Delta steps per mm - how to calculate?

RepRap forum,

I'm in the middle of spec'ing out parts for a home built delta. I want to expiriment with lead screws, knowing full well they will be much slower than their belted brothers, plus added challenges of backlash. I'm going to explore cnc and other items with this build, basically making a beast of a machine, in a delta format.

But, therein lies the question. Steps per mm are easy to configure in a cartesian, where you just mark the axis you want to measure, move it x # of mms, then measure the difference and adjust.

Delta's with 3 axis' moving at the same time definately puts a challenge in this. What's the easiest way to figure this out on a delta that's opting to go with nema 23's with a 8mm lead screw on each arm?

Thanks for any info that can be shared.

The steps/mm that the firmware needs to know is for the carriages. So you can measure it or calculate it from the leadscrew pitch and microstepping, just as for a Cartesian printer.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Thanks for responding. Yes, I think I understand the math:

NEMA 23 Specs:
Part No.: 57BYGH627
Frame Size: NEMA23
Step Angle: 1.8 degree
Voltage: 3.0V
Current: 3.0 A/phase
Resistance: 1.0Ohm/phase
Inductance: 1.6mH/phase
Holding torque: 19Kg-cm
Rotor inertia: 480 g-cm2
Detent torque: 0.68 kg-cm
Number of wire leads: 4
Weight: 2.20lb (1.0 kg)
Length: 78mm

Lead Screw Rod Specs:
Diameter: 8mm
Pitch: 2mm
Lead: 4

Printer Controller Board (SmoothieBoard)
5 stepper motors drivers, 1/16 microstepping


Based on this info, and using the RepRap calculator, I should be sitting at 1600 Steps per MM - Pic attached. But... Please bear with me and bring me into some learnun'

What I'm basically asking for, is how do you go about verifying that # is correct? I mean, I've had to adjust my multitude of i3's because the belts weren't using exactly G2 pullies, etc. How would one go about doing the same sort of trick used in Cartesians, like measuring the rod from start to finish, and adjusting the difference between how far the carriage really moved, vs how far the software thought it did?

But seriously, thanks for chiming in on this, it's much appreciated.

Few things, first you should be using the lead and not the pitch to calculate the steps/mm for a leadscrew

Second, you are still using an old method to "calibrate" your Cartesian printers, you should stick to the calculated steps/mm adding an arbitrary scale based off a 100mm move will get you accurate distances only for 100mm and all other distances will be off. You have identified that your belt is not exactly GT2 so understand that this is a limitation of your printer. (I still dont know why people try to "calibrate" down to micron accuracy for something that is usually built on the kitchen table)

Third, have a look at delta setups, the math is a lot more complicated for moments, but to break it down it uses the delta radius and diagonal length to calculate the position of the head.

Finally, I dont know why you would want to slow down a delta with leadscrews, The major draw to the delta design is the speed. Alo you would need some stiff rods to make the effector stiff enough of light milling work, which will add more weight and reduce speed.

For a 2 in 1 type machine I would look at converting a mill to take a hotend. The forces and stiffness needed to mill are a lot more than to print, so if you start will a mill it will be easier to to convert and you know it can take the forces. Trying to make a printer into a mill will cost you a lot of engineering time. There is a sub here for people who are making "reprap" mills have a read through.

Understood. Thanks for the insight.

I do have a couple delta's already - one kit that was built, and another Flux machine. As far as speed is concerned, I see no real improvement between my cartesians and their delta brothers - when/if I get to bleeding nose speeds, the quality suffers beyond belief. I've basically set the print speeds relatively the same across all my printers.

That being said, the lead screw build is an absolute experiment, and nothing more. If it does not work out, and the machine is unsatisfying, then I'll break the parts down and make a pure cartesian cnc.

If you don't mind me asking, you mentioned you use the lead to calculate the steps per mm and not the pitch of the rod. So in my math, using the calculated variables should be as per attached?

Edited 1 time(s). Last edit at 04/10/2017 08:33AM by usslindstrom.

Yes that would be the steps/mm for that setup. Lead is equal to pitch when there is only one helical screw wrapped around the rod. Lead screws us multiple helical screws warped around the rod called starts. To determine the lead (mm per full rotation of the rod) you multiply the pitch by the number of starts. So with a lead of 4 and a pitch of 2 you would have 2 starts. Look at the end of the lead screw and you should see the different screws.

My responses was based off your first post that made it sound like you have never worked with a delta before and I don't mean to be rude, but how can someone that has two deltas still ask that question? It makes me think you never set them up correctly in the first place, maybe why you are not seeing any speed advantage out of them.

No worries about coming off as rude. You're helping me, and I appreciate it very much. I would be asking the same questions if I were in your shoes.

To answer the responses, let me see if I can do it inline:

"Yes that would be the steps/mm for that setup. Lead is equal to pitch when there is only one helical screw wrapped around the rod. Lead screws us multiple helical screws warped around the rod called starts. To determine the lead (mm per full rotation of the rod) you multiply the pitch by the number of starts. So with a lead of 4 and a pitch of 2 you would have 2 starts. Look at the end of the lead screw and you should see the different screws. "

Understood. Thank you very much for this clarification.


"how can someone that has two deltas still ask that question?"

Of the two Deltas I own, only one is a kit, with the other being the OUTSTANDING machine called "Flux". While the Flux didn't need calibration, as it's from the factory, I used the Reprap calculator for the belts to determine the steps-per-mm for that, which at no surprise for a belt-driven setup is at 80 for all axis'. Geometry is absolutely perfect with this setup, so there hasn't been any need to adjust.

"you are still using an old method to "calibrate" your Cartesian printers"

Agreed. And I will post no argument here. On my cartesians, I have played with multiple pully sizes, from Ebay purchases / used parts / etc. Even though it's "old", it still works like a champ. I've never attempted to bring it into sub micron accuracy, but to ballpark the whole setup. I.E. New pully moves the print head 80mm, when the software told it to move 100mm. Do a quick math adjustment and call it good. You are correct that even remotely imagining to get these things into NASA tolerance isn't something that should even be attempted. Close enough is good enough.

"It makes me think you never set them up correctly in the first place, maybe why you are not seeing any speed advantage out of them."

No arguments there. The closest delta I've been able to get to decent speeds, was the kit delta I put together from Ebay. I upgraded the motion system to use linear rails instead of the wheel rollers that came with it, which resulted in a VERY noticeable improvement on the speed I was able to push the thing to, but it's still nowhere near the videos I've seen on Youtube from other Delta users pushing their machines to the extreme. I've also upgraded to a BondTech extruder on the build which has helped a lot as well. But I still say that when it comes to quality parts, the Cartesian brothers (at least in my case) can output similar speeds (when configured as Bowden that is).

-- But that brings us back to the Delta stuff. For the kit I put together, the RepRap calculator was more than perfect. I just opened the question that since I want to try and play with lead screws in this build, I wanted to field the question on how to determine the appropriate math should the measurements be WAY off. I think I've got the math down, with the help of what you showed me in the last post - thank you for the assist. If I'm missing anything, please don't hesitate to call me out on it.

I welcome the help. Thanks.