That's been my experience. Although the programming aspect seems painful on older machines.
Well, my CSM84 is an "older machine", and it does have a teach mode, but that seemed totally insane, so I have NEVER used it.
I got the pick and place file out of my CAD/CAM system, and hand-edited the first placement file, got it to work, and then wrote a C program to convert the P&P output to the file format the machine required. This is all pretty much automatic. I do have to create a component to feeder list for that program. The issue there is assigning feeder locations optimally, and remembering that feeders larger than 8mm use up two feeder slots. So, it is best to group all larger feeders together rather than alternating 8mm 12mm 8mm 16mm etc. which wastes more feeder positions. The most-often-used components are placed next to the board, so that XY travel is minimized.
Older systems were restricted somewhat by the computing platforms they were based on, on top of that the focus was seemingly entirely based on production with the concept that an engineer setup each and every product before letting operatives loose on building them. There was little to no attempt at ease of use, integrated setup tools, speed of changeover or integrated intelligence, that just isn't how modern systems work, some of those features might be optional extras which would be unfortunate if you picked one up 2nd hand without them but on the whole anything from the last 10 years at least shouldn't be making you jump through hoops to carry out trivial tasks. However that is the ins and outs of operating a PNP machine. I maintain the process itself remains basically simple until you start attempting the funkier components such as the tall, the high power
One other area is the solder stencil! Getting the right shrink of the apertures is a tricky business, and it took me a couple years to get it figured out MOSTLY right.
For the vast majority of parts you should be able to get away with 1:1, where you avoid this is on things like large pads (that you would probably crosshatch anyway) or on pads that you know extend further under the device than normal. Reasonable summary here
http://www.surfacemountprocess.com/a-guide-to-effective-stencil-design.html you can also probably get bags of advice or indeed a folder full from your stencil supplier (there are advantages to not using China for everything). If your footprints often cause you trouble you need to be looking at them or your pasting process. Of course some issues you cannot see with the naked eye, like voiding and that can require extensive work with both profile, stencil design and even reflow process if it is critical (vacuum helps $$$).
But, yes, there are MANY steps, and fouling any one up could make a big mess. Loading the wrong part into a feeder slot will get you a bunch of rework. Hope it isn't a many-leaded chip.
This is one of the many key arguments for intelligent feeders, you can still load the wrong part into the feeder of course if you aren't barcoding them on, but feeders that know where they are loaded and what is loaded on them and machines that alter the pick routine based on where they get slotted in... godsend, would never go back.
And, there are some parts that just are NOT made to do P&P on just any machine. I'm thinking of MELF and round-top LEDs. So, avoid those types if they will cause a problem on the machine you have.
Interesting, I wouldn't have had these down as machine issues, thats a nozzle issue, MELF has been around since forever, I spotted some in an retro computing article about an early Apple, I can place them wihth the same nozzle I'd use for their flat counterparts although you could opt for a nozzle with a notch cut out of it. Equally domed LEDs, I place 1000s with a standard kitted nozzle that is bevelled on its inner edge, the pick/placement parameters have some overtravel to compensate for the fact some of the device is in the tool. Those LEDs with a weird sticky gel lens however cause lots of people problems and often need a special tool.
All machines have some height limit, so you have to be careful to not exceed that, or parts will go flying. Some machines have programmable trajectories, mine doesn't. You can, MAYBE, trick your way around this a bit by placing parts from back to front on the board, assuming the feeders are on the front. But, the machine still has to get the parts out of the tapes, so you really can't handle much taller parts, anyway.
Jon
Tall parts are new, old machines were often more akin to chipshooters than flex placers (even when they weren't quick) so they can't handle them, sadly the prime cretin here is the electrolytic which may get rather more prevalent once conflict mineral rules kick Tantalum in the ass in 2021. Depending on the machine you may or may not be able to trick it, on Essemtec it knows the maximum height for each feeder type, it won't let you tell it a tall component the machine or feeder cannot handle is located in one. These days its fairy reasonable to expect a machine to either include collision avoidance routines or offer you a way of defining rules, that way if tiny part A is right next to giant part B, the machine will avoid putting down B if A is not fitted. Equally 2 giant parts next to each might mean it has to alter its routine to not start z travel until it has reached destination. Invisible tasks on modern platforms, possibly not so on old ones.
Home
Help
Search
Login
Register
