I still think what you are describing can be attributed to the settings. The techs may not have the same finesse of setting everything up - an example from another industry: when Triumph Motorcycles were resurrected a few years ago, the factory was purchased with all the existing machinery. However when the new machinists were trying to make motors, they could not maintain the needed precision machining. Finally a retired machinist was able to come in - he had a lot of items like drag sticks that could be wedged against a turning shaft to limit motion - if left free you may have a bore that is 7 thousandths to big, but with the wedge in the mill, the bore would repeatedly come out with less than half a thousandths variance....Sadly its not an easy fix!
I'm one of the ones having issues with the FC9000 cutting off. Mine doesnt go off much... but its nowhere near as accurate as the 8600.
The first row will be fine, then it'll slightly shift - so small stickers are an issue... but anything big is ok / accurate enough, but it makes most 4" or under stickers look lopsided.
I've had techs out a half dozen times, they replaced a ton of parts, factory reset everything... but it still tends to "Drift". Even the graphtec techs were stumped - Then on the other 9000 I have access to, it'll cut the exact same file perfectly.... I've been dealing with it for months with many visits, we're on the verge of demanding a replacement cutter be sent in, everytime we do what the graphtec techs ask, they have more stuff for the rep to try - Grimco is getting pretty pissed because they have to come out once a week to do new suggestions from graphtec, and not one thing has made it any better....
Seems to be a batch of bad graphtecs out there or something. For 99% of the stuff I do it doesnt matter, but for smaller stickers... everything looks soooo lopsided and unacceptable. My problem unfortunatelly doesnt fix with 4 corners, so its a bit different than the others... It's like my wheels are pushing the material at an angle and it doesnt correct it properly. Even with 4 corners, first row is fine and then it slowly goes out more and more... I told them I'm close to buying a summa and throwing this on craigslistIve had an 8000, still have a 8600, and 2 9000s... and only1 of the 9000s does this, the others are dead accurate
think about all the forces that go to your media while you are plotting... there is acceleration, jerk, drag, slip, and is all directions - the acceleration is only in and out, as this is the only way the media moves, but while this is happening the blade is dragging through the substrate, but the tip of the blade holder is also dragging (ideally you want enough down force to guarantee the blade plunges through the substrate and stays there, but no more as more down force causes more drag). When cutting near a roller, there are different torques generated than when you are at the maximum distance from a roller... I am positive that most if not all of the "problematic" 9000s will register within spec if all the forces are taken into consideration. (I have had a lot of people tell me that you cannot cut banners, and that even posters look terrible with the perf cut - but they are trying to do fine surgery with a hammer - I set the blade depth to within a fiber, or less of cutting through the banner on the UP stroke (all the way through on the down stroke) - when you pop out the final posters or banners, the edge is clean because there was most like only a halfway cut through pulp fiber holding everything together.
PATIENCE and realizing even a single degree of turn on the blade depth adjuster actually makes a difference (though you do also have to do thing like making the final movement of the depth adjustment being the down, ensuring there is no slack in the fine threads of the adjuster. on a guillotine cutter the last movement of the back gauge is always forward for the same reason - if you move it back to measurement, you can push the gauge a thousandth or so back as it takes up the slack in the threads....) if EVERY adjustment on the machine is thought through with this type of detail, your 9000 will plot perfectly...
