Chinese Production SMT Machines???

Don't go down that route, Juki cut spares support a few years ago on these. Although bits are available from China, you do not want to be swapping used cards. They are both also dog slow in real life whatever you do to them. Especially a 760. We're still nursing a couple and we have excellent techs.

If you must buy a bargain basement machine, find out what is used local to you and how good the support network is in your State so you are not spending big money flying people in all the time.

Mydata's are very easy to program and fix, and the old Philips/Assembleon GEM machines are solid, easy and you'll get 10K placements an hour from a Topaz. However you may find that your $10K machine soon gets bills for double that in year one.

We get conveyors, buffers, loaders, feeders, nozzles, cable cutting machines etc from China, and have great contacts out there, but we still wouldn't dream of buying an unknown generic brand machine - we have seen too many people end up with unstable ovens, flaky wave solder machines, patchy AOI etc.

That's all even before you look at the machine safety aspect....

reply »

Sparkfun users (professional hobbyists?) will have a totally different metric for what is acceptable compared to anyone who is doing this at a proper production level. Where I work I can at least empathise a little with the kind of business you might be running, others here run huge setups and have never seen the other end of the spectrum. When we started doing surface mount we used one of these : https://www.youtube.com/watch?v=ZaqvEftw7aI a UK built essentially tabletop placer that on a good day places may 1300cph. In USD that was probably ~$30k with a selection of feeders. Printing was done manually and we ran a small 4 -zone IR oven. All standalone like you intend to run yours. Today we still run standalone but using an Essemtec Paraquda, this manages ~5.5kCPH and personally I would suggest this is probably about as fast as you want a standalone setup to run. This is because you (puny human) are now the limiting step between each process and if you are trying to multitask or are otherwise called away form the line your work rate will fall and quality will suffer (you'll do stupid things like run the place process twice, reload a reel into the wrong feeder, put the board in backwards or just simply not be there when your budget machine goes wrong). Equally while it was discussed I wouldn't have wanted to run multiples of the slower machine either. On low end machines or machines where your simply pushing the limits of what they can do you will end up "tweaking" the placement with a pair of tweezers. We were doing that with the Versatronics and I know of other places who were doing it when running 0402 through an Essemtec FLX (machines which would have stated they placed 0201). You can certainly see the need for this step on the Chinese machines used by sparkfun types in their Youtube videos - accuracy is terrible.

As an illustration here is a UK dealer in 2nd user machines http://www.shawline.co.uk/product.php?id_product=98 who does list a price if you register. Listed is a perfectly respectable Europlacer machine with a good service history @ £23.5kGBP by the time you add feeders delivery installation and a bit of training that number probably doubles but that figure is pretty common for usable old machines. To get cheaper there is typically a catch, cannibalized spare parts only, known issues, software that only runs on a very specific custom computer, software you need to pay to transfer it to you. Or indeed the hardware is so crap no broker/dealer will touch it and you're getting it directly from the desperate seller.

Your stated target place rate easily justifies buying brand new. It also easily justifies getting a well equipped contractor to build them for you at a decent price - they might well have lines running at 30, 60 or 250 kcph. There are contractors for every type of job. We specialize in the 1-1000 PCBs range for instance, which is exactly where lots of people who have just designed something and can't get a foot in with larger manufacturers, or can't deal with the 6-12 week lead time) need us to be.

If you really want to some of your own manufacturing get started small and simple, take your time, learn the ropes use a reliable supported brand. In the US I would probably look at something from Manncorp like this https://www.manncorp.com/pick-and-place-med-high-volume-mc400.html In the UK I'd maybe buy one of these http://www.intelligentdrives.com/imp.php or a TWS Quadra, Essemtec used to be an option too but they've moved their machines further upmarket which combined with being Swiss, can make them expensive. And one last tip, do not take your manual soldering tips from DIYers Youtube videoes, the technique demonstrated by most of them is awful, paste should "roll" into apertures, they mostly hold the squeegee so near flat its pushed through. The stencil behind the squeegee should be left clean after the stroke too.

reply »