Crazy axles idea design

[gonper]

Today I was able to start the building.

I trimmed a little bit the gears in order to avoid them grinding into the ballbearings. I do not have a lathe so I have to be imaginative, and did the trimming with a stand drill and a Dremel.
As well I was able today to print one of the parts of a gearbox, looks like tolerances are ok. Now 2nd part is printing. I hope it closes and works properly... and I do not need to modify more for now.

[CM9000]

oh my lanta.. where did you get those gears from? #bringthewormsback

[gonper]

Originally Posted by CM9000

oh my lanta.. where did you get those gears from? #bringthewormsback

I got them on Aliexpress, you can find manny models:
https://www.aliexpress.com/wholesale...Text=worm+gear

As well you can get them from here, but much more expensive (provably better quality):
https://shop.framo-morat.com/en/Worm...r-distance-25/

[Gandalf The White]

Couldn't you have re-printed with a little more clearance? Either way, looks sweet!!!

[gonper]

More parts made or modifyed.
I have a problem with the metal parts as I do not have a lathe or a CNC so I have to handcrafted them, so sometimes they are not as accurate as posible but hope that enough for low speeds.
As well I have problems with the 3Dprints layer adhesion. For this testing parts I am using a cheap ABS that I had around and provably the bad layer adhesion issue is because the bad quality of the material.
I still have not decided which will be the final material that I will use for the final printed parts.



And this is how it looks assembled from inside and outside,
After this first assembly test, this parts will need some minor modifications.

Couldn't you have re-printed with a little more clearance? Either way, looks sweet!!!

Clearance, where?

[Almighty Malach]

I know it seems expensive, but Shapeways prints with SLS and it's held up really well for me.

[Gandalf The White]

Originally Posted by gonper

As well I have problems with the 3Dprints layer adhesion. For this testing parts I am using a cheap ABS that I had around and provably the bad layer adhesion issue is because the bad quality of the material.

You could look into filament drying, usually baking in the oven at 100C ish, that can fix a lot of issues with print adhesion. Also, PLA is way easier for testing, not quite as strong, but usually unnoticeable different.

Originally Posted by gonper

Clearance, where?

You said you had to shave the worm wheels down to clear the bearings, and I thought you might be able to just add a little space.

[gonper]

Quote:
You could look into filament drying, usually baking in the oven at 100C ish, that can fix a lot of issues with print adhesion. Also, PLA is way easier for testing, not quite as strong, but usually unnoticeable different.

The problem is just the quality of the material, nothing else. I used it for testing because is what I had around...
Now I am testing good quality materials. PLA, HIPS, ABS, ASA, Nylon, PET, etc.

You said you had to shave the worm wheels down to clear the bearings, and I thought you might be able to just add a little space.

The clearance I was looking for is between the gears and the bearings and I didn't want to add seams for clearance so I just shaved a little the gears.

[gonper]

I sped the day preparing the 4 ESCs, check here: Modified BECs

[gonper]

Functional testing parts are starting to come out of the printer.
And first assembly testing are happening.

I did not realized how big it is going to be.

[Gandalf The White]

That looks like a sweet start! The size in cad is always deceptive, I know what you mean,

[gonper]

ITS ALIVE; ITS ALIVE !!!

After several the 4th or 5th iteration I got the first running test of the gear box. It is very difficult to get the right tolerances on FDM 3D printing when a precision around 0.1-0.2mm is needed.
As I supposed the gears supplier did not provided to me with the correct gears mesh so it is very noisy and will requiere another iteration, hope the last one.

[Gandalf The White]

Awesome!! That looks like it will have pretty decent top-end as well! I am surprised there were no specs, but that looks very close to perfect. A little tight based on the sound? IDK, never tried building anything with worm drives.

[gonper]

After a little pause due some work traveling I managed this week to do some progress.
I printed one of the "final" axels and tested the steering system, it is still missing the steering link between the two wheels but looks that it works quite well.

Here is a video of it in action: https://youtu.be/KY8-GKs9CE4

[Gandalf The White]

Looking good!

[gonper]

I am facing a problem that was already itching me since I started this project.

I am using 4 x 380 size motors that in total have about the same power as 2 x 540 size motor or a bit more. But in this case there is one motor on each wheel and they are not connected mechanically by any manner, that means that each motor has to be able to generate enough torque to move the vehicle, for example when climbing with the traction of only one wheel.

In other words, on a MOA crawler all power of a 540 motor can go to only one wheel, but not in this case.

The necessary force at the wheel surface in order to do a vertical climb with that wheel is around 30N and the motor I have chosen is able to get very close to that theoretically but the real thing is that I am getting only around 8-10N

Most of the looses of that torque are on the worm gears that have lots of friction.

[CM9000]

small brushed motors have not enough speed. you would have to go with a 380 sensored design. I know some are out there, but the choices are slim.

also.. I believe they might only be able to run on 3s or smaller.

good work on design though. keep at it!

[gonper]

Originally Posted by CM9000

small brushed motors have not enough speed. you would have to go with a 380 sensored design. I know some are out there, but the choices are slim.

also.. I believe they might only be able to run on 3s or smaller.

good work on design though. keep at it!

Thank you for your input.
Well, the problem is not the speed of the motor, there is available very high revs. ones, but then the gear ratio will need to be changed and in this design it is not an easy thing...

If it does not work well how it is now, this is what I am thinking:
- Plan A: Change the gear ratio from 1:30 to 1:40 so higher revs. motors (more powerful) can be used and at the same time use smaller pitch gears (from 1M to 0.7 or 0.5M) in order to reduce friction.
- Plan B: Go the brushless route ... Can you let me know where 380size sensored motors can be found?

[TheLetterJ]

Maybe 380 sized outrunners could be packaged better and avoid the need for sensors?

[Gandalf The White]

Originally Posted by gonper

Thank you for your input.
Well, the problem is not the speed of the motor, there is available very high revs. ones, but then the gear ratio will need to be changed and in this design it is not an easy thing...

If it does not work well how it is now, this is what I am thinking:
- Plan A: Change the gear ratio from 1:30 to 1:40 so higher revs. motors (more powerful) can be used and at the same time use smaller pitch gears (from 1M to 0.7 or 0.5M) in order to reduce friction.
- Plan B: Go the brushless route ... Can you let me know where 380size sensored motors can be found?

What kind of rpm are you getting out of the current brushed motors you have? That is a key number when looking for brushless stuff, as the kv is always given. A 700-1200kv bl motor would be about the right speed be my calculations, on either 2s or 3s depending. That will put out a LOT of torque, and are pretty easy to get in 28mm diam. 36mm length.

You are going to have a hard time finding 380 size sensored stuff that has torque. Most of them will be targeting 30,000-50,000 rpm unloaded, which is way to much.