Polar Printing: Polar Axis Rotation Issue

dc42

@reprapper , congratulations on being the first person I am aware of to use the Polar support in RRF!

I agree, it looks like a firmware bug. ~~Can you confirm that the centre of rotation of the bed is somewhere inside that square shape you are printing?~~ Edit: I just realised that's a stupid question, it must be inside otherwise it couldn't print a square without reversing the direction of rotation.

dc42

I found the problem. The main kinematics calculations have worked out that it needs to take the short route between -179 and +179 degrees, but the speed and acceleration limiter in the PolarKinematics class thinks it is going the long way round. So it's reducing the speed for that bit to about 1% of normal. Now I just need to fix it.

dc42

@dc42 said in Polar Printing: Polar Axis Rotation Issue:

I found the problem. The main kinematics calculations have worked out that it needs to take the short route between -179 and +179 degrees, but the speed and acceleration limiter in the PolarKinematics class thinks it is going the long way round. So it's reducing the speed for that bit to about 1% of normal. Now I just need to fix it.

Meanwhile, you could use this "feature" to calibrate pressure advance, i.e. see what value of pressure advance improves the print the most.

dc42

I believe this issue is now fixed and I have put a new firmware binary at https://www.dropbox.com/s/m0r3ldy424mxf4v/DuetMaestroFirmware.bin?dl=0. Caution: this firmware contains some unrelated changes that have not been fully tested, so use it with care.

reprapper

@dc42 Thank you for the amazingly quick response! The firmware binary fixed the kinematic issue.

Still lots of tuning and tweaking to do but the kinematics seem to be working correctly now.
Here's a short video showing what polar's do best, printing vases: https://drive.google.com/open?id=1T2Wjb3OejqWlkfOUjZREdYk8xY59UvRD

@fcwilt My homing files are the same as outlined in the polar setup guide: https://duet3d.dozuki.com/Wiki/ConfiguringRepRapFirmwarePolarPrinter

pir2z

@dc42 Hi David, I see Polar are the hight problem... I have exactly the same effect on the Duet Wifi, please could you find the same solution on it?
Thank for all.
Best regards
Pierre Zanetti

dc42

@pir2z said in Polar Printing: Polar Axis Rotation Issue:

@dc42 Hi David, I see Polar are the hight problem... I have exactly the same effect on the Duet Wifi, please could you find the same solution on it?
Thank for all.
Best regards
Pierre Zanetti

I'm sorry, I don't understand what problem you are having. Please explain.

pir2z

@dc42 Sorry for my english level... I met the problem explained above. When Y=00 , the printer does a long stop (no error occured). After many seconds all restart.
I hope I'm clearer now.
Thank a lot
Pierre

dc42

Try this firmware https://www.dropbox.com/s/fyvibzm0zl92hiy/Duet2CombinedFirmware.bin?dl=0

pir2z

Thank a lot I try it and I keep you in touch. Thank for all

pir2z

Hi David,
I tried the new version you sent me and unfortunately it doesn't work.
The starting problem is not resolved and the movements are very jerky now.
The movement does a "stop" when heater is arround Y= 0 (between -0.5 and +0.5) it's not very constant.
I would like to know if it can come from the mathematical problem between radian or degree, or a division by 0.
let me know what I can send you to help you to analyse my problem.
Best regards
Pierre

dc42

Please try version 2.03beta2.

pir2z

Hi David, Ok I'll try it and keep you in touch.
Thank for all

timcurtis67

@dc42 said in Polar Printing: Polar Axis Rotation Issue:

I found the problem. The main kinematics calculations have worked out that it needs to take the short route between -179 and +179 degrees, but the speed and acceleration limiter in the PolarKinematics class thinks it is going the long way round. So it's reducing the speed for that bit to about 1% of normal. Now I just need to fix it.

Just out of curiosity, Is it possible the issue relates to 180 degree moves? Would dropping or forcing 90 degree moves help with the long pauses?

Maybe none of this is related but back in "the day" a lot of program movement issues could be cured by forcing arc type move to be broken into 90 degree segments.

Again, disregard this post if none of this applies to the current issue.

dc42

@timcurtis67 said in Polar Printing: Polar Axis Rotation Issue:

Just out of curiosity, Is it possible the issue relates to 180 degree moves? Would dropping or forcing 90 degree moves help with the long pauses?

No. But I fixed this issue in firmware 2.03beta1.