starting small. have a few questions

[ejholmgren]

I'm going to make my first attempt at building a small cnc mill as a winter project, and I have a few questions before I start scrounging up the parts.

The specs are as follows:
3 axis, ~18" x 18" x 3" travel (x,y,z) on a gantry system
3/8" Acme threaded rod and rods w/ sleeve bearings for linear motion
3 NEMA 23 size steppers
Dremel for the spindle
Either a Linistepper or HobbyCNC 3 axis kit for the driver
Linux box running EMC2 as the controller

My biggest question concerns the minimum size stepper motor I need to run a machine this size and not lose any steps. Do I need at least 100oz steppers for the x/y axis, and maybe something slightly less for the z axis? Also, what are the pros and cons of using bipolar vs. unipolar motors?

I plan on using using it for pcb prototyping, milling pine for printmaking, and maybe some plexi cutting as well.

Thanks,

Eric

[project5k]

what pitch are you going to use on the screws?? the more agressive the pitch, the more motor your gonna need (assuming a direct drive situation) and the larger travel per step of the motor...

[ejholmgren]

3/8" dia would be 12 tpi, but maybe I should just go with the standard 1/2" dia 10 tpi to make things easier in the end. still trying to decide between direct drive with spider couplings or small cogs/belts and offset the motors an inch or so to the side of the lead screws.

[project5k]

hmm, well my thinking would be to go with the thicker screws... reason being that you will be able to turn them faster without whip issues... also, i would personally go with the timing belt/pullie setup, this will allow you to change the gearing if you wanted to... run 1:1 or 2:1 or over drive at 1:2, but if you go ahead with it this way, you will have the easy options later...

[spalm]

Ejholmgren, Welcome.

Winter project huh? I see you are in Minneapolis, so you have about eleven months to finish this. (I grew up in Northfield so I know how long the winters are up there.)

As far as leadscrew pitch; 200 steps per rev x 8 microsteps x 10 tpi will give you 16000 steps per inch. Do you really need more resolution than that (or even that)? Adding belts and an offset motor will probably add more backlash (the slop you get when you reverse direction) than that. You need to look at the system as a whole; speed, accuracy, backlash, flex, and a whole bunch more.

You get more torque from a stepper if you keep the rpms down in the hundreds, maybe up to 1K. Not sure what kind of speed you are looking for but when cutting plastics you can not stay in one place too long without melting. Don’t make this machine too slow while worrying about accuracy. Cut something with your dremel or router by hand and notice how fast you are moving it (time it). You probably move it faster than you first thought. My recommendation would be to go with 8 turn 2 start ACME (4tpi). It is more efficient than standard ACME, you will more than double your speed, and keep the rpms of the screw down which will reduce whip and keep the motor in the higher portion of the torque curve. The 1/2” version is cheap at Enco.

You are alluding to sleeve bearings running on shafts. Seriously consider supporting those shafts. The flex in unsupported rails will make the above steps per inch accuracy inconsequential. Maybe look at split bearings running on shafts that are attached to a support that runs the length of the shaft. The cost will not be much more, but they will be a bit more of a challenge to incorporate.

Lots to think about,
Stay warm,
Steve

[ejholmgren]

R = RPM/TPI is the forumla for calculating rate of movement at a given rpm and threads per inch, correct?

... so maxing out at 1k rpm, I could still see speeds of up to 100 ipm w/ 10 tpi.

@ 1k rpm / 4 tpi I'd be moving at 250 ipm

I'm not looking for blazing fast speed or ultra fine accuracy with this first machine, just trying to make it a good learning experience.

[ViperTX]

Well firstly look at the torque curves for the stepper to determine the RPM, you'll see that the torque of the stepper decreases rapidly as the RPM increases.....so, go with a stepper in the 400 oz-in torque, remember ACME screws are only 40 to 50 % efficient....so, you'll lose half of your torque just turning the screw, directly couple your steppers, use a 3/4 inch ACME screw....why....are you turning your own ends of the ACME screw?

[spalm]

Hey,

Touché. Your speed calculations are correct. Speed can be over rated. My main point was that the screw will probably not be the concern for accuracy. Other flex and looseness in the system will be.

Seeing other home built machines around here, 150 ipm is very good, and this is usually just used for rapids. More like 50-80 for cutting. The 10 tpi screw will work very well. But the 4 tpi screw will allow you to stay down in the torque curve, give you more speed for a given rpm, give you less whip for a given speed, not cost much more, not affect your accuracy, and a 2 start is more efficient than a 1 start. IHMO.

More important is to actually build a machine.

Steve

[ejholmgren]

Steve, I see 5/8" dia 6 tpi acme rod in the enco catalog, but nothing with 4 tpi until 1 1/2" dia. Also, I don't see any mention of 2 or 4 start, so I'm assuming this rod is single start ... maybe I'm looking in the wrong place?

http://www.use-enco.com/CGI/INPDFF?PMPAGE=596&PMCTLG=00

Thanks for the tips about the torque curve and the greater efficiency of multi start threaded rods ... enco is great parts source too!

[ger21]

http://www1.mscdirect.com/CGI/NNSRIT...MT4NO=12093077

[spalm]

Hey,
I don’t want to push anything too much, but I just did not want you to jump to belt driven screws so you could get more accuracy. I see prices have gone up a bit in the last year.

1/2 10 tpi 1 start – 6 foot $28.50
http://www1.mscdirect.com/CGI/NNSRIT...MT4NO=12093076

1/2 8 tpi 2 start (4 tpi)– 6 foot $37.00
http://www1.mscdirect.com/CGI/NNSRIT...MT4NO=12093076

Here is a link of good stuff charts from Nook, a premier maker of this kind of thing. The second chart is quite interesting for speed. The site in general is worth some study. These particular charts do not include any torque or efficiency data but do show that that the ‘fixity’ or how it is held in place have an affect on rotational speed.
http://www.nookind.com/acme/acmecharts.cfm

What are you planning on using for linear bearings?

Steve

Hey Gerry, while I was typing, you were searching. I need to learn to type faster.