Delta: 0.9° vs 1.8° with 2:1 reduction?
-
I'm planning a delta build in an either hexagonal or rectangular frame / chamber which gives me some room for the motors. I'd like to use Nema17 steppers and I'm planning on having them offset with either a coupler or a timing belt to reduce the stress on the bearings.
With 48V solutions and low inductance 1.8 motors now available I'm trying to figure out if it were a downside or an upside to run 1.8° motors with either 2:1 or 3:1 reduction?
I'm running 0.9 deg motors on my current delta and can easily hit my desired speeds - I've seen videos on CoreXY builds where the 1.8 motors are capable of 3+ times more speed and acceleration than their 0.9 counterparts - hence my thinking 2:1 reduction would be the same and 3:1 might even be a nice gain in resolution and error on a delta build?
-
@bberger
Why not try out how far you can get with 48V/0.9° steppers? You need a reference, before you can quantify the gain of another motor setup.
Question is, how do you deal with a geared motor and backlash. You could use a closed loop belt/pulley reduction, but the extra weight/inertia of the pulley is a downside.I wouldn't trust videos of CoreXY gantries, if they don't show the speeds at 45° moves too.
-
@o_lampe I'm just spinning ideas in my head and wanted to check if they're complete nonsense or possibly have merit.
And yes, the idea would be timing belt + pulley reduction.
It's less about absolute speed for me, I'm actually in a happy place with my current 0.9° delta on 24V and cooling becomes an issue before I can outrun the motors. My gut however tells me that 1.5m long 6mm belts might be a limiting factor and I'm worried about motor bearing life when thinking about tightening 9mm or 12mm belts to spec.
So the idea behind my thinking is:
- does it make sense to use the additional grunt from 1.8° steppers to use a reduction as I feel the need to reduce the load on the bearings anyways
- does a system with 2 belts eventually help to smoothen out motion and void some of the resonances? From my understanding steppers are more prone to get caught up in resonant frequencies if driven at low speed. Which is true on a delta especially in the center of the build area - where also accuracy and error are the worst. A 3:1 reduction would give me (at least in theory?) a 33% improvement in resolution and 33% less positioning errors compared to a 0.9° stepper directly driving the carriages while still being on-par with the 0.9 steppers speed wise.
At least that's what I come up with comparing:
- LDO-42STH48-2504AC (very low inductance, 1.8deg steppers)
vs - Moon's MS17HA6P4200 (my current steppers)
Maybe there's a unicorn 0.9 stepper I haven't found yet that makes all my thinking obsolete. Maybe a reduction system even is a more valid playground for servo motors (clearpath, odrive, ..) as it possibly eliminates most of the cogging issues at low speed?
-
@bberger I am currently preparing iHSV servomotor mounts with 3:1 drive reduction for my corexy printer. I have designed and printed the mounts to directly replace the original motor mounts and will be installing and testing them in a week or two (waiting for Duet expansion board to arrive). I had previously tried driving the printer directly with the servomotors but found the resolution of their Hall effect encoders was insufficient for 3D printing. The new motor mounts use 20 and 60 tooth pulleys with F625 bearings and 5 mm shafts to do the reduction. Here is one of the two motor mounts:
https://photos.app.goo.gl/Zc6jSDT3b6pQi9ok7
The motors are good for 3000 rpm, so with 3:1 reduction driving a 20 tooth pulley in the corexy mechanism, they should be good for up to 666 mm/sec. That's faster than I would ever expect to print, but larger drive pulleys would allow higher speeds. The motors have plenty of "grunt" to move the mechanism- much more than the steppers.
