We've been using fids as badmarks on panelized boards, with reverse recognition so that an uncovered fiducial is a good mark. It's worked fine with 1.5mm fiducials but now we've run into problems with a board with 1mm fids and we can NOT get it to work properly. Sometimes I get false failures, sometimes the vision check on the badmark results in a "block will not be populated" message but it puts parts on the board anyway.
Any suggestions? As it stands I can skip blocks in the program but it would be nice to not have to make program edits for every X-out.
We have a small 2mm square solid white box silkscreened on each board for badmark recognition. You may be able to do the reverse of putting something like "white out" in a specific area to act as a badmark.
I am not completely sure why it would tell you "board will not be populated" and then build it anyway, but if the only difference between the two programs (working one, not working one) is the fiducials are smaller, did you try reducing the search area? I use 1mm fiducials for badmark recognition on our Topaz, and our Emerald. If you like I can tell you how we set up our Mark info for these. but doing a Parameter search on your fiducials should give you similar results. I will check out our settings and then get back with you.
I am using a search area of 2.5mm and I use a Red dot sticker to cover up board fids to not be built. I have it set to reflect. threshold is around 100 but that could be different depending on what you are using to cover up your fiducials with. I will play with our machines and see what may cause this. The above settings work on most of our boards though...