Trying out 0.9 degree stepper for MK8 style direct drive extruder

[Jetguy]

Just figured I start a thread for the latest experiment that may turn into a mod folks might want to try.

 

Unlike X and Y axis motion, the extruder moves relatively slowly, requires high torque, and very specific ratings.

The typical Mk8 stuff being sold on Fleebay, Amazon, AliExpress, and others simple aren't the right motors.

they are often low inductance and higher than normal current ratings to achieve desired torque.

My honest opinion is that in 98% of all cases, the common stepper driver we all use should never exceed 1.5Amps of current for a number of thermal and electrical reasons.

Given many of these motors are 1.68 A to 2A or greater rated, they simply never reach rated torque.

 

By comparison, MakerBot, FlashForge, Wanhao, CTC, et al, use roughly 0.8A to 0.9A rated motors, reasonable inductance, and easily achieve the torque required in a given MK8-MK10 extruder.

 

Given the criteria of a motor between 0.8A and 1.5A rating, NEMA17 size, short body (30-34mm long) I found some 0.9 degree motors.

Now, just for another educational point, motors in general, and stepper motors specifically, require that the rotor slightly lag behind the magnetic field in order to generate torque.

That means there is always some inherit position error between where the motor should be and where it is under load. An extruder represents a fairly heavy but semi stable load as the motor tries to force filament into the extruder.

The problem is, 1.8 degrees is a fair movement of filament given the typical gear. Microstepping is balancing the field between the physical poles in the motor and therefore even less accurate in a high load situation like an extruder.

A 0.9 degree motor would have roughly 1/2 less potential positioning error VS a 1.8 degree motor.

Some folks use gears and the problem with gears is always the vibrational rotation aspect of a stepper beats the gears to death and results in noise and wear. There are tons of pictures of 5:1 geared steppers with worn pinion gears at the stepper motor all over the web. Again, an example of where a 0.9 degree motor is direct drive, no parts to wear, yet gives us the mechanical (magnetic) pole advantage in the very specific case of a direct drive extruder.

 

And of course, it seems as if it's no longer available. http://www.amazon.com/gp/product/B00W98OYE4?psc=1&redirect=true&ref_=oh_aui_detailpage_o00_s00

 

* Manufacturer Part Number: 17HM15-0904S
* Motor Type: Bipolar Stepper
* Step Angle: 0.9°
* Holding Torque: 36Ncm(50oz.in)
* Rated Current/phase: 0.9A
* Phase Resistance: 6.0ohms
* Recommended Voltage: 12-24V
* Inductance : 12.0mH±20%(1KHz)

Physical Specification
* Frame Size: 42 x 42mm
* Body Length: 40mm
* Shaft Diameter: Φ5mm
* Shaft Length: 22mm
* D-cut Length: 15mm
* Number of Leads: 4
* Lead Length: 300mm
* Weight : 280g 

 

[Ryan Carlyle]

In general, I think this is a really good idea. Just need to find the right motor. Did you plug it into my stepper simulator? 

[Ryan Carlyle]

I've also been wondering if 0.9 degree steppers would be a better option for regular XYZ motors, assuming you've got a controller that can handle the higher step rate anyway. Nobody really needs to go to the >400mm/s you can hit with typical 1.5A 1.8 degree steppers.

[Jetguy]

No, I should have.

I really need to build a http://softsolder.com/2011/08/25/stepper-dynamometer-mechanics/

Given this pile o-drivers I have and various motors, start making cool graphs and get a lot more scientific.

 

Compare ideal VS reality.

[Ryan Carlyle]

I have a simple little shaft & frame coupling model that I've been using to get back-emf measurements. Also bought a bunch of low-resistance, high-wattage resistors to use for braking and power measurement. Haven't had a chance to set it up though. 

[Ryan Carlyle]

Hmm, this might work if it's wired bipolar instead of unipolar. http://www.kysanelectronics.com/Products/Detail.php?recordID=3171

[Mark Napier]

Hey Jetguy,

How did this stepper work out for you?  Did it provide a more even flow?

OT, but I'm considering trying out a Printrbot extruder.  I like it because of the two drive gears and it provides a good guide path for the filament below the gears into the hot end.  Also, since it is an aluminum body I can attach a heat-sink/fan on the front and use the body as the cold block.  A round piece of aluminum clamped in the J mount will be tapped for the heat break (the Prometheus heat-break/nozzle would work here).  That will conduct the heat away from the break and into the body.  Only other thing is an angle bracket between the stepper and the extruder to hold an overhead fitting for the PTFE feeder/guide tube.  End result is a really grippy MK8-ish style extruder that can also work with rubbery filaments.

Cheers,

Mark Napier

[Jetguy]

No, I found these motors did not provide the same torque and could skip steps making the whole point of finer resolution  a wash.
It was an interesting test but did not provide the final results we really wanted to see.

The problem is, short of sourcing very specific motors, this off the shelf motor did not work. It was marginal, and if that's me and all my experience and it was marginal, then I hate to see what happens in a less than ideal setup.

[Mark Napier]

If you don't mind me asking, do you have a good recommend for a direct drive stepper? I know its a balance between weight and torque. If it matters I'm doing a 24V system.

Thanks,

Mark Napier

[Ryan Carlyle]

I still use the standard Makerbot/clone motor for most of my direct-drive extruders -- Moons 17HD4063-xxN. 

0.84A

0.23 N-m

5.75 ohm

9.3 mH

38 g-cm^2 rotor inertia

I would only use this for direct-drive 1.75mm extruders, or fairly light-weight motion stages, around parts that can withstand heat so you can run it at the rated current. It's a good amount of torque to stall the motor before stripping the drive gear grip, but still be able to push pretty well. 

[Jetguy]

Another good motor that is ever so slightly longer and the same basic rating is now used on the 5th gen Makerbots and also, I know that my Flash Forge Dreamer used them but they were unmarked on the dreamer.
Here is that motor taken right out of the 5th gen electronics and in my highly modified setup https://www.flickr.com/photos/90025904@N04/22155120414/in/album-72157660402602670/
http://3dprintersuperstore.com.au/products/stepper-motor-17hd4063-05n

Again, in the Dreamer, they were unmarked and in a Makerbot, they were labeled 17hd4063-05n
Because it's a slightly longer motor, it still has the same basic coil winding and current rating, but produces a tiny bit extra torque to prevent skipped extruder steps.

The older makerbot like the T-O-M and the Replicators previous to 5th gen used Moons Stepper Motor 17HD4063-01N
Again, what is funny is Flash Forge sells motors for XY and Z, but not the very specific and ever so slightly longer motor used for extruders as i found in my dreamer. Electrically they ohm out the same rating.

[Jetguy]

Here is that Flash Forge unmarked compared to a MakerBot standard 17HD4063-05N

Picture from the stock Dreamer album https://www.flickr.com/photos/90025904@N04/15244626920/in/album-72157648277675866/
But no markings!! https://www.flickr.com/photos/90025904@N04/15244623040/in/album-72157648277675866/

I know, maybe not the most useful info, but just telling you what I know and what I used.

[Mark Napier]

Thank you both for the information. The amazing thing is that these sell for more than the equivalent Oriental Motors steppers. I must be missing something.

[Ryan Carlyle]

They're customized to original purchaser spec. The two number digits in the -xxN suffix of the Moons part number represent the set of special options picked by the purchaser. The 17HD-4063 part is the standard motor code. 

When I asked Moons for a datasheet for the 17HD-4063-03N a while back, they were confused and asked where I had encountered the motor. It seems to be something they whipped up for Makerbot, but that's speculative.

[Mark Napier]

Where would you get a 17HD4063 stepper?  Jetguy's link is for one that works out to nearly $70 USD shipped.

Thank you much,

Mark Napier

On Tuesday, January 10, 2017 at 11:07:01 AM UTC-5, Ryan Carlyle wrote:

[Ryan Carlyle]

Hmm. Makerbot used to sell them for about $25 iirc. Looks like they stopped. FlashForge has them for $38, don't know what shipping will run. http://www.flashforge-usa.com/shop/flashforge-motor.html

It's probably time to find a good commodity replacement... shouldn't be THAT hard to match specs with something else out there. 

[JF Kansas]

There were some on 3dp2go for $10 or so. I didn't compare specs though. 

[Jetguy]

FYI, that specific motor is pretty close to identical specifications as the Flash Forge motor you were looking at $70 for.
https://www.3dp2go.com/nema17-stepper-motor-34mm-length-08a-p-242.html

But, we need to clear something up. That motor was chosen as the best match for the typical mightyboard driven printer running Sailfish. The logic here is that we know the board and drivers even though we have digipots, the system and resistors chosen results in roughly 0.86A current max even when the digipot value is maxed out. Basically, it's a matched set of motor, board and driver, and also fimrware and settings if we round it all out.

If you are building with different mainboard and drivers. You might be able to go a bit more current, but still very reasonable with this 1.2A rated motor that has ever slightly lower inductance and since it's longer, more torque to boot.
https://www.3dp2go.com/nema17-stepper-motor-40mm12a-nema17-17hs300120b-p-213.html

Basically, we are comparing 34mm 0.8A rated motor to a 1.2A rated 40mm long motor.

[Jetguy]

Also, that printer built for Jeff- that Core XY that was derived from the ones for Fred, I used those 0.8A rated motors from 3DP2GO https://www.3dp2go.com/nema17-stepper-motor-34mm-length-08a-p-242.html for X and Y, Z is a standard Kysan 1124090.

The issue is, you asked for extruder specific motor and I've not yet found a 40mm and 0.84A rated motor on the cheap. As I said, the one in his and the one I have in front of me were Dreamer extruder motors, and the store doesn't sell them and they aren't marked. I know the specs but have no good way of getting another one.

The other motor I would consider very good and used for a bit was from a 5th gen, and here is a picture from my 5th gen + of that motor.
MS17HD2P4100  basic rating is 1A and 40mm long. Nice motor and could be a really good match for TMC2100 drivers in stealthchop for an XY application (hint, what is basically in the Raise 3D N series for XY).
http://www.moonsindustries.com/Products/Steppermotor/StandardHBSteppermotor/MS17HDSeries/

[Jetguy]

Again, if I was building one here is what i would want for the extruder direct drive.

If using a mightyboard, I need a 0.8 to 0.9A rated NEMA17 with a body length of 40mm. That will get you right into the sweet spot of torque VS mass, and the proper ratings electrically.
If I was using a different control board, then a 0.8 to roughly a 1.5A motor could work, I try to go for say 1.2A -1.3A if i had no other choice. Again, 40mm I find is better than a 34mm long motor if you have the mounting space.
There is a mass difference as well, but i feel never skipping a step on an extruder pays off.

[Ryan Carlyle]

A few rules of thumb worth noting here: 

• A large fraction of, but not all, steppers are sold under an unrealistic "two coils on at rated current" torque value, which makes it really difficult to compare motors unless the datasheet specifies which version you're looking at. The "two coils on" torque is about 40% higher than you'll actually get in practice. Measured torque curves are preferable because they don't have this issue.
• For two motors with the same NEMA size, and comparable electrical specs (resistance/inductance), stepper torque at rated current is proportional to the length of the black stacked laminations section. Which basically means torque = weight.
  
• When you look at different motor sizes with about the same weight (eg pancake NEMA 17 vs single-stack NEMA 14) the bigger NEMA size will have more torque but also lower top speed.
• Motors that need more current to produce a particular amount of torque are able to run faster. 

Basically, your top-performing strategies for 1.75mm direct (non-bowden) extruder steppers are:

1. A very small motor with a high-ratio gearbox, like the PG-35L tin-can extruder sold by Ultibots, or maybe a ~20:1 NEMA 11. 
2. A pancake NEMA 17 with a low-ratio gearbox, like the 3:1 E3D titan.
   
3. A short-stack NEMA 17 in the 34-40mm length range. 
   

If you want to focus on smaller nozzles and super-precision prints, go for #1 or #2. Normal or big nozzles, go for #2 or #3. If you want to push print speed limits or power through extruder jams, use a dual-hob extruder with a low-ratio ~5:1 single-stack NEMA 17 (like a Bondtech). 

[Jetguy]

100% agree with Ryan and everything he said.

FYI, my N series Raise 3D printers use 40mm NEMA17s, 1A rated, but not the Bondtech QR gear reduced version that Ryan called out, the direct drive variant called the "mini" that is dual hobbed as well.
Here's the difference between those 2 products.
The QR is an amazing beast, but it's a motor + a heavy metal gearbox. It can be used with an E3D or other hotend direct drive or remote bowden. because of the geared solution, it has an 8mm shaft and thus a roughly 11mm hobbed gear set. Basically, the motor produces huge torque via the 5:1 ratio geabox, and then the hobbed gear is pretty much the same as the common FF and Makerbot Mk8 gear many are familiar with.

The mini is much, much, much lighter. It also costs a lot less than the QR since you are basically buying the $50 set of gears and using your own motor and printed housing or buy a kit with an SLS printed nylon housing.
It may not have the same torque as the geared variant, but used a smaller hobbed gear set that fits the 5mm shaft directly on the motor. This results in the same torque producing more force on the filament VS the typical larger MK8 style 11mm gear. This works out to 140 steps per mm @ 1/16th stepping, vs the typical MK8 gear on the same motor @1/16th stepping with roughly 95 steps/mm. Just trying to show the roughly 50% steps difference and thus that much more force on the filament and finer control.

[Mark Napier]

Thank you both again for the replies.

One of the links gives the option to buy the longer Flashforge stepper for the Dreamer extruder so I'm going to get one.

Since you mention the TMC2100, that's what I've got. Four of them for a RADDS shield. Didn't know at first how limited they are in thermal management so I'll use lower current motors (1 amp or less) and run in constant current "SpreadCycle" mode.

For my X, Y and Z axis I've got a few Vexta PK245-01BA steppers. These have been around for a long time. In bipolar mode they are rated at 0.85 amps.

New they are over $70 but can be found for less surplus or used on ebay. Just saw this link:
http://www.interinar.com/vexta-pk245-01ba.html

Nice thing is the torque and speed they generate on 24V.

http://catalog.orientalmotor.com/item/stepping-motors--1068/pk-series-stepping-motors/pk245-01ba
http://catalog.orientalmotor.com/browser?&plpver=11&prodid=3001048&itemid=62095&assetid=26295

So for instance I figure the max speed for the gantry at 150mm/sec. For a 20 tooth gear that equates to 225RPM. The torque is pretty flat there.

For the X and Y I'm also considering the 0.9 degree PK254M-01BA esp. since I have a pair of them.
http://catalog.orientalmotor.com/item/all-categories/pk-series-stepping-motors/pk245m-01ba?cid=1002&prodid=3001048&itemid=11997&backtoname=Item+%23+PX245M-01BA&pane=sb&bc=100%7C3001048%7C3001065x
http://catalog.orientalmotor.com/browser?&plpver=11&prodid=3001048&itemid=11997&assetid=26304

Only thing is that the torque falls off to 1/2 by 225RPM. Also, they specify these speeds using their clean damper. I have them but don't know what the extra spinning mass means for dynamic performance; going around sharp corners comes to mind. I've been toying with the idea of the Mechaduino driving these to recover for any lost steps and dynamic torque control.

One annoyance of the "A" suffix (which is what I find for cheap) is that it is 440 thread - more screws to keep track of.

I think the PK245 will also be fine for the Z axis. Pretty simple mount so if need be it can be replaced with a 1A NEMA 23 stepper.

I looked up the BondTech mini extruder. The gear set looks nice. They seem to be about the same diameter as the Printrbot gears but the driven gear has a much more compact bearing design. Closest thing I see for me would be the $100 kit for the Wanhao i3.
http://shop.bondtech.se/ec/upgrade-kits/bondtech-mini-for-wanhao-i3.html

I think I could make a cold block to bolt to the bottom and provide a mount. Not perfect as it does raise up the motor some. Wish they had a kit that was a replacement for the MK8/7 Makerbot extruder and made for the standard cold block. Also can't tell from the picture if there is a good way to mount a reverse Bowdan tube on top. Failing a custom kit I think hacking the Printrbot extruder may work out better.

Cheers,

Mark Napier

[Jetguy]

Where did you get your TMC2100s from? I got mine at Filastruder and can walk you through how I set mine up for silent mode to match the ones on my Raise 3D printers that used their own version.
Basically, there are a handful of tiny resistor jumpers to move. I know you can do it other ways, but this worked for me for plug in compatibility.
Again, stock, the filastruder ones are not in spread cycle. removing the circled resistors will set it for you. I know, a total pain but figured I'd share my configs.

[Mark Napier]

Hello,

I've got a pretty early version from Watterott electroncs.  Have the matching schematics (for several versions) and datasheet for the TMC2100.

So are you (and Raise 3D) running in "SpreadCycle" or "StealthChop"?  I know the StealthChop is quiet and low power but I thought that it had a problem with poor delivered torque and missed steps.

Cheers,

Mark Napier

[Jetguy]

It must be stealthchop because it's dead silent, but i'd have to go back and read the schematics and compare the jumpers.

[Mark Napier]

The Flashforge Dreamer extruder motor came in.

Exterior dimensions look like the unmarked one Jetguy showed but this one is labeled.

BOHONG STEPPING MOTOR

42HB40F08AB-03

DATE.2015. 05. 17

Measured length of the case is 1.55 inches.

Measured resistance is 2.8 ohms.

Measured weight is 290 grams.

When I google that part number I get a couple of hits for people reselling a Flashforge Dreamer extruder so I guess its right.

I'm surprised that a 0.8 amp motor would have a resistance that low.

Here it is next to a Vexta PK245 for reference.

On Thursday, January 12, 2017 at 10:41:03 AM UTC-5, Mark Napier wrote:

Thank you both again for the replies.

One of the links gives the option to buy the longer Flashforge stepper for the Dreamer extruder so I'm going to get one.

<snip>

[WifiGuru10]

Hey Jetguy, i just picked up a used Dreamer dual extruder assembly. Motors, hotends, fan. How much modification would be needed to fit it to my flashforge? Does it have the same hole spacing as the regular FF extruder x-axis carriage? Are the stock steppers ok for it?

[Jetguy]

No mods required, rail spacing , rods holes, mounting all generally match.