>>Just sold a customer a GSP III for about $15K MORE than a MPM AP-20 I was
trying to push to him that was equipped with a Pro-head, stencil wiper, paste
dispenser etc. Go figure! These people out here are Fuji brain washed, but hey if
that's what they want, I'm not going to argue.<<
geez, hopefully those dopes are one of my competitors ;-) it is kind of funny what people will do when they don't understand a process. that is, unless fuji corrected their zero downstop feedback problem.
hey, maybe their stencil house recommended the fuji ;-)
i've seen a newer version of the mpa and see that it has the same problem moving from part to camera to pcb. maybe they have another model i haven't seen yet.
>>Speaking of MPM, I was reading an article about a technology they are beginning to use that "injects" the paste into the blades through an enclosed pump head that will eliminate the paste exposure to the outside elements and reduce viscosity changes. <<
:-) :-) :-)
i have to smile. i'm familiar with their auto dispense, too. one problem is that it cycles across the stencil and then proceeds to dump solder paste all the way across the back part of the stencil including the mesh! ;-) you also have to make sure that it cycles after the appropriate number of prints or else you can have quite a bit of solder paste on your stencil. especially if the ops decide to manually dispense solder paste not realizing that the auto dispense will shortly dispense even more. it can get messy.
another problem is that we have to change paste every two hours. in time, though, i think it will bear out to be a real positive once some of the negatives get ironed out.
we use mpms. some are "hi es." our cycle time isn't that low. we're looking at about 18 seconds for normal cycles and then about 50 seconds for clean cycles. we're actually ok except for the fact that when you have a series of bad prints and/or lift the hood (eg, changing solder paste tubes every two hours...) then you actually take away from production.
i'm currently using demming's problem solving cycle to minimize some of the root causes of our bad prints. i have a few ideas but i don't really want to start until i have the data necessary to conclude how effective, if at all, the changes are that i make. i'm working on setting up some spc charts to do just this.
btw, gate and bottleneck, as you pointed out, are exactly the same. |
