Successfull test --> Duet3d --> rotary delta

Hi to all.

I spend whole day testing DC_42 suggestion on reducing acceleration and he was wright.
The robot was bouncing and moving without control before that but now i can control it.

HW: Duet3d Ethernet board, 1,8 degrees nema 23, 3A stepper motors.
Weight of arms and head is 100 grams.
Main power supply: 24V / 10A

Btw.am using rotary design on linear delta firmware, so this could be also a part of the problem
and i dont know if this is effecting stepping calculations...

Updated video of robot moving on X axe :
[ufile.io]

This is the code that am using:
G28
G00 X35
G00 Z-5 ;HIGH CORRECTION
G4 P100
G00 Z0
;G4 P50
G00 X0
;G4 P50

G00 X-35
G00 Z-5 ;HIGH CORRECTION
G4 P100
G00 Z0
;G4 P50
G00 X0
;G4 P50

Problems:
--> i have tested bunch of options and it looks like that this is the maximum speed that i can get before robot starts to "loose himself".
--> i hope that this can be tweaked somehow because if this can not go faster than i must change the motors and add external drivers
--> i noticed that robot starts to loose himself when head is near or over maximum radius and sometimes it looks that it has not enough power.
--> speaking of that, could it be that the speed of the robot is reduced just because of lacking some stronger external power supply ?
--> i noticed one additional problem in my setup. Each time when head make small press on surface, robot looses his coordinates and
after that he starts to act fully out of control. Becuse i have to pick up a plastic bottle, this can be a problem... is there some auto calibrating option in such cases ?

Arnix

Edited 2 time(s). Last edit at 07/18/2018 12:33PM by Arnix.

I have replied to your similar post on the Due3D forum. In summary, you need to change the mechanics so that the motors rotate much further for the same amount of effector movement, either by altering arm lengths or by using gearing..

---

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].

I wanted to see your video, but the site asked me to remove my 'adblocker'. Sorry, NO WAY!

Power supply is not a problem. You have 10A power supply now. And you need at most 3*3 = 9A for motors. It is very easy to check whether it is your power supply. Connect a voltmeter to it while motors are moving and check whether the voltage drops below 24V.

Gearing will help. It will also increase your resolution which is probably rather bad now.

I could not watch the video as well. And no way in hell I'm switching off my 'adblocker'.

Thank you all for your reply's !

@DC_42
- i read it, thank you for explanation. I will keep you updated.

@O_lampe && Hercek
-guys, you dont need to disable adblocker.
Under red warning button you will see red link with "continue without disabling", click on it and then on slow download.
Download will start and maybe some notification window will pop up. It's free and secure site, dont worry.
Btw. my add block is also active...

Gearing
OK. My gearing system will be ready next week so i will send some updates.

@all
- after successful X axes test am planning to draw a square but it looks am missing something
because after robot reaches end length on X axes and when he needs to start to move on Y axes, it just "drops" and starts to move in all directions.
Like described in original post. Maybe this is also problem with gearing... Anyways, i will try again it and let you know.

Arnix

OK, I saw the video.

You will use the gearing well. Your 1/16 microstep size is around 0.56 mm now. That is a lot. Hopefully your gearing is at least 10:1 so that you get somewhere to 0.056 mm per microstep.

Your 1/16 microstep size is around 0.56 mm now. That is a lot. Hopefully your gearing is at least 10:1 so that you get somewhere to 0.056 mm per microstep.
*** funny that you mentioned that :-) I was thinking about that before day or two, but didnt take much notice on this.
Gearing is 10:1...

Btw. am using x32 micro stepping so this should be even worse :-) ?

M350 X32 Y32 Z32 I0 ; Configure microstepping without interpolation
M350 E32 I1 ; Configure microstepping with interpolation
M92 X16 Y16 Z16 E663 ; Set steps per mm // DEF 80
M566 X2500 Y2500 Z2500 E ; Set maximum instantaneous speed changes (mm/min) // DEF = 1200
M203 X14000 Y14000 Z14000 E0 ; Set maximum speeds (mm/min) // DEF == 18000
M201 X100 Y100 Z100 E0 ; Set accelerations (mm/s^2) // DEF == 1000
M906 X2100 Y2100 Z2100 E800 ; Set motor currents (mA) and motor idle factor in per cent
M84 S15 ; Set idle timeout

A.

Quote
Arnix
Btw. am using x32 micro stepping so this should be even worse :-) ?

It will be better with x32 microstepping of course.
Holding torque at the microstep boundary will be about half of the torque for x16 microstepping). But your overall holding torque will be increased by about 10 times due to the gearing. You may be OK as for as precision goes. It will be probably somewhere around 0.028 mm. No clue whether it is enough for you. If it is not or if holding force of the steppers will not be enough (heavy arms/platform or too high accelerations)) then you will need even more gearing.

There was i typo, but you notice that :-)

then you will need even more gearing
*** i can add bigger gear no problem.

After you have see the video and after you know my setup main question is:

Would those motors, with 10:1 gear system ( or bigger ) and with extruder/arms weight under 200 grams,
able to archive this speed : [www.youtube.com]

This is the most important thing for me because all the rest can be changed in day or two, but ordering motors and
constructing the base will take some time...

Arnix

From your post in the Duet3D forum, I has the impression that you don't yet have the 10:1 gearing installed. Can you confirm that? If so, then I suggest you try your existing motors with 10:1 gearing before you consider using different motors.

---

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].

No, gearing will be ready next week.
Am just impatient to define final HW so i can change and set final design.
You can call it professional defect :-)

Edited 1 time(s). Last edit at 07/20/2018 01:11PM by Arnix.

Gearing will help a lot even with the motors you already have. You may have precision problem at higher accelerations (ringing). If so then just increase gearing ratio 2 or maybe 4 times. Or stronger motors.
I cannot tell you how your device will behave at high speeds and accelerations (how much ringing will be there). You would need either to simulate your printer in CAD or describe it with equations and solve them.

I don't think ringing is an issue for a 'pick and place' machine? Especially, when 'place' only means: dump the part in the left or right container.

Edited 1 time(s). Last edit at 07/21/2018 02:40AM by o_lampe.

I dont know what this ringing means :-), but i noticed one interesting thing.

When the acceleration is low, robot draws straight line but as soon as i increase acceleration speed,
robot starts to draw "waves". He is moving on X axes but he moves in waves. Or to be more exact,
in half sinusoidal path.

If i increase speed further, it will make bigger waves / faster moves, until it looses it :-)
I guess this is also a problem with counting steps and gearing.

A.

Ringing is the platform vibration which happens because of elasticity in the system. You can see it on corners of 3dPrinted parts: [reprap.org]

What is the wave length of the waves? If it is smaller than 16 mm (while you do not have the gearing there) than it may be related to weak holding moment of the steppers. You will see whether the wavelength gets 10 times smaller when you add the gearing.

What is the wave length of the waves?
** Its hard to measure this but i would say that they are more around 2,5 cm.

With this setup:
M350 X32 Y32 Z32 I0 ; Configure microstepping without interpolation
M350 E32 I1 ; Configure microstepping with interpolation

M92 X10 Y10 Z10 E663 ; Set steps per mm // DEF 80

M566 X2500 Y2500 Z2500 E ; Set maximum instantaneous speed changes (mm/min) // DEF = 1200
M203 X14000 Y14000 Z14000 E0 ; Set maximum speeds (mm/min) // DEF == 18000

M201 X100 Y100 Z100 E0 ; Set accelerations (mm/s^2) // DEF == 1000

M906 X2100 Y2100 Z2100 E800 I30 ; Set motor currents (mA) and motor idle factor in per cent
M84 S10 ; Set idle timeout..
---------------------------------------------

The part that i dont understand is the height movement.
E.g. if i try to draw a line from X30 to X-30, extruder lowers the height for about 5 cm.
In other words, hight is dropping in linear "mode" after he starts to move from 30 to -30 position.

I didnt add any Z coordinates, just plain movement:
G28
G00 X30
G4 P500
G00 X-30
G4 P500

Has this something to do with gearing as well or is this because of missing rotary firmware ?
Btw. this X30 is equal to 28 cm.

------------------------
I have dismantled the robot so next two days i would not be able to make tests.
Regards !

A.

It will not work without firmware supporting your machine type. You need to get the proper firmware support first.