Pulley and Pinion sizing advice

[ClarKent101]

I'm attempting to size a pinion and pulley for my Y axis and eventually the gantry X axis. Can someone who knows better cast an eye over this and see if this makes sense?

Using a MOD1 rack and pinion

3:1 Ratio with a belt reduction from pinion to T5 pulley

T5 Pulley size: 15 teeth
Pinion size: 44 teeth

Other info:
Plasma cutter: Powermax 45xp
Steppers: Nema 34
Belt width: 20mm
Linear rails: HGR25 (HWIN)
Machine operating speed: 1,270 mm/min (55 ipm)
Moving / jogging speed: 6,350 mm/min (250 ipm)

Is there anything else I should be taking into account here?

[robertspark]

suggest the use of a 20T pinion
https://www.apexdyna.nl/en/why-does-the ... e-20-teeth

my gear driven 3:1 reduction drive uses 60:20T

I used mod1 but some believe mod 1.5 or 2 to be optimum

a nema34 on mod1 rack... bit big imo,

I run
NEMA 23, 4.2A, 3.2mH, 3.1nm / 439oz in holding torque
mod1. 20T with a 3:1 gear drive reduction and get 600ipm at 0.6G comfortably.... it can go a LOT faster + higher acceleration before stall... but no point

do you have a sketch of your design?

do you have more info on the NEMA 34 motors

[ClarKent101]

Hi Rob,

I've attached a few pics of what I have so far. I've been trying to ensure that I maintain a certain amount of accuracy, using some of the parameters given on the Hypertherm site, they recommend around 0.25mm for travel (rail and traverse axis).

https://www.hypertherm.com/learn/artic ... r-xpr300/

So if my stepper motor is 1.8 deg = 200 steps per rev.
Linear distance: Pi x Pitch Diameter
3.142 x 50mm = 157.1mm linear travel per revolution
157.1mm / 200 steps/rev = 0.75mm linear travel per step.
0.75mm/0.25mm = 3

So my gear reduction should be 3:1 to get that accuracy of 0.25mm. Is this correct?

Also, big gear on pinion mating with rack, small pulley on stepper?

[robertspark]

I am confused, how did you get:
3.142 x 50mm = 157.1mm linear travel per revolution

where did 50mm come from?

https://www.linearmotiontips.com/differ ... of%20teeth.

you have a MOD1 rack and a 44T pinion.

1 MOD x PI (3.142) x 44T = 138.23mm of linear travel per pinion revolution

Stepper motors are SELDOM run at 1:1, and are normally run at a higher microstepping resolution for smoothness. I set my drives at 10 micro steps (this has a LOT of different opinion .... choose what you feel is correct!)
Below are some of my choices in why I used 10 microsteps and no more.... but I'm also running a metric setup so the numbers are easier (5:10 etc) than 4:8:16:32 etc which tends to align better with parts of an inch (1/4; 1/8th; 1/16th etc)
https://www.geckodrive.com/support/step ... ution.html

https://e2e.ti.com/blogs_/b/industrial_ ... -confusion
https://www.faulhaber.com/en/support/te ... realities/

so if your machine was mine:
~138.23mm / (200 full steps x 10 microsteps x 3 (the drive reduction ratio))
= ~0.0230 mm / step

I'd still look at a 20T pinion as that is what seems to get used by most setups + for the reason I posted earlier

These may help you for ideas:
https://www.avidcnc.com/pro-rack-and-pi ... p-227.html
https://www.finelineautomation.com/prod ... ion-drive/

I designed and made one very similar to this using standard gears and parts with laser cut plates:
https://www.ebay.com/itm/132994274061?h ... xyiRlScaKM

[ClarKent101]

Thanks a lot for this info, there's some good resources. I keep getting pulled away from it this week but I'm going to figure it out!

Yes the 20T makes sense, if I can achieve 0.023mm / step accuracy that would be great. I expect some inaccuracy from things like backlash, belt tension and ??? something else will add up at the end. What sort of accuracy do you get on your cuts when all said and done?

[acourtjester]

If you do the steps per calibrations correctly you will get a very accurate movement on your table. With plasma cutting there are trade offs like with cutting holes due to the taper of the plasma stream and the thickness of the material. Cutting speeds around corners can affect the following of the G-code, there are tricks to help with those, with G-code planning. Material and construction methods could affect how ridget the tabe is and this will have effect on the accuracy too, I am not a fan of bolted together tables.

[robertspark]

plasma is still a blunt axe (at least the run of the mill air plasma systems)

it's not as accurate as laser.

each to their own.... beats oxy acetylene / jig saw or grinder.... perato principle.... you'll send a lot of money and time chasing that last 20% of accuracy https://en.m.wikipedia.org/wiki/Pareto_principle

there is also the natural bevel to account for as Tom mentioned

I do get some things laser cut sometimes

[ClarKent101]

Yes I'm ok with <1mm accuracy but would like to get that to <0.5mm if that's realistic. I know what you mean about getting that last drop out of the machine, not sure I'll be chasing that either. I was originally going to build the table to bolt together but decided to move my welder to where it's going to reside and build it in situ. I'm starting to think I should make the water table removable though, in case for future wear or corrossion.

As far as the micro stepping goes - I think I'll go with the 20T pulley and 10 micro stepping like suggested. If I have a 200 step/rev motor then a 2000 pulse/rev driver should do ok?

So if I'm using a 20T pulley at 3:1 gear ratio = 0.010472 steps/mm

And the pulley attached to the stepper should be 20T/3 = 6.66 = 7T

Edit: I'm looking for a 20T pinion which has a larger gear attached, similar to the picture below. This would allow me to go to 10T on the minimum stepper side and 30T on the pinion side.
I'm having trouble finding a pinion like this, is this a configuration that has a name, which can be purchased together? 20T pulley with gear

gear.png