aviphysics

In the United States, there is a device called the Silhouette Cameo that is similar. There are more heavy duty devices than the Cameo, but they are more expensive. Also, the Cameo has a mechanism to detect index markings on prints, so you can precisely cut out printed work. I love using mine to cut out papercraft stuff. http://www.amazon.com/Silhouette-Electronic-Cutting-Starter-Bundle/dp/B00NAX7H78/ref=sr_1_1?ie=UTF8&qid=1426480014&sr=8-1&keywords=silhouette+cameo

I have seen those black blocks somewhere before. If they are what I think they are, they also hold a clothespin spring inside, to help with tension.

But then each nozel is only being used during its turn. If you had several machine with several beds on a rotating platform, all the heads could be used at the same time. It would be like having a production line where each machine does its own part of the process, accept each bed could work on entirely different parts. I can think of a few reasons why it isn't a superb idea, but for the right application, it would be awesome.

You can reverse the direction of a stepper motor by swapping any pair of wires. Your build looks like it came our real nice.

Crazy idea: What if instead of switching print heads, several machines were mounted upside down in a circle and connected to a rotating axis. They would at the same time print their material for that layer on separate print beds and then rotate to print their material on the next bed. It just seems so obvious and insanely great.

Slop is really negligible on a factory UMO+ or UM2 (though UMO+ would benefit from GT2 belts and sprockets). The bronze bushing seem 99% fine. A design that used proper linear rails might last longer,but it adds cost and, more importantly, weight to the xy carriage. The Type-A Series 1 printer has a carriage along these lines.

I have been wondering what the shortest time to change print heads might be.

If you are custom building, why not get electronics that support the voltage you need without having to mod?

Why beagle-bone over smoothieboard? The thing that confuses me about bone's description is that the implementation didn't sound that different from having an Octoprint feeding an Arduino running Marlin. Isn't it just an ARM chip pushing gcode or similar commands to a microcontroller? That was the basic impression I got, but it was pretty hard to separate the product from the crowdfunding hype. Edit: I am actually thinking of the T-Bone cape for BeagleBone. Regardless, I would hate to be in the market for a new controller right now. Just too many options and too hard to tell what platform will be successful in the long run.

In summary, I wouldn't focus on copying the UM2 specifically. Just copy the design points you like from both machines. Besides construction materials one is not supremely better than the other.

I am not really sure what the substantial difference between UMO+ and UM2 is. Both have the same z carriage, same motors, same xy carriage structure, same heated bed and same electronics. If you are building a UM variant, it would be pretty hard to tell which UM it came from. Main differences I am aware of are the following. - UM2 structure is not wood. - UM2 xy carriage switches which rod is on bottom and which is on top. - UM2 has a slightly different hot end design with two fans. - UM2 stock feeder doesn't have enough torque to print fast and rough. - UMO+ has more mods developed for it, though there has been a lot of activity from UM2 users lately. - IIRC, the UM2 comes with GT2 belts and sprockets. I have a hard time understanding the UM2 appeal. It is a great looking machine, but for the less than the $1000 price difference, I could do a lot to make a UMO+ as good or better. The only difference that is not so easily corrected is the wood frame, but I haven't seen evidence of it being an issue.

I make an effort to have each pair of sprockets lined up when slipping the belts on. Now that my new gt2 sprockets have two set screws each, it is a lot easier.

Retract while combing plugin would affect the gcode, not the firmware. It really doesn't make sense for it to be a firmware issue. I guess you could try seeing how your firmware handles g0 and g1 moves that travel in all 3 axis. If it goes up or down before traveling along the xy axis that might explain it.

The sand thing sounds cool. For thin copper tubing, I bet you might be able to get away with printing a pipe bender, or just buy one. For soldering, I would look at how plumbers do it and do it that way. That is the same as the way I have scene the pros do it for this kind of stuff. The only difference is that you are soldering the sides together instead of the ends.

Hmmm. I remember that my UMO used to do that, but at some point there was a change in the order of the start code and now it lifts and primes after homing. It then goes directly to the start of the print. I can't remember why I had problems fixing that behavior at the time. Seems like it should all be in the startup gcode.

The fan is just powered from a simple transistor that is right next to the fan header. I believe it is the main limiting factor in the current you can draw. When I get home, I can look up the part number. I wonder if water is really necessary. You could try pumping air with an aquarium air pump. Water just seems like a mess waiting to happen. At work we use neumatics for cooling parts inside a vacuum chamber. They work pretty darn well and a small leak doesn't destroy the system. For water cooling, I would suggest that you solder a copper tube to the top of your plate. Would be easier if they are both copper.

Thanks for pointing me in the right direction. Regarding the artifacts, I don't actually see an issue when using opaque plastics. I do however normally use at least 2 or three boundary layers, which might explain that. Regarding the quality of the fill, this is what I was thinking. 1) When I have printed with 0% infill, but left combing on, the combing strings do a pretty good job bridging gaps a wide as 25 mm. 2) Generally, when I do use infill, I use about 10% to 15%, where the longest gap in the grid is substantially less than 25 mm, so the layer above it should be able to bridge that pretty easily. 3) Normally, the main reason I use infill is to support a flat surface at the top of the print. For this purpose, the infill just has to be strong enough to support the layer above it. I do however think that randomizing layer rotation would only be useful for sparse infill. Unfortunately, that decision is made further down in the code. Just makes it a little extra work, if one wants to make that decision a configurable option.

If you want a terrible screw, all you need is to grab a 1/4" 28. Those things are just about the worst and will vibrate lose if you so much look at them wrong. M5 is a far superior choice. Also, for some reason nearly all SAE screws/bolts I come across seem to be total junk compared to the metric screws. I don't know why, but metric screws just seem to be made to higher tolerances.

Here you go. $20 for a pair of digitial calipers that read in both Inch and Metric units. http://www.harborfreight.com/6-in-digital-caliper-with-sae-and-metric-fractional-readings-68304.html This one is only $10 http://www.harborfreight.com/6-inch-digital-caliper-47257.html

Yeah. I am gonna have a lot on my plate when this thing arrives. It will mean moving my UMO out to the garage, which will also probably mean building an enclosure for it (which might as well be heated), which will mean finding a way to cool the steppers and electronics. That last one seems like it will either require keeping the machine thermally separated from the print some how, or some sort of water/air cooling lines. Both of those would be easier if the printer stayed in on place. If the printer is at the top, I would need a ladder for a lot of stuff, so having the print at the top would really help.

The code for generating fill has a rotation parameter, but I am having some trouble finding the code that calls the fill routines and therefore sets the rotation parameter. Would someone please point me in the right direction? What I would like to do is try randomizing the grid rotation at every layer and see what that is like. I think it would not be as strong, but might help provide support with less clearly visible artifacts (e.g. my objects printed with support in translucent filament currently look like they are wearing pinstripe suits.) Would also like to try randomizing the translation of the grid, but that looks like more work, cause there isn't an existing parameter for that.

In the USA, a lot of places fill the glass until it overflows a little and all the foam runs off.

Would there be any advantage for turning the assembly upside down, and/or maybe lifting the print bread instead of the UM?

In the model, you can make tiny perforations around you screw hole that are too small to print, but Cura will still add wall thickness around them. You could try experimenting with that.

I had this problem pop as well. It went away when I switched to gt2 belts. My theory is that the original belts had gotten slightly stretched out, so they weren't in perfect alignment with the grooves. I think it is part of the problem of having a belt/sprocket with a square tooth profile. Never figured out why it always happened at the exact same position along the belt.