Belt Z axis WormDrive issue

Hello everyone,

Having some trouble with the Wormgear Nema17 i am using for the Z axis in my new printer.

The output shaft of the gearbox has some play when moving the shaft manually, which makes sense.

I will try to explain the issue with an example;

I electrically move the Z axis down xx mm. The Z axis moves down, after this i want to check how sturdy the Z axis is, so i try to move the Z axis down by hand, which i can for about 4mm axis travel (the amount of play in the wormgear)
If instead i try to move the Axis up by hand, i can't.

If i electrically move the Z axis xx mm up and do the same thing, the Z axis can't be pushed down, but i can push it upwards for also the amount of play in the wormgear.

It seems i am currently not understanding this logic and how to use it for my advantage.

Since the Z axis has to move down after homing, the issue would be resolved if the axis can't be moved down after travelling down. The required force to push the axis down is very minimal and will give issues.

Photo of the axis in attachment, advice is much appreciated!

update: Maybe it would help switching the motor assembly in the other direction?

Edited 1 time(s). Last edit at 05/14/2020 05:59PM by Govahnator.

Does the bed assembly drop without the belts attached? It should.

You're not using counterweights are you?

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

The assembly does drop without belts, moving without belts the axis is also very smooth from bottom to top.
I am not using counerweights. Bed does not drop when turning the power off either.

It sounds like you have backlash but it shouldn't be possible. The weight of the bed assembly should be keeping the gears engaged at all times. You should not be able to push the bed down because it should always be sitting at the lowest point the gears will allow.
When you say the shaft of the gearbox has play, is that rotational or longitudinal? Do you have a link to the gearbox?

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

To try to counter the issue, i did push the Z axis down, so that the motor position is in the lowest of the backlash so it can't be moved down, only up. Then i did multiple electrical movements z upwards , after these movements i am not able to push the z axis downwards , which is fine.

As soon as i do electrical movements downwards, the backlash reveresed and i am able to push the axis down again after every move.


The play/backlash is rotational.

This is the exact motor:


[nl.aliexpress.com]

Edited 6 time(s). Last edit at 05/15/2020 10:53AM by Govahnator.

That gearbox has a bunch of gears in it, including a small worm gear. It may be that the friction inside the gearbox is enough that the weight of the bed assembly doesn't keep the gear surfaces in contact. Try adding more weight to the bed. Which gear ratio did you buy?

The higher the gear ratio, the more likely you are to have backlash with a complex gearbox.

The simple worm drive I used has one worm gear and one disc gear so the gear teeth are in contact all the time, resulting in zero backlash.

Edited 1 time(s). Last edit at 05/15/2020 10:48AM by the_digital_dentist.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

The gearbox has a 1:17 gear ratio.
I will try to do the testing with some KG's of filament on the assembly.

My assembly is very light, it uses 2040 and 2020's, the light weight versions. The bed i have currently tested is 220x220mm 5mm (of my previous printer).
The bed i will use is 340x340x6mm, will still be a relativly light weight Z axis.

I think the very light weight is the problem.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

You are right. I placed some filament spools on the axis (with no alu plate attached), 2 spools of a kg, not enough, 3th spool did drop the axis.

My 1.9 kg cast plate is not sufficient to make the axis usable, it needs another Kilogram.

Solution: I have to buy a 420x420x8mm plate of +-4kg Or make the axis heavier on the frame supporting the plate. I have 2020's laying around that i could attach to the 2040 holding the bed.

Thanks for the help DD

Edited 1 time(s). Last edit at 05/15/2020 02:17PM by Govahnator.

Does the motor have any trouble lifting it with the extra weight?

I did belt stretch tests by loading the bed with 4 kg (in addition to the bed assembly's own weight). I did that because I figured I'll never print anything that heavy, so it was a good test for belt stretch and to see if the motor would have any trouble lifting that mass (it didn't).

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

It does not have trouble with the weight, i did a test with 7kg on the Z axis and it goes fine apart from the steel core belts starting to making a cracking noise.
I really dislike the steel core belts and will be replacing them even though the glassfiber belts stretch more.

I will make a mount for my dial indicator and do some tests aswell.

Although not the most aesthetically pleasing, i implemented a solution as seen in the pictures below:

[ibb.co]
[ibb.co]
[ibb.co]

Update:

Although issue much less pronounced, it is still there.
Even with 5 kg extra wait on the axis, it still moves a bit down (gears driving the axis move a bit) when adding extra wait after moving the axis down by controller.

Like mentioned before, when adding wait to the axis after moving it upwards, the axis is rock solid and does not move.

Not sure how i can try to become the opposite.

What if i flip the Z motor to the opposite side, so it's movement would be inverted. Would this help?

Edited 3 time(s). Last edit at 05/19/2020 06:40PM by Govahnator.

The shaft couplers aren't twisting, are they?

Maybe flipping it over would help, but I doubt it. Only one way to find out...

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

If you mean twisting in like squirming ,no they don't.

They just move a little bit until the lower limit of the backlash in the motor is reached.

I just don't understand the logic behind the Z motor being in the lower limit of the backlash after a z+ move, and being in the uperlimit of the backlash after a z- move.

Maybe i should make a video showing what i mean, English is not my native language so i am not sure i am putting it straight.

I will try to change the motor position, don't think it will help either but i have to try.

So i tried changing the motor to all possible different directions, still no luck.
I did some testing with the printer placed vertically, even without "the gravity" interfering i still have the same issue.

Made a video of the vertical testing which you can find in attachment.

Basicly, after moving the motor in whatever direction there is further backlash in the same direction.
I guess i can't do anything appart from getting a different wormdrive or making the Z axis much more heavy.


[streamable.com]

[streamable.com]

Edited 4 time(s). Last edit at 05/21/2020 07:05AM by Govahnator.

You can contact the "engineer" who designed this worm gear box. His name is Sum Sing Wong !

Note, the behavior described in the above post is normal, just that the gear play seems quite big.

Edited 1 time(s). Last edit at 05/22/2020 04:41AM by MKSA.

---

"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)

So i ordered a new cast tooling plate to increase the weight. 412x412x8mm = 3.6kg.
However it seems that even this will not be sufficient. Oh well , we will see.

Hi everyone

Update:
I managed to resolve the backlash issue and all issues mentioned in this topic by opening the Nema17 gearbox and reassembling it.

It was fairly easy. I disassembled the gearbox and realized that if i could untighten the 3 screws,move the whole gearbox more to the right and get the gears closer in the worm wheel on the Nema17, the backlash might decrease.
With a screwdriver i was barely able to reach them, but i managed to loosen the screws, i then was able to move the gearbox a few mm to the right.

After assembling i did not feel any backlash in the shafts, i could not remember if the shafts needed load to make the backlash appear though.

I decided to spend an afternoon reassembling my Z axis to refit the gearbox and Nema17, i did not expected an improvement but to my surprise the issues mentioned in this topic are resolved.

The bed/Z axis does no longer drop when i put weight on it.
Z movement seems to go fine, did a test print today with a few layers and it seems to be able to do small movements like it should.

The issue i had with the gearbox was an assembly issue, not a faulty design.

Further testing will point out if any other issues come up :-)





(A gear in the gearbox is missing in the picture)

Edited 3 time(s). Last edit at 11/12/2020 07:13PM by Govahnator.