Sunday 4 December 2011

While Sod's Law=True; Something Breaks; Fix It; EndWhile

The last few weeks have been eventful, apart from turmoil in personal life, I've made a few changes with my reprap, and taken bloody ages to get round to finally blogging this. There is still more to come afterwards too, so stay tuned.

My Adrian's Geared Extruder got jammed (again) so I took apart that old extruder to replace it.
White ABS quite stained from heat and fumes. I've seen this before somewhere.
The filament drive was quite clogged, and after later cleaning it out with a brass brush I found that it wasn't just that grip had been lost and clogged it, the notches on that brass M4 insert really had started to wear out and dent flat in line with the filament path.

That part having been squashing and grinding at the plastic as usual, this came out:

So, with a bracket that I customised for shorter bolts from another design someone sent me, I hobbed an M8 bolt and constructed a Hinged Accessible Wade's Extruder to GregFrost's design.
My first attempt, not too bad (moved along slightly at one point when I found a slim washer that would allow me to cut at a better distance along the bolt), though I did another one later that was better.
I found that the hole for the PTFE insulation was too small, and after beginning to try and carve it bigger, thought 'screw this' and plugged the hole with a piece of bottle cork with a hole drilled through. :)
The oversized cork later got squashed very slightly as it took up the compression from the hot-end fixing bolts.
I found that this extruder actually wouldn't fit on the regular x-axis carriage due to the PEEK block orientation; I earlier thought it would be OK when looking at how the bolts would sit in relation to the carriage insides, but didn't realise that on installing the extruder, the top of the PEEK block would be higher than the bottom of the x-carriage, so it couldn't fit in there. :(
For a temporary X-axis I used Prusa's standard one for PLA bushings, since I had a spare one sitting in a box that looked ugly due to previous problems with backlash.
So far so good...
I started trying to print the final version of a clip that I had been designing for a while to go on top of the accessible extruder and stop the filament coming in from being bent too far to the side out of the hobbed bolt's grip (since there is no filament guide to enable easier cleaning, hence the name accessible extruder.
Annoyingly, the extruder started to jam when I was trying to print with PLA, and a few other things that I tried to print did similarly.
An earlier design for the clip that didn't work well due to the axis it was printed in (the bottom print surface is on top in the picture, and the bolt clip snapped off by delamination/peel stress).
Unusual about the picture above; the extruder seems to be designed to use M3 bolts to push the idler in, but I couldn't find any long enough, so I cut the captive nut holes out to M4 size and used the bolts pictured above. The spring was the stiffest small spring that I had on hand at the time (before I took apart the old extruder).
Bases of a pulley that I wanted for my x-axis idler, and the filament guide, on the left and right respectively.
When I went to the reprap IRC channel for troubleshooting suggestions I was advised to thoroughly clean out the extruder nozzle using acetone to help, since it seemed to take a lot of force to push filament through it by hand while hot, though I couldn't accurately remember whether it had always needed that much manual force.
I soaked the nozzle in acetone for about 3 days while very busy with other things and when I got round to taking it out, I found that there was still a significant amount of undissolved plastic stuck in there, which was mostly easy to remove with hand-applied torque on successive sizes of drill bit.
The nozzle outlet, which is 0.5mm in diameter, was a pain to clear out as I couldn't find a pin or needle that thin anywhere, let alone a drill bit, so I eventually used a short length of copper wire from a bag of scrap wire insulation that I'm hoping to find a sensible way to dispose of.
Very frustratingly though, upon trying to put the hot-end back together, I found that the threaded end of the PTFE insulation no longer fit the thread inside the brass nozzle, and was starting to strip the tread when I tried to put it in.
PTFE? Warping? Oh right, yes that does happen sometimes. -_-
It looks like I ought to get another insulation block as spare, but anyway after much messing about trying to find the correct thread size to re-cut it (needed an M8x1.25 die if I remember right, thanks again pete) so that it would fit properly, I got the extruder back together to try again.
While trying to run a print again, I wasn't happy with how much friction the PLA bushings add to the sliding of the x-carriage, and how loosely they held onto the smooth rods, so I decided to design a new x-carriage with the combination of properties that I need.

When I came round to trying to print parts for my new x-carriage in ABS, I was faced with a couple more extruder failures, so I tried successively reducing the distance that filament retracted during moves from about 1.2mm down to 0.8 then 0.5 without it helping much, also increasing the target printing temperature from 240 to 245C. Sadly, this didn't help, and I was unconvinced that the clogged nozzle was my main problem, as on trying to manually push filament through the hot-end again, it seemed to be just as difficult (though easier at slightly higher temperature, hence trying that).
The filament drive definitely wasn't gripping evenly; sometimes extruding, sometimes not. That made it difficult to calibrate...
Eventually I tried a combination of reducing the retraction distance hugely to 0.2mm, the print speed to 30mm/s, and replacing the small extruder idler spring that I was using, with one of the 4 larger ones that originally held the idler on my old extruder, as I had a hunch that maybe the spring I was using wasn't applying enough pressure, against the advice I had to the contrary. This solved the problem of the extruder jamming just nicely, albeit producing quite a few loose threads of ABS in the print.
However, a short while into the print I noticed a couple of other problems; with the spring now making the filament drive tight enough, too much plastic was being extruded, making the extruder bump against layers of plastic in . I ended up sitting by the printer alternately using one hand to hold the carriage onto the rails, and the other to draft this post, for nearly 2 hours. That's not something I want to do again in a hurry.
When we talk about repetitive labour...
Part of the way through I even innovated something to make the job slightly less tedious (though still very tedious), as the way I was mounting my cooling fan before (too small for the holes on the carriage which were for a 40mm fan), was allowing it to drop down too low, cooling the hot end too much when I wanted to keep the filament drive and motor cool. So I did this:
Wire twist - easily installed with one hand (obviously I took this picture afterwards)
Here I show the dodgy carriage and express my thoughts on the situation:

Eventually I got this result:
Fugly and stringy (until I cleaned it up, then just fugly... but usable)
I put my new x-carriage top together with the old bottom part, and on trying to fit the extruder was reminded of why I was using the Prusa carriage in the first place. While I had made plenty of room in the top part to turn the extruder around, the old bottom part didn't have room for the diagonal alignment of the PEEK block on the hot-end, so not to be set back anymore, I had a go at the part with my pen knife.
It works... what more do you want?
Finally, order has returned to the system (for now) and it's working better than before, since my re-designed x-carriage allows the timing belt to be tightened on-the-fly using a neat trick that I've seen used on a couple of other designs now; an M3 nut and bolt embedded in the belt-clamp:
The new carriage has mounting holes for both this tiny 32mm fan, or a 40mm one :)
Since testing by printing a few standard Prusa structural parts, with the filament retraction returned to normal, it seems that spring pressure was the main issue here.

Good news on a design front though... I've finally got round to starting to do some CAD on printable parts for a cowled wind turbine, having only had ideas for it stuck in my head and a few scrawled sketches so far, while a group in Michigan Tech Uni have made what sounds like a better attempt at a relatively cheap plastic recycler/extruder, for which I hadn't thought of a good way to drive filament through yet, after mentioning the work of Delft Uni students before. So if their system works well and is easy to source parts for, I might not have to worry about that anymore. :)
I've also recently met someone else through TZM who seems interested in coordinating a project to start prototyping DIY rotary-hydroponics solutions into something we know works and is easy to build, which we hope can get the assistance of Douglas Mallette, as I hear he's also interested in developing that.

No comments:

Post a Comment