× Forum Header

stepConf and actual running precision issue

More
06 Sep 2016 18:55 - 06 Sep 2016 19:13 #80087 by fannlow
fannlow created the topic: stepConf and actual running precision issue
Hi there,
i need some help, for something which i might have missed out.

I was running StepConf for a 4-axis hotwire cutter.
from the options, i have opted for XYZA (since we don't get XYUV) and mm for measurements.

motors per revolution : 200
driver microstepping : 5
Pulley teeth ratio : 1:1
Leadscrew Pitch : 82.5mm/rev
Max. Vel : 508mm/s
Max. Accel : 1016mm/s

Home Location : 0
Table Travel : 0 to 2650mm
Home Switch Location : N/A
Home Search Velocity : N/A
Home Latch Direction : N/A


during StepConf, I have some issues, where X and Z uses the same pitch, stepping and rev.
however, for X, the distance travelled is precise, but on Z, somehow the distance travelled seems more.

on the other hand, for Y and A, its the same, while Y is precise while in StepConf, A is travelling more.

so without any changes, I accepted and use XYZA in linuxCNC, but when I get inside the program after homing, I get some issues where X is now travelling more than it should be. while Y is also travelling more than it should be, but A is travelling like at least 6 times shorter.

but when I up the scale of A to 6x the original scale, it is still short of like 25% travelled.

is there an issue with my software or setup or is there something which I have missed in all the confusion?

i am uploading the .hal and .ini files for referencing.
- updated... i jz found out that my .hal and .ini files are blanks!
- I will upload it again later.

thanks for the help very much! =)
really appreciate it
Last Edit: 06 Sep 2016 19:13 by fannlow.
More
06 Sep 2016 20:49 #80091 by tommylight
tommylight replied the topic: stepConf and actual running precision issue
Something is not right with your numbers, pitch of 82.5mm is only if using pulley's, or rack and pinion.
Velocity of 508mm/s is 30.480 meters per second, so it is way to high for any hobby machine. Also acceleration of 1016mm/s/s is to much as it would try to get to the speed you set (over 30 meters/s) in half second.
Try setting as follows
Velocity 50 to 100
Acceleration 100 to 200
and be sure to check the pitch.
More
07 Sep 2016 05:18 - 07 Sep 2016 05:30 #80114 by fannlow
fannlow replied the topic: stepConf and actual running precision issue
hi, I checked it.

the issues seems to be when I changed
4 axis XYZA to 9 axis XY ZA BC UVW.

for some reason, FYI is running according to distance quite precisely.

but V is running way short.
I ran G1 F5000 V-100. the actual measured is only 15.5mm

any ideas where gone wrong?


or...

do i need to add these axes for my .HAL?
thanks
Last Edit: 07 Sep 2016 05:30 by fannlow.
More
07 Sep 2016 07:25 - 07 Sep 2016 07:26 #80117 by fannlow
fannlow replied the topic: stepConf and actual running precision issue
Hi,
I did fixed the issue of scaling for all other axis, except for my V axis, which now travels more than it should when moving to the negative side , and lesser than it should when moving on the positive side.

The changes I made is in my HAL file for the axis, now all change to corresponding axis in .ini file.

eg:
G1 F5000 V-900

results in a travel of -934 or somewhere near.
then i use G1 F5000 V0

it travels back not to 0, but 34mm short of 0.

not understanding what is happening.

i am uploading my hal and ini files for help.

-=======-
also i keep getting bad character 'z'. while starting up linuxCNC and also while touch off.

my Y and V are using same stepping of 2000 now
which means at the stepConf, i use 200 for rev and 10 for microstep

Y axis is OK, nothing is wrong and it travels quite well.

my 5 start ACME screw is 12.7mm per rev from the supplier.
leadscrew ratio is 1:1, its a direct stepper to acme screw.

===============
last thing, i still can't get the view to work like in the sim-9 axis view.

hope someone can shed some light for me, but priority given to my travel issue :( :( .

thanks
Attachments:
Last Edit: 07 Sep 2016 07:26 by fannlow.
More
07 Sep 2016 08:49 #80120 by fannlow
fannlow replied the topic: stepConf and actual running precision issue
updated some minor changes in order for linuxCNC to work.

got rid of the bad character 'z'.

i ran stepConf again to double check.
and all the axis ran the correct amount of distance it should travel.

all Driver Rev is 200
all microStep is 10

not sure what went wrong?
Attachments:
More
07 Sep 2016 17:33 #80144 by fannlow
fannlow replied the topic: stepConf and actual running precision issue
I have been at the machine the whole day. still no luck.

and this is weird.

I was using 2.7.5 I believe from 2015 and it worked all good.

then I had a wire not working and I changed it.
then I downloaded the 2.7.6 from 2016.

but I forgot all my previous setup and had to do some trial and errors.

I went with drive rev 200 and micro at 10 tiger 2000 per turn right? as the settings on the driver.

then I did an assessment with all the leadscrews and also asked my suppliers.

came out X and U are 82mm per one full motor turn (360°).
and Y and V at 12.7mm per one full turn.

and i inserted them into my stepConf but I didn't get an accurate distance travelled.

it turns out X and Y are accurate while U and V are not accurate. going to positive and negative both gives me different results.

example: I start from 0 and go to 100. it would stop around 70 or so. then I return back to 0. and it will either pass the 0 mark by a bit or short of it by abit.

however on linuxCNC it shows 100 on the axis and then 0 when I return it.

X and Y are both accurate with this same settings.

any one can pls shed me some light please?

am about to pull off all my hair soon...

will changing my driver steps per turn disturb this travelling accuracy?

thanks in advance
More
07 Sep 2016 18:48 #80156 by fannlow
fannlow replied the topic: stepConf and actual running precision issue
After another round of trying to correct the machine.

this time i will write down all the settings and the issues I face too

StepConf :-
=============
Base Information
=============
Machine Name : my-mill
Configuration directory : ~/linuxcnc/configs/my-mill
Axis Configuration : XYZA
Reset Default machine units : mm
Drive characteristics : (multiply by 1000 for times specified in us or microseconds)
Driver type : Other
- Driver Timing Settings
Step Time : 3500 ns
Step Space : 3500 ns
Direction Hold : 5000 ns
Direction Setup : 5000 ns

o One Parport
Base Period Maximum Jitter : 47000 ns

=============
Parallel Port Setup
=============
Outputs (PC to Mill): Invert
Pin 1: ESTEOP OUT NO
Pin 2: X Step YES
Pin 3: X Direction NO
Pin 4: Y Step YES
Pin 5: Y Direction YES
Pin 6: Z Step YES
Pin 7: Z Direction YES
Pin 8: A Step YES
Pin 9: A Direction YES
Pin 14: Unused
Pin 16: Unused
Pin 17: Unused

Input (Mill to PC): Invert
Pin 10: Unused
Pin 11: Unused
Pin 12: Unused
Pin 13: Unused
Pin 15: Unused

Parport Base Address: 0

Output pinout presets: Sherline (default)
More
07 Sep 2016 19:06 #80157 by fannlow
fannlow replied the topic: stepConf and actual running precision issue
=============
Axis Configuration - X and U - AXIS
=============

Motor steps per revolution: 200
Driver Microstepping : 10
Pulley teeth (Motor : Leadscrew) : 1 : 1
Leadscrew Pitch : 82 mm/rev (or is this rev/mm?)
Maximum Velocity : 127mm/s
Maximum Acceleration : 1016mm/s (or is it in s2?)
Home Location : 0
Table travel : 0 to 2650 (mm or is this in linuxCNC units?)
Home Switch location : 0 (default)
Home Search velocity : 0.05 (default)
Home Latch direction : same (default)

=============
Axis Configuration - Y and V - AXIS
=============

Motor steps per revolution: 200
Driver Microstepping : 4
Pulley teeth (Motor : Leadscrew) : 1 : 1
Leadscrew Pitch : 12.7 mm/rev (or is this rev/mm?)
Maximum Velocity : 127mm/s
Maximum Acceleration : 1016mm/s (or is it in s2?)
Home Location : 0
Table travel : -1550 to 0 (mm or is this in linuxCNC units?)
Home Switch location : 0 (default)
Home Search velocity : 0.05 (default)
Home Latch direction : same (default)

=================================

here is the issue:

1. while running "TEST THIS AXIS" on X-axis
Velocity : 25mm/s
Acceleration : 1016mm/s
Test Area : +/- 10mm

X axis seems to go slightly more than the intended 5mm from mid point (covering a total of 10mm distance), hence when I hit RUN again to stop, it doesn't return to the starting point, it will be short of a few mm.

using the same values on V axis, it will go about the same distance, and after a few continuous runs, it will start to move further away from the starting point. and so when I hit the RUN button to stop it, it will stop no where near the starting point.

for Y and U, amazingly it is quite precise, as after I test to run 100mm, it is quite accurate.

=================================
issue

2. so i tried with these settings and run on linuxCNC, with the edited HAL and INI files to setup U and V in HAL as well as for the 9 axis in INI.

I also corrected in HAL the axis for U and V, changing 2 to 6 for U and 3 to 7 for V.
in the INI file, I simply copy the information from X and paste it for axis 2, 3, 4, 5, 8.

in linuxCNC, i ran G1 F5000 X100, and it went off the chart, with way more distance, and when I do G1 F5000 X0, it will over shot the 0 point in distance.

However, in linuxCNC, it indicates that my X has indeed travel 100mm and back to 0 again, quite different from the actual physical distance traveled.

the same can be said for my V axis, it will go almost 100mm but it will not return to 0 point. It will be a few mm short of 0mm.

again, Y and U axis are fine and the motors also sound smooth and normal to me.
X and V motors are more aggressive in terms of their sound and movement speed.

=================================

I hope this can shed some good light into what has happened and how confused and somewhat frustrated I am, as I am depending now on this to finish up my work.

hopefully kind soul here sees this and is willing to help me get on with my linuxCNC.

thanks again,

if there's any questions, please ask me right away, cause i really need to get this done and fixed.
More
07 Sep 2016 20:32 #80171 by tommylight
tommylight replied the topic: stepConf and actual running precision issue
List your hardware, how are motors mounted and to what, what drives, anything and everything, pictures would help.
I understood the part with 12.7mm pitch screws, but why is the "82mm" on some axis and on others 12.7, lower the acceleration to 100 and test with that.
Do a test with velocity at 10 and acceleration 100, see if that is ok, then you can test with higher values.
The following user(s) said Thank You: fannlow
More
07 Sep 2016 20:42 #80174 by andypugh
andypugh replied the topic: stepConf and actual running precision issue
Looking at the last INI file posted, the SCALE parameter for the V axis is completely different to all the others.

Just edit that by hand. Stepconf might be trying to scale for degrees not mm.

In fact, just accept that Stepconf is now finished with. You should now deal directly with the INI and HAL files.
The following user(s) said Thank You: fannlow
Time to create page: 0.082 seconds
Powered by Kunena Forum