UM original with direct extruder

[foehnsturm 970]

After several months without any 3D printing at all I finally found the time to fix some issues with my latest project. So this is my modular printhead, now fitted with a direct extruder: a NEMA 8 stepper with a 1:40 worm gear, MK8 drive gear, 0.55 amp stepper current, microstepping set to 2, 760 steps/e.

As expected, retraction works like a charm. Torque hasn't been an issue so far but the weakness of several idler arm designs.

Right now it's reliable running with approx 70 mm/s @ 0.2 mm layer height. I didn't try to push it further as i'm printing some christmas gifts.

Please don't look at the wiring :oops:

 

 

[rigs 735]

very interesting

 

[ultiarjan 1,223]

Thats a nice and compact design!

Does it have the option to adjust the sping pressure?

I assume it's 1 bolt for removal of the bearing arm, for cleaning ?

Would like to see result on flexible fillament ...

Can you just directly use this Nema8 on an UM / UM2 board?

offtopic;

I was planning on using a nema 14 for another project on my UM2, but I'm not sure how to know if I can safely use a stepper motor. It seems the UM2 motor is a SY42STH47-1684MA (2.8 Volt , Current 1.68 A) Is my assumption correct that is has to be one with the same voltage, and same or lower current ? what about the other specs like resistance per phase etc ... basicaly my question is how do i know if I can safely use an other stepper.

 

[foehnsturm 970]

Hi UltiArjan,

No spring pressure adjustment so far. But that shouldn't be a big issue,

Yes, it's one bolt for the bearing arm.

For the steppers: As far as I know, voltage isn't important but current. Just adjust the stepper driver to the desired current (which should be equal or lower than the stock setting). http://reprap.org/wiki/Pololu_stepper_driver_board. Most steppers can be operated in bipolar (4 wire) mode.

First print with Formfutura FlexiFil. There is still some room for improvement (layer adhesion).

 

[foehnsturm 970]

I did some extrusion tests. 5.8 mm3/s is the save limit at the moment. Which is very much ok for most prints.

Above 6 mm3/s filament starts to slip. Increasing spring pressure helps but soon the stepper starts skipping steps. Right now it's a NEMA 8 with 1.8 Ncm holding torque.

I'll change to a NEMA 8 with 3.6 Ncm. If everything works as intended, it should be possible to get a direct driven extruder capable of 10 mm3/s with only approx. 100 grams additional weight.

 

[Dim3nsioneer 558]

Interesting. Can you say something about shadowing?

 

[foehnsturm 970]

I can't see a big difference so far.

In theory it should be worse than a UMO or UM2 but way better than direct extruders with bigger steppers.

BTW, can anyone post the weight of a UMO or UM2 print head?

 

[foehnsturm 970]

I did some test prints today. https://www.youmagine.com/designs/make-2015-3d-printer-shoot-out-test-geometries

Extrusion temp seems to be a little high but for a first try I'm quite satisfied with this.

Sliced with Cura 15.01.

layer height: 0.1 mm (bridge 0.15 mm)

speed: inner shell 60 mm/s, outer shell 40 mm/s (except resonance test), top/bottom 30 mm/s

I had to learn about max revs/min of steppers. With a 1:40 worm gear 45 mm/s would transform to 5400 revs/min for the extrusion stepper ....

Settings are now: 0.3 mm distance @ 5 mm/s

 

[nick-foley 5]

Loving this upgrade. Have you considered switching it to 1.75mm filament to change the balance of feeder torque vs. filament speed?

Is there a BOM or project page on Youmagine if I wanted to try setting this up on a machine?

 

[meduza 191]

1,75mm would require even higher rpm on the motor and he already has problems with the high rpm.

Me and a friend (johank on the forum) were discussing a similar solution a few months ago, but we were looking into brushless rc motors for the speed and power, mounting an encoder to the motor and drive it with real servo feedback, that would give us precision enough, but more speed and power.

For the same weight as a high torque Nema8 stepper (90g) you can easily fit one of these and a encoder: http://www.hobbyking.com/hobbyking/store/__25556__AX_2810Q_750KV_Brushless_Quadcopter_Motor.html%20and%20it%20should%20have%20the%20same%20or%20more%20torque,%20but%20over%20a%20wide%20range%20of%20rpm%20instead%20of%20the%20steppers%20rapidly%20declining%20torque/rpm%20curve.

 

[aviphysics 9]

1,75mm would require even higher rpm on the motor and he already has problems with the high rpm.

Me and a friend (johank on the forum) were discussing a similar solution a few months ago, but we were looking into brushless rc motors for the speed and power, mounting an encoder to the motor and drive it with real servo feedback, that would give us precision enough, but more speed and power.

For the same weight as a high torque Nema8 stepper (90g) you can easily fit one of these and a encoder: http://www.hobbyking.com/hobbyking/store/__25556__AX_2810Q_750KV_Brushless_Quadcopter_Motor.html and it should have the same or more torque, but over a wide range of rpm instead of the steppers rapidly declining torque/rpm curve.

 

I wonder how the power curve through a cycle of rotation compares to a stepper motor. The other potential issue is that a normal brushless motor won't have much holding torque. In this case, the holing torque might mostly affect its ability to quickly come to a stop at a specific point in its rotation.

Also, brushless RC motors are designed to operate most efficiently at relative high RPM's. They might not work as well at lower speeds.

I think you could do it, but I think it would be pretty non-trivial and you might not be able to use a normal off the shelf RC brushless motor.

 

[foehnsturm 970]

I'll upgrade to a NEMA8 with 300gcm and 80g weight (SY20STH42-0804A).

 

If I remember correctly, steps/e on the UM2 is set to 282. So one might consider to operate the NEMA8 in full stepping mode which gives 380 steps/e. Additionally, as I read somewhere, half stepping gives the lowest torque output (but differences are within 5-10%).

 

At the moment, a BOM or detailed project description is a little complicated: I machined down the stepper drive shaft from 4mm to 3mm, removed half of the MK8 drive gear and glued it etc. There is really not much room as you can see.

 

For the next iteration I'll use this worm gear http://www.ebay.de/itm/Schneckenrad-Schneckengetriebe-Schnecke-Modul-0-5-Messing-Stahl-/191027866545?_trksid=p2054897.l4275 (German)

 

 

[kvones 5]

Very interesting! I am looking to build a similar setup with a syringe and a non-captive linear stepper motor for my UMO+ to do paste extrusion. I am struggling a bit to figure out exactly what electrical specifications (current, Voltage etc) my stepper needs to adhere to in order to plug directly into the UMO+ electronics and work (I think they are the same as UM2's). Could you shed some light on this for me? Also, how much does your direct extrusion print setup weigh?

 

[foehnsturm 970]

I just finished the first larger print with the new setup with the stronger NEMA 8.

3 hrs, ColorFabb XT @ 240 C, 0.15 mm layer, 40 to 70 mm/s

Close to perfect results.

The additional weight is almost exactly 100 gr ( stepper 80 gr, worm gear 10 gr, drive gear and bearings another 10 gr)

Don't know about the UM2 or UM+ electronics. But the only thing to adjust should be the stepper current (HW on the UMO, SW on the UM2 and UM+ I think).

 

[foehnsturm 970]

I printed my extrusion test cylinder again, ColorFabb PLA @ 206°C:

Perfect extrusion up to 9.6 mm3/s (120mm/s * 0.4 nozzle * 0.2 layer)

My conclusion so far: a worm geared NEMA8 direct extruder is strong enough.

The additional 100 gr weight is partly compensated with the absence of fans or heat sinks on my print head. The head still remains substantial heavier than the standard head though. However, the possibility to use any filament and the absence of any strings may be of more importance.

.. yes, the motor is intentionally tilted :wink:

 

[foehnsturm 970]

.. works very well with FormFutura FlexiFil:

0.8 mm shell, 0.15 mm layer, no infill, 40 mm/s, retraction 0.4 mm @ 5 mm/s, 215 °C

 

[drayson 75]

Impressive...

Do you see any need of fans on your setup or how is the cooling of the layers guarantied??

 

[foehnsturm 970]

I'm still using the http://umforum.ultimaker.com/index.php?/topic/3890-the-crossflow-fan-approach/?p=30643.

 

[drayson 75]

Ah - ok, sure... :-)

 

[jonnybischof 60]

Very nice!

I'll have to give this a try, too... Want to build up a smaller printer for flexible filaments.

The red XT part looks very nice - the added weight on the printhead doesn't seem to have much of an impact on print quality..

How's the motor sound at 1/2 microstepping?

 

[foehnsturm 970]

How's the motor sound at 1/2 microstepping?

 

... ugly, I have to admit. But as it's a small motor it's not very loud.

 

[foehnsturm 970]

FlexiFil and his friend Lay-Fomm who has to sit in water for the next days ...

 

[restinpieces 3]

Interesting! How would you compare your direct drive extruder to your flexdrive extruder?

 

[foehnsturm 970]

I still like the flexdrive because it realizes direct extrusion with almost no additional weight.

However, I had some extrusion issues with short paths and frequent retractions in between. Perhaps some torsional flex, perhaps my setup wasn't dialed in perfectly. It would be very interesting to hear some other flexdrive experience (Team Ultimaker ?!).

Compared to this, the NEMA8 direct drive gives you absolutely on spot extrusion control (retraction distances < 0.5mm). It's even possible to print flexible filaments with retraction on, because the tiny movement doesn't cause any unwanted filament jam.

 

[drayson 75]

great!!

what are the current dimensions of the stepper itself?

would be interesting if there is a possibility to get it even smaller... with same torque...