rotating ballnut on RF31 Z axis

[Arquibaldo]

Dear Members:

Unfortunately, nobody could help me with a RF-31 Mill-Drill Z axis conversion from
the factory rack&pinion to a rotating ballnut, so I decided to apply my (extremely)
limited mechanical design habilities to this problem!

I will post here my progress(?), hoping that someone will take the time to evaluate
what I propose and offer me improvement tips (or, more likely, tell me this will
never work...)

My requirements are:

1. Extremely low cost.

2. Minimal modifications to the mill (hopefully only a few threaded holes
on the head casting) so that, if this project fails, I will still have my old
mill (flawed, but better than nothing).

3. If at all possible, retain the ability to drive the quill manually, using
the factory handles, with only a simple setup to choose between manual and
stepper driven quill operation.

4. Under stepper power, capable of precisely moving the quill for light-duty
hobby use only, mainly machining aluminum or mild steel.

5. The stepper will initially be manually operated, with a precision multi-turn
potentiometer, or I may even attach encoders to the current handles
for the quill operation - computer and digital electronics are (were?) my area.

I started by reviewing everything I could find online about rotating ballnuts.
I found many interesting designs, particularly on the thread
"Rotating Ballnut - design ideas" by Jonathan, on this site,
but most were sophisticated, expensive mechanisms aiming at high speeds on very
long ballscrews. This is the oposite of my needs: very low speed with a very
modest travel: about 150mm only.

I found only one design that I thought I could adapt to my needs:

https://www.youtube.com/watch?v=iuAQnf6vSAA

So, I started modelling my rotating ballnut!

I attach pictures of my initial attempt.

Actually I couldn't find out how to attach a PNG file to a message.
Will someone help me? This post is meaningless without the
figures!!!

It uses a single 5208 double angular contact ball bearing (downsized from
the original 5212). These bearings are used in the automotive industry
and are very cheap (if chinese...): about 10 US dollars...

I also opted for a SFU1605 ballscrew&nut (which I already have) and a
timing belt & pulleys system with a HTD5M 15mm wide profile: 50/20 teeth
pulleysm and a 360mm belt, yielding a center distance of 89.4mm.
This kit can be bought for about 15 US dollars, plus shipping.

So, here are my initial worries:

- Is a 16mm ballcrew rigid enough in this application? I hope it won't
flex much over a length of only 150mm... I would prefer a 20 or even
25mm ballscrew, but these are more expensive and take more space,
requiring a larger bearing and the whole mechanism might not fit
in the limited space available.
Actually, a 20mm screw could fit in this design with minimal changes,
so do you think this upgrade ie worth the cost: 32 dollars (plus
shipping) for a SFU2005 ballnut&300mm ballscrew or 55 dollars
(plus shipping) for a DFU2005 double ballnut&300mm ballscrew?

- Is a HTD5M 15mm belt adequate here? This choice is not critical:
if this belt proves inadequate, it can be upgraded with minimal
impact on the rest of the system.

Note that this transmission has a 1-to-2.5 reduction ratio and with
a 5mm pitch screw we will have very fine increments per step:
0.005mm per pulse at 400 pulses per revolution, and a tremendous
power gain, so that I think a NEMA23 stepper will be more than
adequate.

Also, this mill has a spring return mechanism which can be adjusted
to just balance the weight of the quill.

Then, there is the problem of where to place this mechanism on
the mill and how to atach it to the quill.

Originally, I intended to mount the system inside the head casting,
but it proved very difficult to mount: there simply is not enough
space.

I think I found a suitable spot on the left side of the head - there
are some obstacles, but they can be overcome.

Attaching the system to the quill is the remaining, difficult problem.
The ideal solution would be to have two screws, one on each side of
the head, working synchronously thus balancing the forces on the quill.
But this would be very difficult to implement and very expensive.

So, I am hoping that my unbalanced solution works:

There will be a very strong (25mm thick) plate attaching
the quill to the ballscrew. Note that, in this machine, the quill is
very beefy (75mm in diameter) and I think it will support the forces
involved.

Note also that this is a preliminary design and many details and refinments
are missing, in particular the means to attach the mechanism to the
left side of the mill head.

Before cutting metal, I intend to build a mockup with a cheap, easy
to machine material (MDF or PEAD) to test the concept. Also, I must
decide which parts will be aluminum and which must be steel... what
do you think?

Here I stand... I hope at least one member has the patience to read all
this and offer some criticism and suggestions.

Regardless, I will keep posting my progress, if any.

Arquibaldo