the block.

--
Rich Webb   Norfolk, VA

How well would it work on PCBs that have parts on both sides?

On 12 Jun 2006 13:25:23 -0700, snipped-for-privacy@gmail.com Gave us:

Much cheaper in the long run. A dremel tool is a lot cheaper than that pizza cutter, which is likely a couple grand.

The board doesn't quite butt against the enclosure; it's separated by about

15 mils. But that's too small for a tab, unless I grind it down, which seems like a pain and which generates fiberglas dust that I don't really feel like breathing.

FWIW, the reason for the proximity is that this is the board that the front-panel connectors and pots are mounted on. The connectors are 3.5mm stereo jacks; the space between the front PCB terminal and the bushing is only just barely enough for the PCB pad. I have not (yet) found any panel-mount 3.5mm connectors with PCB terminals for which that is not the case.

So the only alternative I'm aware of is to hand-wire the connectors, instead of PCB-mounting them. That's what I do now, and it's too slow; I can't meet distributor demand.

Can you think of a different option? The givens are that there will be two

3.5mm connectors and two pots mounted on the front panel; the connectors must be mounted in such a way that connector insertion does not result in undue stress on the solder joints; and similarly, the pots must be mounted in such a way that manipulating them, and possibly dropping the unit on its face, does not result in undue stress on the solder joints. (I've fixed too darn many pieces of gear from other manufacturers where these rules weren't followed!) If there's a good alternative I'd be interested to hear it.

Thanks, -walter

Yeah, you can't rely on the edge routing for much accuracy, nor v-scoring. I think you'll find that drill hits for mounting holes will be more accurate. And yes, depanelizing v-scores does flex the board if you don't use one of the cutting wheels (at $3K+ each).

Follow-on to another post here, who says the tabs have to stick out from the PCB edge? Indent the tab from the edge of the board. If you have two neighboring boards, make the tab like a disc that can break off both boards. It'll probably take pliers to grip the tab for the second break.

To avoid extra points on the rout, you could use an oversized drill hole on either side (larger than the routing bit), and smaller drill hits indented.

:-----------------: :PCB1 ooo : :======OtabO======: :PCB2 ooo : :-----------------:

One approach might be a small PCB for the front-panel that's ribbon cabled to the main PCB - among other reasons, so it can float / flex with the abuse. The pots should be able to be fastened to the panel with nuts front & back, but I hear what you're saying about the mini jacks.

Using a close-tolerance front-panel hole for the mini-jacks and mounting the PCB somewhat loosely would cause the panel to take the lateral forces from the mini-jack movement, rather than the PCB. Likewise, don't let the jack protrude if at all possible, so the panel takes any any heavy insertion forces, rather than the protruding jack.

HTH, Richard

front panel pots mounted on the pcb...

Pots mounted on main pcb is not an abuse proof system, consider what happens when the case flexes slightly during a drop. As you say, hand wiring is slow. One popular solution is to mount the pots initially on the pcb, with a tab or hole style break line, and wires crossing the break. Now, everything is machine assemblable is one go, and you just snap the break line when putting the pcb into the box. Thats it.

NT

It can be tested before snapping as well.

Best regards, Spehro Pefhany

--
"it\'s the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

You route leaving tabs and place 3 holes across the tab aligned with the PCB edge. The tab breaks across the holes leaving nothing protruding from the PCB that can't be wiped away with a single stroke of a file.

We have been here before see if you can find Message-ID: it was in alt.binaries.schematics.electronic so probably not on google.

The idea is to cut the PCBs before placing components.

--
Reply to nico@nctdevpuntnl (punt=.)
Bedrijven en winkels vindt U op www.adresboekje.nl

But the whole point of v-scoring the boards is to place the components while they are in an array (less handling of individual boards, so cheaper assembly).

You mean, like the diagram that Richard H. drew? That seems like it could work.

Got it - thanks.

Yes, that's exactly what I've got. The main board is mounted with screw holes, edges don't butt against anything. The small PCB is just big enough for the pots, jacks, and a ribbon connector to connect it to the main board.

Scoring between the main board and small board could be tabs or v-score, doesn't matter. The "front" edge of the small board, though, still needs to get separated from the big array, and it needs to have a clean edge (or at least concave) so that the jacks and pots can be screwed to the front panel.

The suggestion above seems like it could work for me, with a little bit of conversation with the board house to make sure they don't plate the holes.

Thanks!

Good point. I suppose there are cutters available that address this problem.

--
Reply to nico@nctdevpuntnl (punt=.)
Bedrijven en winkels vindt U op www.adresboekje.nl

BTW, you can combine V-groove and routing, at a slight cost premium.

Best regards, Spehro Pefhany

--
"it\'s the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com