Welcome to the SeeMeCNC Community

I just want to throw this out there given my experience in the last day...

Firstly, I've printed 1000s of hours and KMs of filament on my Rostock in the last year with absolutely NO SKIPPING. Yesterday I started working with some new Proto Pasta Carbon Fiber PLA to print a reel for an article going in Field & Stream magazine. I performed my usual "new filament tuneup" to find the best temp and other settings and then printed. The first two parts printed were perfect, I really like this filament. Smooth as silk. I then started my 3rd part - the larger diameter reel back plate. While printing the first layer it started filament starving in very short segments, I didn't hear the tell-tale extruder clicking. I restarted and watched carefully. Same thing in a different location. Again a restart, increasing the temperature and then I noticed that at the slow print speed I use for my first layer, the extruder is turning very slowly and I could almost see the microstep positions.

For all the benefit of micro stepping, the one negative is they reduce holding torque at the intermediate microstep positions. I realized that the first 2 parts printed a bit faster so I ramped up the speed some and tried the 3rd part. This time it printed fine. I further tested by slowing and increasing the speed and discovered that temperature doesn't have a significant effect but speed does - faster is better! Counter intuitive but when I started thinking about it, at higher speed, the extruder is revolving faster and has more momentum. That just might be enough to counter the "micro skipping' phenomenon I think I'm observing.

The 3D print forums are filled with comments about direct drive extruders. In fact, one of the developers of smoothie was getting on my case for using a direct drive extruder with 92.4 steps/mm (I'm actually running mine with 32 microsteps so I have 184.8, but still low). He said that smoothie was probably not going to work well for anything less than about 400 steps/mm which had been in the back of my mind when I came across this problem. The CF PLA is a bit more viscous than normal PLA and requires more worse to extrude. I can tell the difference pushing it by hand.

It makes me wonder what would happen with a geared extruder. So, I am going to find a suitable one to test. I'm curious if anyone else has seen this, maybe without realizing what was going on?

cheers,
Michael

What nozzle are you using? Just a regular brass one? Ive been reading around and I know Colorfabb and pretty sure ProtoPasta both recommend using SS nozzles with the metal and CF filaments. Nozzle wear and all. That was just a little concern I had.

I saw this on my Rostock Mini when I was using 3mm filament. I never dared go below .2mm layers because I thought I'd get funky extrusion issues. I haven't found a printed geared extruder design I've liked for 1.75 filament. There was a fully printable planetary one Ive seen around that looks pretty decent though. http://www.thingiverse.com/thing:39138 I haven't tried it out for size but it looks pretty cool. He made it because his motor was softening the filament, but it has nearly a 5:1 reduction, so it should fare pretty well.

Ah, I should have said... I run a Kraken water cooled hotend with .4mm nozzles that I bored to get a .6mm bore length. They are brass and I know that SS is preferred. I am making some now but needed to get this reel out. I'm sure a few parts won't wear the brass enough to be a problem.

That's a cool extruder. I'll have to take a closer look at it.

cheers,
Michael

I've had a conspiracy theory about this for a while too. Phil from the MakerHive used to print things with a .35 nozzle at .07mm layer heights, and when he switched from Steve's original extruder to the EZStruder, you could barely tell that it was moving.

I believe that there is definitely an extrusion resolution limit that you approach with smaller extrusions. This is where Brian from Trick Laser has experimented with putting a geared stepper on direct drive extruders.

[img]https://scontent-dfw.xx.fbcdn.net/hphot ... e=55502E64[/img]

Yes, I have found the Kysan 5.2:1 geared stepper works better than direct drive ezStruder. I do more for the added torque, but the increased resolution may help at slower feeds.

This week I am in the process of building a mount to put a 19:1 geared nema 11 motor right on the platform.
That little guy has 208 ozin of torque so it should be plenty of power.
Talking with John Oly last night he was concerned about retract speed, but I don't think it will be a issue.

So the 19:1 gearing will give even higher steps/mm ...should be a good thing?
Might even try running it in full step mode.

bvandiepenbos wrote:Yes, I have found the Kysan 5.2:1 geared stepper works better than direct drive ezStruder. I do more for the added torque, but the increased resolution may help at slower feeds.

This week I am in the process of building a mount to put a 19:1 geared nema 11 motor right on the platform.
That little guy has 208 ozin of torque so it should be plenty of power.
Talking with John Oly last night he was concerned about retract speed, but I don't think it will be a issue.

So the 19:1 gearing will give even higher steps/mm ...should be a good thing?
Might even try running it in full step mode.

I have considered the same but to do a 50mm/s retraction, that little motor is going to be wizzing away at 1662RPM with the standard filament gear where 5:1 its only at 433rpm and direct drive 86rpm The 14:1 option brings you in at 1212rpm.

I'm not sure what the max speeds are of these motors - but I'm guessing you're looking at the same ones I am - they don't list torque vs speed or any important information like that.

I ALWAYS retract PLA slow - 20-25mms and then prime fast (now supported directly in KISS) at 50-55mm/s. PLA is thixotropic and lots of fast retract cycles will lead to jamming.

mhackney wrote:I ALWAYS retract PLA slow - 20-25mms and then prime fast (now supported directly in KISS) at 50-55mm/s. PLA is thixotropic and lots of fast retract cycles will lead to jamming.

yep, on your advice I started doing 20mm or even 15mm retracts with PLA with much better results.
(instead of 60 to 75)
Thank for the tip Michael.

I have not tried it with ABS yet, but think slower won't hurt either. 2 years ago I was using 40 for ABS with OK results.

I'm definitely gonna pick up a couple 5:1 steppers and test, as I've been reading a lot about the testing of other users here. Also, putting two 40mm 12v fans pointed at the sides of the stepper on the front side, near the extruder, helps with similar issues I had been having. There was a TON of heat build up, which could also lead to skipped steps (I presume).

I've also recently slowed down my retracts as others have mentioned. I'm using 30 or 35mm/s.

Is the Kysan 5.2:1 geared stepper a drop in replacement on the ezStruder? I assume that would give 5*92.4 = 462 steps per mm? That seems to be right in the sweet spot.

I'd like to know more about this as well. If it's a drop-in and painless (reasonably) upgrade, I'd be up for trying it.

g.

It looks like an adapter has to be made since the geared side has a smaller mounting hole diameter. Not a big deal, I could wipe up a few of these in aluminum. I am thinking about picking up 5 of these for my set of ezStruders. I found some posts on other forums that really advocate using these geared steppers on direct drive extruders.

I've been on the verge of buying a 0.9 degree stepper for ease of installation. Stock hobbed gear would still work and motor is direct fit. Shaft on the output of the geared units is 8mm...

Won't be as much resolution as the geared stepper, but still twice that of stock.

Yes, I thought about that too. I am intrigued by the higher torque of a 5:1 stepper though.

Adapter plate and 8mm bore drive roller

A little more evidence...

I've started printing more in ColorFAB's copperFILL PLA. I've pretty much dialed in the settings (in KISS) and get pretty good results. But I've noticed that every so often (and usually a couple of times on each layer) I'll get a streak from over extrusion. Even tweaking the flow multiply down in increments to isolate them did not really work. Even with reasonably starved layers, I would still get the over extrusion streaks.

So, I pinched the filament between my fingers as it enters the ezStruder to apply a little resistance. Two things are interesting about this: 1) I can feel the pulses from the extruder and 2) the layer with filament resistance went down with NO over extrusion tweaks. It was like night and day.

I think the more viscous copperFILL highlights this issue where normal PLA just expands/contracts to absorb most of the flow unevenness. But here are two examples of potential flow unevenness due to a low number of steps per mm with a direct drive extruder. I am absolutely going to try the 5:1 geared Kysan stepper on my ezStruder. It might take a couple of weeks to get it in but I'll certainly post back when I do. Meanwhile I am going to make a little clip to apply some tension to the copperFILL so I don't have to sit here and pinch it during a print!

I have an extra EZstruder to set up with a geared motor. I found this one on Amazon: http://www.amazon.com/Gear-Ratio-Planet ... ed+stepper
Does it looks like the right specs for this application? If there are other options, I'd appreciate a heads up and a link

That is this motor: http://www.omc-stepperonline.com/gear-r ... p-140.html

Its from china so keep that in mind when you look at the price, the one from Amazon is from a seller in California who only carries a little bit of what the omc stepperonline store has.

http://www.thingiverse.com/thing:20884
https://www.youmagine.com/designs/diffe ... ry-gearbox

Batteau62, that is a .4A rated stepper. I am planning to get the 1A Kysan: http://www.ebay.com/itm/Kysan-5-18-1-Pl ... 2a3e5f325d

But with the gearing down, .4A should be ok.

Nylocke, I like that thingiverse geared stepper - about 4.4:1 is right in the sweet spot. I am going to print one up and see how it does.

While you're pinching, use a piece of foam or something. I run the filament on both printers through a piece of foam (without pinching) and get an amazing load of fuzz and stuff caught on the foam. Even from a spool fresh out of the bag.

bvandiepenbos wrote:Yes, I have found the Kysan 5.2:1 geared stepper works better than direct drive ezStruder. I do more for the added torque, but the increased resolution may help at slower feeds.

This week I am in the process of building a mount to put a 19:1 geared nema 11 motor right on the platform.
That little guy has 208 ozin of torque so it should be plenty of power.
Talking with John Oly last night he was concerned about retract speed, but I don't think it will be a issue.

So the 19:1 gearing will give even higher steps/mm ...should be a good thing?
Might even try running it in full step mode.

Are you going to offer the aluminum EZstruder at a future date or tell us if you had a source, preferably one that fits the geared stepper.

Nylock and all, I printed out the stepper gearbox on Thingiverse. The print came out great, all I had to do was ream the gear bores to 6.3mm and they spin nicely. But, the problem is, it is a PIA to mount this gearbox to anything! Take a look at the design, it uses four 35mm M3 screws to attach it to the stepper. In order to mount to the ezStruder or anything else for that matter, these must be replace with longer screws. The 40mm I have in my stock are two short. Plus, keeping everything together and aligned while you screw the stepper, gearbox and ezStruder together is a bit tricky. I went ahead and ordered the Kysan. It will probably arrive before I can sort out mounting the printed one.

cheers,
Michael

I wasn't quite sure what would have to been done about that. It looked like a slight possibility. I'm going to go the http://www.thingiverse.com/thing:39138 route, maybe make a derivative that works for mounting it like an EZstruder. I might even make a complete remix that blends that with my EZflexstruder design.