nick-foley

Thanks for doing this, looking forward to more results. I've been interested in making some flexible parts and have been wondering some of the same things. Right now, I think I'm going to print negatives in PLA and cast positives in Silicone because I want a pretty low durometer and high level of detail. Might try some flex filaments if they work well though.

As I've said elsewhere, I'm psyched about this approach. Don't have a heated bed or any sort of enclosure but still get great results. I use this fan: http://www.amazon.com/gp/product/B001NPEBJM/ref=oh_details_o05_s00_i00?ie=UTF8&psc=1 and here's the very simple mounting bracket I use with it (hangs from the UM frame and attaches to the fan with the screws included with the fan): https://www.dropbox.com/s/cegv8beuku64wmt/Crossflow%20Fan%20Mount%20v0.3.STL

What are you hoping to improve by using the E3D on a UM2? I haven't used a UM2, but my understanding is that the hotend has been improved from the UM1. Unless you're trying to print polycarbonate, I'm not sure what major benefits the E3D will offer.

Yeah, I really like this design. Would love to make a derivative heavily inspired by an ultimaker - using the same printed XY blocks and printhead I've designed.

How does the extruder retract the filament, say, 5mm, if reversing the stepper direction is what also causes the other filament to begin extrusion? Is there another actuator, not visible in the schematics, that is controlling which filament is being fed?

Yeah, I think this is a great idea. I was looking into designing my own for a while, before I discovered the UBIS design. The crux is the custom ceramic heater coil that they use. I've looked at other solutions which seem similar (http://www.makergear.com/pages/ceramic-heater-core-instructions) but haven't found anything close to being off-the-shelf or even semi-custom. Lots of industrial suppliers make https://www.google.com/search?q=ring+heater&rlz=1C1CHFX_enUS540US540&espv=210&es_sm=122&source=lnms&tbm=isch&sa=X&ei=RwLBUvPsOIi2yAHlooGYDA&ved=0CAkQ_AUoAQ&biw=1600&bih=798#es_sm=122&espv=210&q=nozzle+heater+bands&tbm=isch for injection molding machines, which are very similar to an ideal heating element for a 3D printer hotend. They generally seem too large though.

Awesome! Just ordered more as well. Going to set up a machine with dual extrusion using this hotend.

I reused the Ultimaker cartridge heater, but switched to the E3D thermistor for temp sensing because the thermocouple doesn't fit. You just need to switch the firmware to the "Best Choice for EPCOS 100K" thermistor table and solder a 4.7K resistor to R23 on the main PCB. Do away with the thermocouple PCB. I would also invest in some high temp heatshrink and teflon sleeving (bought mine from the QU-BD site), since the E3D-provided method of thermistor installation and insulation is a real pain in the ass and doesn't hold up well. http://store.qu-bd.com/product.php?id_product=12 http://store.qu-bd.com/product.php?id_product=17 I think the UBIS is close to perfect for printing PLA, but it may also be close to perfect for other materials as well... just haven't tried it. I know the Printrbot bros do a lot of ABS printing so I suspect it's probably great for that as well

How do you differentiate between the printers from the browser? Does OctoPrint sort this out for you?

If this is really such a problem, just spend the five minutes it would take to model custom support material into your part using whatever 3d software you're familiar with. Auto generated support is convenient, but it's inefficient even in the best of implementations, and often a hindrance when trying to achieve perfection on a large subset of prints.

Increase the xy distance in the expert support settings. Try 1.5mm. Not a perfect solution but it generally works for me.

It's christmas, and all of santa's helpers are working double time. Printing a robot right now at 12mm^3 filament per second flow rate (120mm/s x 0.25mm layers x 0.4mm nozzle) and the results are looking perfect with this hotend. Definitely didn't think the results would be this clean. Edit - need to catch a flight! Sped it up to 13mm^3 and thing still look good.

I wish I could: Disable the auto-nesting buffer that gets used when placing multiple objects on the build platform (no matter how small I set my printhead settings, objects still can't get very close to each other without everything getting auto-adjusted) Right click on each separate body on the build platform and turn support on/off for that body only. (Also, object-level slicing settings would be sweet, but I won't hold my breath for that). Both of these features would make it easier to pack a bed full of parts, start the print, walk away for a few dozen hours, and return to finished glory.

Printed a bunch of knife handles that I designed to work with my favorite puukko blade: Also some ornaments I designed: http://www.thingiverse.com/thing:157592 http://www.thingiverse.com/thing:189376 3D Printed Christmas presents for all!

Oh. Yup. See that now. That works too.

Sweet! Just ordered another one too. So, now that others are trying this setup, It's worth figuring out the best connection strategy. The hotend comes with a male and female Molex connectors (P/N 43640 and 43645). I didn't have those on hand, so I just cut the wires and soldered my hotend in... but it would obviously be better to utilize the connectors (and I've made space for them in the printhead design).

Just tested this and yes, keeping velocity low with acceleration high is absolutely the way to go. Z velocity is set to 16mm/s and Z acceleration is set to 1500mm/s/s - and now the initial bed lowering happens nice and slowly, but the inter-layer z changes are still a lightning fast chirp. More testing is probably warranted here to find the best balance of these two values.

Useful to know, and actually very relevant to my last post. I'm not getting skipped steps during the incremental layer changes - I'm getting skipped steps during the initial bed lowering routine that Cura performs. An ideal setup might then be 1500+mm/s/s acceleration but 15mm/s or so max velocity - maximizing layer change efficiency but preventing crazy speeds during initial bed lowering.

Actually, it seems that I'm still getting some skipped steps when using 1500mm/s/s Z acceleration. I turned it down to 900, seems ok, but I may need to tweak my Z stepper driver to give more current.

Tried another rounds of tests this evening. After watching the machine print for a while, I felt the theory of uneven Z steps causing uneven extrusion seemed likely. I ran 4 more of the same prints while tweaking some settings, and there were some further improvements made. Everything was the same as the previous set, except I was using Ultimachine Yellow PLA instead of Ultimachine White. All tests included the dampening weight on the bowden which showed improvement last time. The first change I made was lubricating the Z lead screw with a high quality teflon + lithium grease made for helical gears. This did not seem to make much of a change in print quality (maybe a slight improvement?) apart from making my Z moves quieter. Second change was lowering the speed and acceleration of Z moves in firmware. I dropped the Z velocity from 40 to 10mm/s, and the acceleration from 700mm/s/s to 100mm/s/s. This noticeably decreased print quality - there was definitely more banding. With that knowledge, I increased the Z values up to 70mm/s and 1500mm/s/s (important note: prior to re-greasing my lead screw, the machine could not handle these values without skipping steps). This definitely increased quality beyond the starting print. Conclusions so far: - Dampening the bowden reduces banding - Increasing Z velocity and acceleration reduces banding - Having the Z lead screw well lubricated is required in order to have successful fast Z moves. I think the next set of tests will revolve around varying the tension of the extruder spring, as suggested by Daid.

Haven't tried beyond PLA temps but the Printrbot is designed for use with PLA or ABS, so I don't suspect it would be a problem to go up to 260°ish

Good snag. I guess I'll keep checking back for them to be in stock. I should mention that one unresolved question with this hotend is which thermistor table to use when building Marlin. I asked this question on the Printrbot help page but haven't received a response yet. Others have asked in the past with no definitive answer found. Anyway, I'm using "Best Choice for EPCOS 100k". It doesn't seem to be a perfect match (or my old thermocouple was significantly off) because I now print Ultimachine PLA using temps of 200-210°, where I was previously using 220-230°. Ultimately, it is a relatively minor issue in practice.

Just use thicker top/bottom values? It looks like those are just way too thin. Set it to 1mm at least.

Wiping the blue tape down with acetone completely solved the problem without going too far. Thanks for the suggestion!

Excellent. I missed the opportunity to buy multiples when they had them in stock a few weeks ago, and now wish I had at least 2 more.... I have a 1.75mm one that I purchased but don't have any 1.75mm filament to use it with. One additional benefit of this printhead setup is that the bowden tube can be removed along with the hotend without any disassembly - meaning switching between 3mm filament and 1.75mm filament can be done in a matter of seconds.