neotko

Thanks! For fonts 12mm seems big enough to handle it. For that I would (on s3d) make the process use no perimeters and increase the infill overlap to 90% so the travel path of the sanding process goes as smooth as possible, also I would set avoid crossing distances to maximum. The small blops can be pressure remaining on the nozzle, you could cut down the extrusion to a minimum while doing the sanding so the process focuses on scratching the layer lines without filling the microgaps. Probably using neosanding 2.0 would work better for fonts, but it has a few drawbacks like no matte finish unless you play with speed/temp.

With a 8od 5id adapter it could be very easy and cheap to add bondtech to the r17 without paying for the new motor

WOW!

I would pick the bcn sigma. I seen users with old models having issues on the feeder and hotend, but not any user with the new R17 models. So for the price, that's lower, is better because: - Has modern steppers without the Zebra stripes issue like all current ultimaker models - Hotends are cheap as chips, for a 230€ you get a full hotend kit with nozzles 0.3, 0.4, 0.5, 0.6, 0.8 and 1.0mm that's for the price of 2 UM3 cores. - Bed is fully usefull, no excuses of how they define 'print area' or clips to block, or the need to install a builtak to raise the bed to avoid the bed clips. Also they use Borosilicate Glass now, no just Tempered glass. It has a 0.1 max error on the surface, on UM they don't even tell you what error ratio you will get. - Bed leveling system is just easy as cakes, I will do a simple video showing it because it's great imo. Also their firmware allows you to readjust on 0.025 decimals manually if you want to fine tune it and actually know what's changing. - You can enable Duplicate/Mirror mode to print two objects at the same time (using half the bed for duplication or a bit less for mirror since the head could collide otherwise) so you can use the printer to print 2 of the same objects, production wise is a nice option (they have it as default for sigmax and I made a firmware after @amedee explained me how to est the marlin compiler to comment 1 line of code), I even made a video about that XD - Firmware is very easy to edit and upgrade like on a normal printer, no need for Signed security firmwares like UM3 or the danger of briking a 3k machine because you edit the UM3 Json files and you mistype a comma (been there done that) Print speed and reliability for me are on par of the UM3, and using BCN3D profiles the prints go nice and smooth. - The feeder is a knurled bolt, that's good and bad, I never got an issue so far, but I see some old sigma users that they have changed it to bondtech for more reliability. So far I don't see the need, but I know that bondtech is a better feeder. Upgrading it it's easy, but on um3 if you plan to upgrade their feeder you can really brick the machine (3 weeks ago an user on twitter did brick his by just changing a wrong data on the machine um3 json. Ofc UM did send him a new board to replace his. - Price of components is more earth-level. For example the main board is 125€ on the sigma (3djake) and steppers are around 100€ for all. So a full electronics for the machine is around 225€. On the UM3 the 'printerboard' (the um3 white board) is 300€ and the Olimex linux board is another 300€ (that's 600€ for just basic electronics). - Talking about price, the sigma is 2200€ (without taxes) and the um3 is 3000€ without VAT. That's 800€ that you can expend on bondtech, a full hotend pack, and you can even put some spare glasses for fast production. - Motion system on the UM3 is a light gantry with thin 6mm shafts to keep the weight down, but holds a 250grms printhead + cores. So it's moving a 300+ grams with thin smooth rods. That's the big speed limit for the um3, and the reason Cura team developed a lot of tricks to improve the ringing, this tricks are also used on Sigma, because it makes the print look better. Ofc the difference is on the infill settings... But that's for another day. Basically Sigma gantry is more heavy, but all rest on a solid frame with linear rails, that unlike smooth rods, don't wooble. On print speed both machines are very close. - Future upgrades on Sigma are very easy to install, and is actually Open Source, you can even download the tools they use to adjust the machine or see everything. On UM3 they took a year to just post some step files, and still they haven't post their firmware (not talking about their marlin 'driver') but the actual firmware. I don't know why they fear so much to being copied, but clearly they have a protect IP fear that blocks them to share back. Now offtopic My plans on the BCN3D Sigma for 2018 are to make the gantry lighter, thanks to the input from @gudo and other user on facebook. Probably using a markforged corexy idex setup. Would remove almost 400grs on the gantry making it a seriously fast dual printer. But well, that's just me, I like to mod stuff. So far I'm using it without mods =)

Oh not really afaik. Extrusion is quite precise. Also mind that I been using 3mm square rod that’s more flexible without an issue, 4mm should just be better (ofc for 2.85). The main issue of all is that noone sells steel key/precision rectified rods for 4mm longer than 375mm (or at least was never able to find any shop online).

Indeed Z wooble due directdrive occurs, with the flexible beta addon is minimal. Gudo and I are moving our printers to Corexy with rails where Z move due directforce disappears completely without the need of the flexible addon. Check this thread I post on twitter for progress. So far been using it without any issue for production

On um3 you could achieve same results, slower ofc. Is a matter of cura settings and many options they added to compensate the speed and improve outside look. Double extruder makes all more complex also, but as a single vs single extrusion the um2+ is better indeed.

I’m very sleepy now but also tou need to use OLDTOOL for coolddown and newtool to separate the before and after tool change.

Seems like a issue of names M104 sets temp and M109 waits until temp is achieved You need to set old to cold, set new to hot and wait new to reach temp T0 M104 Sxxx change old to cold T1 M104 Sxxx change new to heat M109 Sxxx wait for new to reach X temp

If your other machines work but one doesn't produce the same results, then clearly is a problem of the machine. I would inspect the x/y squareness and belt tension. You know, usual suspects to start with. Also, ofc, um3 isn't as good for single prints as a um2/um2+ the nozzles all, is different, and specially the profiles in Cura have tons (literally 10-20-30) options just to make um3 prints 'look' good. It has very fast infills to compensate the slower max print speed, and even the nozzle size extrusion is change to 0.35 for a 0.4 to make the outside 'look' better but on the expensive of accuracy. Me and others have hit over and over to UM about this (some not publicly) and the need for a 'Engineer' mode where all the features that are there to make it 'look' ok get removed in the favour of precision and constant speed, things that are a must to get an accurate part and not just a nice thingiverse click-print. (yea, half rant there, but not unrelated). So I would start by unhiding all the slicer settings and try to set at least, the nozzle size to the real nozzle size, the infill to the same speed of the rest of the print, and disable specially the acceleration/jerk automatic control, it has huge impact on how the print looks and how precise it's vs the object real size (specially on corners)

Found a bug, reporting it: iphone 7s with current latest version. Seems related to the code tag

Indeed it doesn’t seem an overhang issue indeed. Even less at High very slow default quality. That model doesn’t have that hard overhangs for a um3 fan. I would like to see the gcode. Are you using latest cura version and the machine firmware is up to date? Also and very important maybe is just no correctly calibrated on x/y plane. They do calibrate it before shipping it but maybe it just got decalibrated due traveling. Is a simple process and just will take 20-30minutes Other problem maybe is because you changed the infill. Less infill gives less time to cooldown the printed pla, and with so little infill the perimeters might not have enough ‘grip’ with the inside and allow the overhang area to curve more, and more because the print doesn’t cool good enough layer after layer.

Umm where's the edit button... Well anyway.. Now I had time to read the question again XD Leak, you mean 'drip' right? Mark2 drips less, because one head parks while the other prints, except when the second head is in use... But well, using Cura auto-temp-magic this is more or less not an issue.

It shouldn’t leak at all if properly assembled. Post photos so we can debug the issue

I been thinking lately about this bed problems with Cura since now I actually use cura for my bcn3dsigma, and my 2cents about this issue is that the origin (apart of bed glass issues or even if the hairspray is applied) is the crossed lines infill. Cura default cross paths infill does scratch/neosanding/Ironing every time layer after layer, pushing the print layer after layer, making the print adhesion a bit problematic if isn’t rock solid (more than needed imo). What I mean is that if the infill is done on both directions at the same layer there’s material being scratched when the infill lines cross. I’m too sleepy now but I hope I explained myself right ?

If anyone wants to 3dprint talk while closed I’m always ON at http://twitter.com/neotko Anyone else twitter account to talk around?

This was the problem I was having Same spool Same printer Same day Removing the TL-Smooth fixed the issue. Ofc maybe it was that the filament had trouble to be pushed by my feeder, but clearly shows that it lowers the effective torque.

3.2mm

What firmware are you using with the duet3d boards? Info about their firmware https://duet3d.com/wiki/Firmware_differences_from_Marlin,_Repetier_etc. Isn’t marlin, also all firmware doesn’t need to be ‘flashed’ to change any jey setting, some need a reboot but everything is gcode defined. Very handy

Thanks @conny_g no need. Already ordered. Anyhow will be probably my last test of this. January I will order my third duet3d and finally say goodbye to this old hardware, also will change all the printers to gudoxycore

Indeed, also I don’t think that UM has much spirit towards letting people change what they define (and of that I’m sure) Adjusting this thing could be easily done by the user if access to the machine jsons wasn’t limited to ssh or aigned firmware updates. If changing the machine json would be accesible the chances of getting easy updates from the community to improve little things would be faster One could rearrange the prime areas Change steps of the motors Play with the hardware to make modifications (like 1.75) But atm the risk to brick the printer is very high when changing the machine json that controls all that information. IMO the biggest issue the um3 has. One typo on the json comas or dots and you get a white brick that doesn’t boot up.

Yeah 300€ for outdated tech isn’t worth Duet3D or even a smoothie is the way to go. Think of um3 boards, they use 300€ white old board + a20olimex that they sell for another wooping 300€ and you can’t even edit the steps without the risk of bricking the machine (a week ago a friend on twitter suffer that and lost a week of production just because the machine bricked and not even with the recovery sd was able to revive it) Simple and modern stuff without access restrictions is the way to go if you want to have a machine that will endure the pass of time (unless you have so much cash that you just toss them like cakes). Ok enough ranting. Will test the new tl-smooth in a few weeks. So dar my machines one has tlsmooth on x/y only and the other without anything. Christmas production can’t wait

A small update on the old (waiting the 'new') TL-Smooth stuff. I'm so freaking busy to report and post photos but basically using this affects how much effective torque the motors can do, and it DOES affect X/Y motors at speeds of 50-70. Also it affects specially the effective torque of the Extruder motor. So if after adding this you start to see top layers with micro gaps or less extruder power, try removing the Extruder, and if persist remove the X/Y ones. After 5 days of debugging one printer why it wasn't able to deliver quality, changing coupler (that was dead) and opening a brand new spool, measuring, etcetcetc, many wasted hours, even changed the nozzle from the ruby to e3d to remove variables, in the end the fix was just to remove the X/Y/Z/E TL-Smooths I think that removing just E would suffice, but also I'm so busy with Christmas productions that I can't write much. So, if something changes, try removing the stuff and try again.

Even if you had a battery backup, you could still run out the power, so, no matter the setup you could lose power to the printer. I feel this should be something to aspire to as well. Duet3D have something for that, If detects a down on Below 19V saves status and parks the head. Ofc it needs tuning (since duet3d is a one for all solution for many kind of printers). But is doable.

Mándale un msg a @bondtech responden muy rápido