Cnc For Mini-lathe

[Billh50]

Jim,
I was just hoping someone on here had some knowledge about doing these mini-lathes. Guess I will keep searching.

 

[ch2co]

Bill
Sorry if it seems that this forum doesn't seem to be answering your questions.
I'm probably sure that you are familiar with this site:

http://www.stirlingsteele.com

This guy us apparently selling his plans for this retrofit.

CHuck the grumpy old guy

 

[Billh50]

Chuck. Thanks for the link.
I had seen a link to some of his stuff but never saw the link to actually buy the plans. I just sent him an email to see if his plans included what steppers to use as well as drivers for the steppers.

Bill

 

[Wreck™Wreck]

Jim,
I was just hoping someone on here had some knowledge about doing these mini-lathes. Guess I will keep searching.

You have asked a very vague question that no one may answer with any accuracy given the lack of parameters required. Do you require low cost? High speeds and power or something similar to running it manually? Do you require high accuracy when interpolating arcs and angles? Do you require a C axis for threading and rigid tapping? Do you want a conversational control or a straight G-Code control? Do you require a tool changer?

Without any parameters, (best way to convert a mini manual lathe to CNC control?) is an open ended question with no concrete answer.

If indeed you would like concrete information contact one of the several "companies" that sell kits for this sort of thing, they will tell you exactly what you want and sell it to you whether it works in your application or not.

If cost is no object, which you haven't mentioned, then I suggest skipping the stepper motors and use AC brushless servos coupled with 0-backlash ball lead screws designed for this exact purpose.

Narrow down the questions a bit.

 

[JimDawson]

1) Type and Oz of Stepper is best for this but not real expensive.

A quick google search turned up a few good hits, it looks like a NEMA 23, 400 oz/in stepper would do, but these were being used with ball screws. If you don't plan on changing out the original screws, a bit torque might be in order.

2) Easiest software to learn.

Mach3 CNC software is the most common and the least cost as far as I know. Very well supported by user forums and we have a number of users here. You will also need CAD and CAM software to draw the parts, and to generate the G-code. Fusion 360 by AutoDesk is free to hobby users, is both CAD and CAM.

3) Any links to beginner tutorials

Are you asking for tutorials for the build or the operation?

 

[Billh50]

OK....I was basically hoping someone did a conversion on a mini-lathe. I am not looking for great speed. I mainly just want to turn some contours and do some threading a bit easier than I do manually. I already have a couple cad/cam programs. The one I use the most is an older verssion of MasterCam. I am not going to make this into a workhorse as it was never intended for that. I do know where I can get my hands on a couple ballscrews fairly cheap from a friend. So yes I am going to ball screws. My only income is disability which isn't much. So Money is an object. I knew about using nema 23 steppers but I had no idea what size in oz's would work. I didn't want to get something that was so weak I couldn't take a cut without it stalling.
If no one can answer this then maybe this thread should just end and I will have to search more elsewhere on my own.

 

[JimDawson]

Around 400 oz/in are about the biggest you can get in a NEMA 23, so that is what I would go with. Here is an example system, 2 motors, 2 drives, power supply, interface board and cable. Pretty much plug & play with Mach3

http://www.ebay.com/itm/2-axis-Nema...044850?hash=item4615d5d0b2:g:098AAOxyWmxSZdMY

 

[Billh50]

Jim,
Thanks for the help. That price is definitely in my range.

 

[hanermo2]

Some bad advice. Some good.
I did this, back in 2004 or so, on a 7x.

1. You dont need lots of torque. 1.5 Nm or so is enough.

2. You do need resolution. Nothing is enough.

3. Your aim is for rigidity, and more mechanical resolution leads to more repeatability and rigidity.

4. You need high acceleration, for pullouts only, in threading. Nothing else.

5. You dont need high speed.

The best combo, minimax solution, is 2M542 type chinese drivers (40), 48V DC PSU, nema 23 steppers of 2.5-4 Nm (30-40), belt drives of HTD5-15, at 1:2 - 1:3 (15 -20 € or so).
On steppers, use 14 mm teeth minimum for good belt wrap (12 is too low, imo, ime).
Your step size will be 2000 at 1:3 for 6000 steps/turn.
Your mechanical resolution is still 1/2 step, or 400 steps/turn.
On typical 1 mm screw => 400 steps/mm, which is just right.

Mach3 threading is broken. It will sort-of-work but cannot be relied on (lots and lots of gotchas and it depends).

The only good sensor for threading is an optical encoder.
You need a fast, crisp, signal, with predictable delay at widely varying speeds from 50 to 600 rpm.

A fussy sensor will create fuzzy quality threads.

Suggest polabs cncaddon motion control solution, ethernet, and mach4.
120€.

They have excellent drivers/support/tech for cheaper stuff.
It can read an encoder z/ pulse, ie has support for fast signals.
It will support encoder threading.

Bigger steppers will have lower max rpm, and will accelerate slower.

If you need to go faster than nema 23, its about the same price as big steppers/drivers, and 10x better, to go with modern, cheap, ac brushless servos.

Background:
I have been working on high quality threading with machx for over 10 years, and 2000 hours.
My current system is ac brushless servos, of 10.000 counts/turn, AC220V, 750W, on x and z.
Its very expensive, and meant for me, to be able to do "perfect" threads (industrially).
500 kHz on X,Z, C axis.
4000 steps/mm, or 0.25 microns step size (=/= resolution).
CSMIO-IP-S, industrial controller.

 

[Billh50]

Thanks, I really appreciate hearing from someone that has done this before.