Pulley Teeth

[Robo]

Hello,

Just planning to change X and Y axis pully 50 teeth to 20 Teeth for more accuracy.

can anyone tell me pros and cons for both type of pulleys. is it good to change 50 teeth pulley to 20 Teeth pulley?

Thanks

[Michael Anton]

That should be fine, as long as you are not using belts with a steel core, as they have a minimum bend radius, that might be larger than what you would get from a 20 tooth pulley.  I'm not sure what the minimum radius is though.

[Oz-Ron]

G’Day Robo,

 

https://www.robotdigg.com/product/939/GT2-profile-brass-pulley-for-high-torque

 

FWIW, I am using Nema 23 motors with 8mm shaft.  They marry up to the 30 tooth pulley nicely.  Derek can supply 18mm wide GT2 belt to suit on request and it works well for me.

 

Cheers,

Ron

[Robo]

Thanks for reply,

First , i am not using steel core wire so bend radius issue will not come to the picture . see attached pic.

Second i am worried about only steps, means for 50 teeth pulley it was only 64 steps not for 20 teeth pulley its about 160 steps. so is the any chance to miss steps ?

 
Thanks,

[Cri S]

You don't loose steps, only speed.

The max speed is between 40 and 60% of you actual speed. 40% is the direct mapping, but because more steps usually allows better motion curves speed could increase from 40 up to 60% of actual value or even more if the actual pull in/out curves are bad.

You increase the accuracy from 0.125 mm to 0.05 mm allowing 0.3mm technology where with 50 teeth only 0.5mm technology was possible. This are numbers for production, for prototyping where additional rework is acceptable 0.4 was possible with 50 teeth.

--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+unsubscribe@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/65499c7b-8354-4dee-afdd-6bbbf36af0ad%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Marek T.]

Hmm, Friends, what is typical accuracy of your home-made machines build with different NEMA or other stepper motors? Is it this 0.03/0.05 mentioned by Cri?
On my re-trofitted scrap I have now accuracy between 0.02-0.01mm and just trying to get it better - exactly to get stable 0.01, have problems with DC servomotrs vibrations... Am I strange?

[Eagle Media]

On my current setup with 0,9° steppers and 1:1 transmission i have got 0,08mm moving/step with full steps on X,Y.
This will give you 0,04mm accuracy in worst case.

[Trampas Stern]

On home built machines the accuracy is often dictated by the mechanical components more than the stepper motors. For example I had a terrible time with plastic brackets as they would flex, this flexing was more than the 20 tooth pulley and steppers motor accuracy.   

Again my experience has been:

1. No plastic/wood parts - all mechanical critical items need to be metal

2. Invert Anthony Webb's design - put Y rails on bottom of extrusion for more cross bracing and acrylic top (keeps dust off working surface) also allows more z height for feeders. 

3. Use Smart Steppers with at least 24VDC for speed and accuracy (better yet request a 36V smart stepper)  - www.misfittech.net (yes a shameless plug) 

4. Use as wide of a belt as you can physically fit.

5. Nozzles should be multiple of feeder spacing (one day Jason will add having nozzles pick from two feeders at same time)

6. Nozzles should have independent Z control, see number 5

7. For accurate rotation using gear on nozzle, no direct drive hollow shaft NEMA 8. Accurate X/Y does not help if you can not rotate part to same accuracy.

8. Both Y axis need to be driven, I prefer a rod and one motor but have no data this is better/worse than dual steppers.  

9. Only optical end switches - mechanical switches have repeatably issues. 

Again no wood or plastic anywhere in your design for mechanical components, if you want it accurate.  

I have thought about trying to get some people together and work on a good common mechanical design.  I estimate a good design would cost around $2000 for the custom parts (aluminum base, Yamaha feeder plates for 64 feeders, brackets, etc). You would still need to purchase other parts, like motors, electronics, belts, solenoids, feeders, etc. Hence a full machine build might cost around $4000-$5000 depending on feeders.  Would anyone interested in working on such a project or buying the kit parts? 

Trampas

[Marek T.]

Or much more dynamic and fast DC/AC servomotors powered 80-200V :-). But there is many problems to tune it really good :-(. And it's not too cheap solution of course.

[Eagle Media]

Hi,

This is exactly I'm working on allready.

1. Done
2. I don't know his design, I didnt looked-I have got my own.
3. 36V/48V 0.9° NEMA 23 stepper with custom driver.
4. Done
5. Done - I have asked Jason for this. Currently no plans to integrate this option.
6. Done
7. Currently on Nema 8. I'm busy with the new head design. I'll go with 3/6/8 nozzle type head.
8. Done
9. Done

This kind of machine should be unter 2K EUR (net price) incl. all parts.

Regards,
Rick

[Jason von Nieda]

5. Done - I have asked Jason for this. Currently no plans to integrate this option.

5. Nozzles should be multiple of feeder spacing (one day Jason will add having nozzles pick from two feeders at same time)

Something to keep in mind is that this functionality would be pretty easy for you to add if you want it. It's all controlled by ReferencePnpJobProcessor, and the entire job processor can be swapped out with another class just by changing machine.xml. It's designed this way so that people can freely add this type of functionality to their machines.

Jason

[Eagle Media]

Hi Jason,

I meant about the combination of multiple nozzles with intelligent part placing we are talked about some weeks ago.
When I asked You if there are plans to integrate the posibility of changing the placement order of the parts in Software.
IMHO this would be a real benefit for multiple nozzle systems. With speed improvements.

Regards,
Rick

[Jason von Nieda]

Hi Rick,

Yea, falls under the same situation. I don't have plans to do it, or at least it's low priority due to some other important work that has to be done to improve stability, but again, all of that code is in one place and it's easy to swap it out on a machine by machine basis without making any changes to the rest of the system. So, if this is something you really wanted, you could work on it without requiring changes to OpenPnP as a whole.

Jason

--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.

To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/db956809-cc79-474f-879b-16d8bf7da4a3%40googlegroups.com.