For my LR4 Beta build, I originally used cedar fence pickets to support the Y axis and Y axis belts, so I could use 2’ x 4’ plywood as a base without being restricted to a long, narrow strip of working area.
The downside of cedar fence pickets is that they are about as stiff as a pool noodle. When I push on the machine by hand, especially on the XMax side, the belts are very soft on account of the wooden rails bending rather severely. Even the metal rail on the XMax side doesn’t seem to help much.
I had printed most of the major RC1 parts (Core, YZ plates), but I was not fast enough and they became obsolete before I got to the RC1 upgrade. One observation, while doing multi-color first layers: if the first layer is red and the second layer is black, the red looks darker because it is not fully opaque and the black second layer darkens it. By printing a white second layer above the red, it looks much brighter.
For RC2, I decided I wanted consistent colors. For the tail end of a spool, I tried sequential printing for the first time. It worked just like it was supposed to, and I failed two small parts (which were ultimately recovered) instead of six.
While waiting for more filament to arrive, I got started on the rail upgrade, using some hardwood that I had laying around. I think it’s cherry but I could be wrong about that.
I also made the Y axis a bit longer, so the work area is 27" in Y. The idea is I have the full 24" width of the board (which will ultimately turn into a proper table) and I have 3" overhanging the front for ends of boards or for a 4th axis to hang down off the front.
Every time I see a version of this machine it just looks to be “the goods”.
I presume the multi-colour first layer was done using the “tell the machine it has an MMU attached” trick? - the red one in particular looks very clean indeed!
Yes, just a multi-material profile and manual color change. Fairly straightforward to import SVG as a modifier for the first one or two layers.
It does have a minor issue with goobers on the color change. I did try a (short) wipe tower but something about the sequence jammed the extruder 2 of 2 times that I tried it. It managed to pull the filament up, over and to the right of the extruder gear, which was a big hassle to remove.
I think I have a remedy that I will try next time.
There’s a knack to that! When I printed my 100-tile portrait, I had 600 colour changes in about 100 hours so I’ve had some practice!
After the change, there’s a small purge - leave it there, then, when you press the “OK” button after the “is the colour correct” question, it purges another ten mm or so.
If you either hold on to the initial “tail” before it starts to move the whole purge will come off in your finger! I have a brass brush that I use mostly, just as the nozzle starts to move. You have to be quick though.
Yeah I can get those purges but there’s a couple other things.
There is a little blob (pretty minor) left behind when the machine begins the color change.
The default color change gcode purges an extra 0.3 after the color change operation when the nozzle has returned to the part. I think this was the worst offender, and was fixed by removing that from color change gcode.
I don’t take great care to keep my filament dry and it oozes a bit between the color change and moving back to resume printing.
My idea is to add to the tool change gcode to move to a fixed location like (10,10) so it quickly moves away from the part and doesn’t begin/end the color change on the interior of a part.
Then, add my own ‘wipe tower’ of sorts that is a set of concentric tubes, one of each color, with center at (10,10). After the color change, if there is any leftover dribble hanging out of the nozzle, it will get deposited on the interior of the tube and not the exterior of the part. That’s the idea anyway. I have yet to find out if it works in practice.
The motors from StepperOnline come with little zip ties on the wires, which were a minor inconvenience. I decided to get clever and remove them and twist the wires together. It looks nice too in my opinion. But oh boy they do not go through the channel easily and I got one good and stuck and had to cut the end off. But still worth it I think.
The motor shafts do not have the flat along the entire length and for Beta I used a grinder to make the flat longer for the Y motors. Since Ryan updated the design, this time I tried the unmodified motors for Y and it worked fine. Nice!
I tried to figure out the pulley position from the marks on the YZ plate but I couldn’t figure it out. All the distances looked wrong.
The motors with extended flat segment I am using for Z now because last time I had to put the coupler a bit higher than I wanted, which meant I couldn’t go quite as low (a few mm perhaps) as I could have otherwise. Now the XZ plate goes all the way down to the bottom of the YZ plate.
I managed to strip one of the screws for the linear rail. Using my electric screwdriver I set the clutch to super low torque but it kept spinning without even hesitating and I felt no increase in torque. I think by turning it fast, the screw got really hot and really softened the plastic. For the remaining screws I did them slowly in increments and had no problems.
Also, not shown, I installed the min and max braces onto the gantry and left the rest in place. After it is working I’ll cut new strut plates and aluminum XZ plates. I haven’t decided if I’m going to change out the interior braces.
I had this happen to one of my beta YZ plates. The rails are probably fine with a missing screw or even two if the top and bottom ones are tight.
A trick I use is I pre-tap the holes with a hex cap M3 going really slow, and only then try to install the rails with the M3 Phillips screws from the V1 kit. When I back the Hex cap out or each location, it is surprisingly warm.
I think you’re right about heat from running them in with the power driver.
I don’t have a lot of extra M5 nuts floating around. I tore down Beta and assembled RC2 and I was short! I had to look through the parts to find the sneaky buggers hiding in the Z stubs, the Y idler pockets, and in the XMax brace. I think I got them all in the end.
I used a 6-wire ribbon cable and the idea was to match colors, but I made a mistake and the ribbon cable doesn’t have colors in the same order. This bothers me much more than it probably should…
Note red and green are reversed. Now that I see this, I could reverse the wires between the ribbon cable and the connector, so at least the connector is correct.
Looking back at Amazon, it looks like it is not standard, as there are a few different color orderings available.
Anyway, next is some more wire tidying and then I think I can cut some XZ plates.
Yeah, I’d have done that, too. That’s why I just buy the extensions that are all one colour with dashed and dotted lines.
Though then, I would probably deliberately mismatch some of those so that the wire colours all lined up the same at the control board. As is, on the Jackpot, X and Y1 are reversed from Z1, Y2 and Z2.
That would drive me crazy as well so you aren’t alone at all. Seems we all agree on this lol
Your fix seems to be a proper one and is probably what I would have done as well
I always just annoy the out of @vicious1 and order wire from him 100’ at the time. He has told me in the past he has to walk it down the street to measure it out lol. It has the needed colors in it and works well to keep things tidy
Yup, as mentioned elsewhere, I reversed some in the firmware so they all are the same orientation.
I attempted to switch the housings from the Dupont to the MX2.5, but for some reason they don’t fit. Maybe the mechanism by which they catch onto the crimped pins is different. The MX2.5 housings came with their own pins that look similar if you don’t look too close, but on close inspection they are different.
No, I don’t think it is - I think it’s a valid iteration of the process - as the machine improves with each model, it’s really important to pick up this sort of thing holistically. I was terrified when I turned on the LR2 for the first time and had to randomly hope it all worked.
This very systematic approach could make it much easier to build the machine by rote. That may or may not be desirable, but I think it’s worth following up.
out of 
