March 21st, 2017 - The good, the bad, and the ugly [ARCHIVE]

Hello again everyone, and apologies for another long delay in updating after our UL testing. As many of you know from e-mail, we ran into a string of hurdles that kept getting worse and worse - so this update is unfortunately mostly bad news, though there are a few bits of good news sprinkled throughout. The TL;DR is we ran into two more QC issues - one of which is solved, one of which is undergoing endurance testing - and then had a pretty major issue with our shop.

The good

We passed our UL factory inspection! Our HiPot tester works, we passed all of their testing, and after waiting longer than we'd hoped to get approval, we are good to go on that front. At this point, the only thing holding us back is our own in house issues, which we should have solved soon.

The bad

We ran into two more QC issues, one of which we actually brought upon ourselves when we solved our earlier grinder issue. The second, we didn't catch until we completed our final testing after UL where we load, calibrate, and test each unit before shipping.

Issue #1 - grinder interlock is now "too good".

Most of you will remember that we had to design and manufacture a plate to hold the electromagnet, so that the interlocks would disengage at the right torque - this is covered in a previous post. Unknown to us at the time however, is that this solution - while it solved the problem it's been meant to - introduced a new one, in that the interlock would sometimes energize itself to the point of sticking. Other improvements in our production and assembly process also meant we'd removed more static friction from the entire assembly, which is normally a good thing, but combined with the sticky interlocks this meant that the grinder would no longer "pop out" under it's own internal resistance, which of course is not good for safety reasons!

The solution we're using for now to get units out the door is to re-introduce some internal friction so that the grinder will pop out on it's own. This has proven to work so far and we're currently endurance testing a few units to ensure that the internal friction won't "wear away" over time and re-introduce the issue where things stick. We should know definitively by the end of the week, and be able to implement the change quickly to move forward.

Some shims we've laser cut to sandwich at the input of the grinder gearbox, which create just the right amount of friction for the grinder to disengage on it's own!


It might be hard to see, but between the input gear and the housing are our shims. By torquing the input gear bolt to the same torque every time, we get a consistent resistance that kicks out the interlock when it should. 

Issue #2 - stater interference with the auger

This one is a bit more annoying because we should have caught it earlier. While our "In house" stators we used for endurance testing never had this issue, the production versions - made from the same drawing, to the same spec, but with more room for variance within that spec - did. Ultimately, the 4 bolt pattern that holds the hot section, insulator, stators, and cold section together was not sufficient to accurately locate the stators around the auger. The stators need to be centered around the auger or the auger can rub on them at a certain point, introducing tiny bits of aluminium into the filament and clogging the filters far too quickly (which is how we noticed the problem on some units).

Far, far too much aluminum in the nozzles! 

The very tip of the auger sticks past the stators, and is the original / correct diameter. The rest of the auger has been ground away by the stators...not good!

The obvious solution and the one we've implemented is to ensure that the stators are always exactly located around the auger. This involves precisely locating and machining two dowel holes in both the stator and the drive section, which aligns everything exactly on center, instead of relying on the four bolts to hold everything in alignment. As an added safety measure, we're also taking a tiny bit off the diameter of the auger just where it sticks into the stators, to ensure that any slack introduced over time won't suddenly start wearing away at the auger. We've tested this solution already, and  have confirmed it works on multiple units - so now we just have to machine the dowel holes into all of our stators and drive sections.

The old stator on top, the new one on the bottom showing the alignment holes for the dowels.

The ugly

I'd been hoping to post an update on the above two issues early last week, as it was getting to be too long since our last update and we wanted to get the news out. Then something much worse than two solvable QC issues happened - our laser cutter caught on fire. While fires in laser cutters are relatively rare, they are possible, and cutting flammable materials like acrylic and fiber board can leave behind very flammable residue that, if not cleaned regularly, can catch fire. Unfortunately for us, this proved true, and despite catching the fire almost immediately a combination of flammable particles left in the bed, flammable material being actively cut, and an air pump shooting right at the fire can escalate things quite quickly. We were able to put it out and no one was hurt (though the fumes did require a temporary evacuation of our unit), and the damage isn't too severe or costly...but we do have to completely take apart the unit to check for anything else, completely clean off any residue whatsoever off all of the optics, rails, and electronics, wait for the new parts to arrive, put everything back together, and re calibrate for use. We're currently planning for this to take about a week (2 weeks total), and between fixing the stator issue and getting the laser up and running as quickly as possible, production is on standby for the next few days.

The laser right after we put out the fire (and the air cleared from the fumes) - there's extinguisher dust on *everything*

The laser after we'd removed all of the damaged components, and started to do some of the cleaning work. You can see where the fire was on the left side of the bed and on the left side of the window.

This is what's left of our "X" axis gantry belt...

Our limit switch also got pretty badly melted, and will need replacing, along with some other wiring and electronics mounted on the gantry.

Unfortunately between all of these setbacks, we are nearly a month behind our most recent schedule: ~1 week from waiting on UL, ~1 week in solving the QC issues, and ~1 week due to the laser fire. As always, we're happy to answer any questions or concerns you might have over e-mail, and we are endlessly thankful for your patience while we work out the struggles of producing and shipping our product. It's taken us much longer than we'd hoped and setbacks like this are incredibly disheartening, so it's extremely helpful to have some of the best customers a team could ask for supporting us!

Hopefully our next update in a week or two is that we've started shipping, and until then, wish us luck!

-The ReDeTec team