Getting the most from your ezStruder

[JFettig]

I need a reality check, I'm picking my motors for my next printer and I am sizing up some geared stepper motors, torque isn't much issue, it'll be more than the stock kysan most use 1:1, but more rpm. It seems that most are shooting for 5:1.

I did some quick tests on extruder speed with ABS, it seemed like the most I could get out of it at 240C is 5mm/s, which is 300mm/min, or about 8.5RPM at the motor, 5:1 is only 43RPM and 14:1 is at 121RPM. I know torque drops as speed goes up, but not at that extreme.

As a reality check - does 8RPM and 5mm/s filament speed sound right? If so, I might order a smaller(physical size) motor and 25:1 gearbox setup so that maybe I can do direct drive right on the platform.

The motors I'm looking at are from this page:
http://www.omc-stepperonline.com/geared ... s-c-4.html

I think what I neglected to think about is if I wanted to have the ability to retract up to 50mm/s a 5:1 gearbox will put me at 433rpm and 14:1 at 1212rpm

[KAS]

Just a wild shot in the dark, but what about moving to a larger stepper like a nema23?


Now, I'll be honest in my hopes you say " hell ya" because I have 4 of those and geckodrive controllers collecting dust...



[img]http://www.geckodrive.com/media/catalog ... ng_stc.jpg[/img]

[Polygonhell]

Moving to a nema23 wouldn't really help unless you use a larger driver.

[KAS]

Moving to a nema23 wouldn't really help unless you use a larger driver.

/waveshand someone design something for this... /thread back on topic, sorry

all joking aside, I was using this to build a CNC machine but lost interest and picked up other hobbies. I might go ahead and finish it someday.

[Jimustanguitar]

I ended up using Blue tape to take up the slack in the aluminum piece. So far it's working well.

[img]http://i.imgur.com/LaipezK.png[/img]

[626Pilot]

In the past, when a PTC connector has failed, I've sanded the top of the plastic lip and the end of some PTFE, then put a coating of Superglue on the last couple inches of the PTFE before fitting it to the PTC. I put some filament through to act as "floss" in case any Superglue should leak to that area. After letting the glue cure, I'd put on a second coating, making sure to get it all over the PTFE and plastic lip. That setup lasted for a good long while, hundreds of hours. I think it helps print quality a little because it doesn't allow the tube to move at all during prime/retract.

That setup finally broke down a few days ago. I had some of the plastic "BSPP" fittings that they ship with E3D hot ends, so I took out the machined aluminum adapter (like in the above post) and Superglued the E3D fitting in place. The fitting had a little rubber O-ring, which I took out in order to make sure it would seat properly. It's necessary to take a little care when lining it up, but it's not rocket science. It will sit pretty cleanly on the plastic part of the EZ-Struder mount, with some space left on the sides that's perfect for reinforcing with Superglue. The PTFE passes through all the way to the filament channel, and it's a lot easier to feed filament than it was with the old fitting. You can see it pretty clearly in this photo:

[img]http://thingiverse-production-new.s3.am ... atured.jpg[/img]

It would be nice if there was a PTC fitting that didn't have any play. When you prime or retract filament with these, the little plastic lip lets the whole tube move a little. The old Superglued PTC didn't let it move at all, because the plastic lip and PTC body were physically bonded with Superglue.

[Jimustanguitar]

I've seen designs on Thingiverse for a screw down collet style bowden connector. I've always meant to try it, but never have.

Is there an industrial compression fitting to try?

[Eaglezsoar]

Moving to a nema23 wouldn't really help unless you use a larger driver.

/waveshand someone design something for this... /thread back on topic, sorry

all joking aside, I was using this to build a CNC machine but lost interest and picked up other hobbies. I might go ahead and finish it someday.

photocqi.jpg

The parts you have will work fine on a 3d printer. The voltages and current for the drivers will be completely different so you would have to use a dedicated circuit for the motors.
You would also have to figure out how to mount the larger motors. You may want to sell all the parts pictured on ebay then pick up the normal 3D components.

[Eaglezsoar]

Moving to a nema23 wouldn't really help unless you use a larger driver.

/waveshand someone design something for this... /thread back on topic, sorry

all joking aside, I was using this to build a CNC machine but lost interest and picked up other hobbies. I might go ahead and finish it someday.

photocqi.jpg

The parts you have will work fine on a 3d printer. The voltages and current for the drivers will be completely different so you would have to use a dedicated circuit for the motors.
You would also have to figure out how to mount the larger motors. You may want to sell all the parts pictured on ebay then pick up the normal 3D components.

[bvandiepenbos]

@626Pilot, if you pull on the PTC lock ring as you push tube in and bottom tube out it eliminates most of the play. Then clip something under lock ring to prevent lock ring from moving.
Something like this http://repables.com/r/444/ but larger for the E3D fitting. I have made lock clips from a large paper clip or the lever from a binder clip.
You can get commercial made clips, but not sure size is right.

[Nylocke]

I'm using one meant for the Ultimaker PTC fittings resized for these. The Ultimaker ones are for 3mm filament, and they are the same ones that E3D sells.

[Jimustanguitar]

I've been thinking about doing something like this as a test.

[img]http://thingiverse-production.s3.amazon ... _large.jpg[/img]

http://www.thingiverse.com/thing:11864

[bvandiepenbos]

Here is a lock clip for E3D fitting that I drew, but not printed yet.

PTC_lock_clip-E3D.stl

(89.54 KiB) Downloaded 432 times

put it on repables.com also...
http://repables.com/r/479

[626Pilot]

@626Pilot, if you pull on the PTC lock ring as you push tube in and bottom tube out it eliminates most of the play. Then clip something under lock ring to prevent lock ring from moving.
Something like this http://repables.com/r/444/ but larger for the E3D fitting. I have made lock clips from a large paper clip or the lever from a binder clip.
You can get commercial made clips, but not sure size is right.

Dang, that is cool! I had been thinking of doing something like that, but never got around to it. I think pulling on the collet while connecting the PTFE will take out enough play, but I wonder if one of those clips isn't a better idea - it would take the strain off the collet, I think.

The only thing that bothers me about these fittings is the possibility of the collet squeezing the tube enough to add friction to the filament path. I think an ideal solution would be a collet that's really long, so it can apply less pressure over a much larger surface area. Or, perhaps a part that you fuse to the PTFE by holding it over a stove and then clamping it in place, so it can melt its way a fraction of a millimeter into the tubing without deforming the inside. Something designed specifically for 3D printing, rather than running dental drills!

[mhackney]

+1 626Pilot. These type of fittings can squeeze the PTFE. Perhaps a rubber cone would provide the necessary friction without distorting? Not sure where you cold get one but you could print them in Ninjaflex.

I was thinking of the "fused part" idea myself last night. There is special PFTE adhesive that would be just the ticket without heat. This part would be a simple collar around the tube, maybe 1 cm long. Conceptually, this would replace the cone in the Brian's drawing. This would allow easy removal, could be 100% printed, allow the PTFE to pass all the way up to the EZStruder and would not move in/out - and it would still allow disassembly.

I like it, I think I'll try something out. The base part could actually integrate the aluminum "donut" that comes with EZstruder and eliminate that part altogether.

[Jimustanguitar]

These type of fittings can squeeze the PTFE.

Very true. The forum member that was working on the triple hot-end mentioned that his threaded bowden design squeezed the tube enough that he had to ream it out afterwards.

[duvdev]

Hi all,
A small qestion, I want to add the geared nema 17 ro the extruder, do I have to make any cahange in the eprom settings of the max v2?
If so, what are the changes?
please advise - Thanks

[Jimustanguitar]

Hi all,
A small qestion, I want to add the geared nema 17 ro the extruder, do I have to make any cahange in the eprom settings of the max v2?
If so, what are the changes?
please advise - Thanks

You'll have to change your steps per mm for the extruder in your firmware.

[mhackney]

And the best way to do that is to measure it to get the most accurate results. The value will depend on your geared stepper ratio so it's best to measure your specific extruder anyway and it is easy to do. I think Gene's manual has a section on extruder calibration, if not, search here or the web and you'll find lots of info. Takes 5 minutes.

[Nylocke]

The procedure I generally follow is unhook the bowden tube from the hotend, snip the filament with diagonal cutters to be flush with the end of the bowden tube, extrude 100mm of filament, measure the value that it extruded, take 100 and divide it by the value you got, multiply that by your steps per mm, and plug that into your firmware for steps per mm. If your off by a significant amount the first time (5% or greater) do it a few times until you're happy with it. Then I set up different profiles in my slicing engine for all of my filament that I have with the average measured diameter of the filament. This works well for most rigid materials (ABS, PLA, Nylon, etc) but for flexibles I generally adjust the flow multiplier to get it right. That one is more of a trial and error.

[626Pilot]

The procedure I generally follow is unhook the bowden tube from the hotend, snip the filament with diagonal cutters to be flush with the end of the bowden tube, extrude 100mm of filament, measure the value that it extruded, take 100 and divide it by the value you got, multiply that by your steps per mm, and plug that into your firmware for steps per mm. If your off by a significant amount the first time (5% or greater) do it a few times until you're happy with it. Then I set up different profiles in my slicing engine for all of my filament that I have with the average measured diameter of the filament. This works well for most rigid materials (ABS, PLA, Nylon, etc) but for flexibles I generally adjust the flow multiplier to get it right. That one is more of a trial and error.

How do you find flexible filament with these printers? I've never tried any, but I understand it can be difficult with the Bowden tube.

[mhackney]

I've been printing it since it came out. Go slow (25mm/s or so), use minimal retract and it prints fine. But, I also have modified nozzles that reduce the back pressure so I am not sure if stock E3D nozzles would work as well.

[Nylocke]

I haven't done quite enough testing to have conclusive results. I have V6 nozzles that I bought specifically so I could have a shorter transition zone without drilling them by hand. I'd have to take another look at E3D's drawings and take some measurements of my own but if what sanjay has said in previous posts its true, the length should be around 1mm or less now.

The main use I've had for Ninjaflex has been a "composite" part of sorts I've been developing for a customer. He's planning on injection molding them in the future for mass production so I've moved away from the dual material design for simplicity sake. I would like to do some more testing with my extruder design, refine it a bit more maybe, but I'm all but certain your design for the extruders mounting on the arms (which I have fully adopted since you first posted files for) will help a significant amount with printing flexibles.

[JFettig]

Brian/others who're running the Kysan 5.18:1 geared stepper - what are your settings for acceleration and max extrusion speed? I've been messing with mine and I'm getting random missed steps on retracts. It seems like some retracts are going faster than others.

-edit- seems that smoothieboard is ignoring the max speed in the firmware, more experimentation needed

[mhackney]

I'm running 100,,/s max feed rate and 6500 mm/s^2 acceleration on mine. Still running Azteeg X3 Pro on this machine though.