rotary vs linear deltas

[Peter Shabino]

Anyone have any thoughts on rotary vs linear deltas? Wondering if the linear would be easier to get precise repeatable movements between multiple machines over the rotary since there is less 3d printed parts in the critical accuracy path?

Just looking at my FPD and was thinking I have 80 to 90% of a linear delta sitting here. Just need Thomson rods, bearings, longer belts, and some mounts and adapters. All the motors, electronics, and sensors / cameras could be reused in the new machine. Was also thinking since large Z depth is easy on linear machines could also have multiple levels of feeders that the head pokes into to pick up parts. 

Thoughts?

 Peter

 

My projects:
http://www.wire2wire.org/

[Karl]

FPD has a nice rectangular frame that facilitates accessory mounting. 

Linear deltas typically have three dedicated vertical rails that cannot by used for accessories.

This doesn't matter for 3DP, but PnP seems to attract a lot of accessories.

3D printed parts affect accuracy and require calibration (e.g., firenodejs eventually).

With linear delta, your challenge becomes extrusion alignment. FPD doesn't really

care about the squareness of the frame, but linear delta frames must be precisely assembled

and manufactured. And you've got that "120 degree" requirement for the frame.

[Rich Obermeyer]

How about putting a square frame the same size as the FPD into a typical linear printer (see below) and you should be able to attach most of the items that could attach to the FPD.  A few blind spots but it should work.

I am just waiting for a good feeder that can be connected to it to finish it off.

Then openpnp is suppose to do the rest.

I need to try firenodejs to see how precise it can be.

I would be interested how multiple levels of feeders could work.  My printer has 16 inches of height!

--
You received this message because you are subscribed to the Google Groups "FirePick" group.
To unsubscribe from this group and stop receiving emails from it, send an email to firepick+u...@googlegroups.com.
To post to this group, send email to fire...@googlegroups.com.
Visit this group at https://groups.google.com/group/firepick.
For more options, visit https://groups.google.com/d/optout.

--

Rich

[Karl]

It's worth a try. FireStep doesn't support linear delta yet, but if you want to write that in Javascript,

I could plug it into firenodejs.

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

To post to this group, send email to fire...@googlegroups.com.
Visit this group at https://groups.google.com/group/firepick.
For more options, visit https://groups.google.com/d/optout.

--

Rich

[Peter Shabino]

The thing that lets you do multiple layers of feeders is the fact that the arms fold over the feeder head when you go to one of the extremes. So if you move to almost one of the extremes you should be able to put another layer of feeders just above the highest point of the head + arms (give yourself ~1" so you can pick the part of the pocket then move the head out with out bumping into anything. 

This would work on the FPD as well in theory but the work envelope is so small there is just not enough space for multiple ranks. The other method would be to stack the feeders like organ keys. That way the first row is level with the base and close. The next rank is slight up and back as not to overlap the pickup point of the first and so on. Downside to this is a very custom multilevel feeder. 

Once the holidays are done should be able to work on stuff more. Due to other work have not been able to follow things much. How is calibration going? Douglass any progress on the I2C temp code? 

Later,
Peter

My projects:
http://www.wire2wire.org/

---

From: fire...@googlegroups.com <fire...@googlegroups.com> on behalf of Rich Obermeyer <rich.ob...@gmail.com>
Sent: Sunday, December 13, 2015 2:00 AM
To: fire...@googlegroups.com
Subject: Re: [FirePick] Re: rotary vs linear deltas

 

[Karl]

Peter, calibration is coming along well. The introduction of LPP paths seems to have locked in acceptable precision for FPD.

I'm currently working with Andrew on TinyG and cartesian kinematics for firenodejs.

After that, it's back to accuracy testing for FPD. 

Enjoy your holidays :D

[Douglas Pearless]

I2c temp code has been put on hold until Holiday period as I have been working crazy hours for my clients and have literally no spare time for another 2 weeks :-(

Sent from my iPhone

On 14/12/2015, at 6:27 AM, Peter Shabino <wi...@hotmail.com> wrote:

The thing that lets you do multiple layers of feeders is the fact that the arms fold over the feeder head when you go to one of the extremes. So if you move to almost one of the extremes you should be able to put another layer of feeders just above the highest point of the head + arms (give yourself ~1" so you can pick the part of the pocket then move the head out with out bumping into anything. 

This would work on the FPD as well in theory but the work envelope is so small there is just not enough space for multiple ranks. The other method would be to stack the feeders like organ keys. That way the first row is level with the base and close. The next rank is slight up and back as not to overlap the pickup point of the first and so on. Downside to this is a very custom multilevel feeder. 

Once the holidays are done should be able to work on stuff more. Due to other work have not been able to follow things much. How is calibration going? Douglass any progress on the I2C temp code? 

Later,
Peter

My projects:
http://www.wire2wire.org/

---

From: fire...@googlegroups.com <fire...@googlegroups.com> on behalf of Rich Obermeyer <rich.ob...@gmail.com>
Sent: Sunday, December 13, 2015 2:00 AM
To: fire...@googlegroups.com
Subject: Re: [FirePick] Re: rotary vs linear deltas

 

How about putting a square frame the same size as the FPD into a typical linear printer (see below) and you should be able to attach most of the items that could attach to the FPD.  A few blind spots but it should work.

I am just waiting for a good feeder that can be connected to it to finish it off.

Then openpnp is suppose to do the rest.

I need to try firenodejs to see how precise it can be.

I would be interested how multiple levels of feeders could work.  My printer has 16 inches of height!

<image.png>

[Reef Morse]

For linears you now have to calibrate for actual angle and distance from center, rod lengths, offset from axis to rod ends for rails and end effector. Software exists to do that automatically in Marlin.

[Peter Betz]

Hi Karl,

The work you are doing is amazing, and it sure seems to be critical to moving forward. The message below seems like a notable achievement, so congrats!

Do you believe that accuracy will be an easier task than precision? I am trying to make time to finish the calibration kit and perform calibration on my FPD.

Peter.

[Karl]

Well, considering that precision was beastly hard (who'd ever thought of something as bizarre as LPP would be needed?), yes, I think accuracy is no

harder than precision.  

[Peter Betz]

Excellent! Exciting times....

Is my assumption of LPP correct in that you achieve precision by taking a predetermined path to (among other things) load up all the joints in a repeatable manner? This essentially compensates for looseness and backlash etc?

Peter.

[Karl]

Perter, that's our hypothesis. Similar treatment can be done for cartesians by "always come in from the left" for example.