0.9 degrees Steppers - User Experience

Hello

I know 0.9° steppers have been brought up from time to time in the past. What I would like to do is to build a high accuracy RepRap based on the Prusa i3. The fact that the 0.9° steppers are twice as accurate as 1.8° stepper motors in terms of step precision is intriguing when repeatability is one of the main design goals of my printer.
Unfortunately, information and vendors are not as abundand as for the 1.8° versions, so I would like to ask anyone who has used the model 42BYGHM810 by Wantai Motors whether the torque is still high enough for higher rpms because the website lacks the pulse-torque characteristics diagram.

Also, I have read in another thread that for layer heights of <0.15 mm, a 0.9° motor used in the extruder would provide better results. Can anyone confirm this?

I use 42BYGHM809's on all my machines. I drive them @1/8 instead of 1/16. No mechanical issues so far (user-generated ones for sure, mostly dumbass moves on my part).

Not sure you want to drive 0.9's @ 1/16 to increase accuracy - it may lead to a loss of precision (lost steps), but I cannot verify this as I haven't attempted it.

But IMHO 0.9's are the way to go - mechanical accuracy >software accuracy.

As far as the extruder goes I suppose it would help for superfine moves, but with the gearing associated with the extruder I don't think it's a big deal. Maybe I'm wrong here.

---

- akhlut

Just remember - Iterate, Iterate, Iterate!

[myhomelessmind.blogspot.com]

Thank you very much!
Now that I take a second look at the motors, your choice makes much more sense.
I think we are on the same page here as I would drive them with 1/8 steps and add a pulley system to double available torque and resolution. Maybe this will already be overkill for the arduino because the firmware allegedly can generate 30000 pps. With this high a motor resolution, only slow movements may be possible.

Unfortunately, I cannot seem to source this motor from reliable suppliers in Europe (fortunately, Sparkfun stocks them, though). Well, additional shipping and tax, here I come!

akhlut Wrote:
-------------------------------------------------------
>
> Not sure you want to drive 0.9's @ 1/16 to
> increase accuracy - it may lead to a loss of
> precision (lost steps), but I cannot verify this
> as I haven't attempted it.
>



can you elaborate that?, microstepping vs lost steps?

Quote

Not sure you want to drive 0.9's @ 1/16 to increase accuracy - it may lead to a loss of precision (lost steps)

For a given speed, the time between reaching full steps is always the same, regardless of the microstepping in between. And only full steps can be lost. So, if microstepping looses steps, full stepping does, too.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

Quote
For a given speed, the time between reaching full steps is always the same, regardless of the microstepping in between. And only full steps can be lost. So, if microstepping looses steps, full stepping does, too.

Yeah, I someone I know also mentioned that the microsteps had a tendency to deviate towards the full steps under load. This of course may lead to irregular results and higher tolerances.
Maybe this problem might be mitigated by designing higher torque drives. Those could be more powerful motors or a pulley system (which I favor because of the compactness and lower weight).

If accuracy is more of a concern, maybe it would be better to gear down your motors instead? You'd get greater accuracy and higher torque at the cost of speed.

Gearing down would induce backlash from all those gears and add a lot of size. The phidgets geared stepper motors have backlash of 1°.
What I am striving for is not the absolute highest possible resolution, but positional accuracy. And for that, the motors should be as precise as possible. After that, I will add a pulley system similar to this: youtube.com
The advantage of that being cheap and easy to make while doubling the torque and resolution. The current arduino based hardware and firmware combination allegedly can't put out higher step rates than 30kHz, so this system would be about the maximum I could build without sacrificing too much print speed (yeah, I also thought about double pulleys and whatnot).
If I were to try an ARM based higher powered board, I could theoretically gear stuff down via pulleys considerably.

The newer design flips around belt so I can use bearings instead pulleys on the carriage. It also makes it easier to tie the ends down. This video shows it off probably better than others. [www.youtube.com]

That is also like how I want to build mine.
Is there a reason why you used belt for your Z axis? It's quite the unusual design, but looks like it could be more compact than leadscrews.
Your printer sounds quite distinct and nice, by the way

Decent leadscrews are expensive and using Z belts keeps the cost down. Leadscrew coupling is also not as trivial as it might seem.

That's true. I am currently designing my printer and the Z leadscrews really give me quite some headaches. They don't really want to fit anywhere because they have to be centered above the stepper shaft, thus taking away space. And because I want to "do it right", a lot of complexity is added by support system.
The thing I like about your design is that in case of a catastrophic failure, the printbed has the chance to sag away. Unfortunately, the opposite is true for a X axis coupled to Z like all the current standard designs, which, when considering using a little mill on a sturdy printer only exarcerbates the potential dangers of this configuration.