A-axis, can't find hole problems

[soren]

me and my collegue are both students and have recently been working on calibrating the litplacer. However we are stuck with some problems. if we try to calibrate the needle, the A motor won't work properly. When it's calibrating it just wobbles the whole time and not turning like it should do. When the calibration is done it go's to find the FID's on the PCB. our setting are done correctly buth sometimes it find's the FID's and sometimes it doesn't and we get the message can't fin circle. Sometimes when it finds the 4 FID's it trys to find the components buth it can't find the reference holes. anyone that has the same problems that could help us ? maybe it's just our settings that are wrong. And an other little problem is that if we set the PCB zero and click go to is does it correctly buth if we try it again the day later, it goes to a complete other location.

best regards

[soren]

if we try the x driver on A motor its still the same problem. If we try the A driver on X motor it works perfectly. propably a problem in the A motor, can we fix this with the liteplacer programm ?

[JuKu]

Hi,
The issue is likely one of these:
- The tube does not rotate freely. It should be easy to turn by hand.
- The wiring for A motor is not good.
- The current setting for A motor is not enough. But beware, the A motor is small, TinyG can send enough current to kill it. Monitor the temperature of the motor when adjusting it.
- The motor is already dead. Does it run hot? What is the resistance of the coils? You should see about 10 Ohms between green and black as well as between blue and red and no connectivity to other wires.

Could also be the TinyG board, but you already checked that it can drive X, so this is not the issue.

[mrandt]

Hi soren,

I believe you have either a mechanical or electrical problem on the A-motor.

Try to remove the belt which drives the hollow shaft with the needle and see if the motor turns correctly.

If it doesn't, measure the coils with a ohmmeter and check your wiring.

If it turns without issues, check if the tube hollow shaft rotates freely. Mount the belt again and don't overtighten it!

Lastly, you *could* increase the motor current on the TinyG with the small trimmer, however the small motor should not require much current.

For the computer vision problems, you should try to improve the lighting and / or cameras. Are you using the original cams from the kit?

Also, you have to play with the filter settings until you get it right. Juha's recommendations in the instructions are a good starting point.

Good luck and have fun!

[JuKu]

> if we set the PCB zero and click go to is does it correctly buth if we try it again the day later, it goes to a complete other location.

Have you homed the machine before setting the zero location and before trying to go to the location again? The locations are saved when exiting the program, if the program crashes or it does not have writing permissions to its own directory, it doesn't get the saved value back. If it does, it is shown in the location box.

> When the calibration is done it go's to find the FID's on the PCB. our setting are done correctly buth sometimes it find's the FID's and sometimes it doesn't and we get the message can't fin circle. Sometimes when it finds the 4 FID's it trys to find the components buth it can't find the reference holes.

These are video processing issues. To be able to help, I need to see your tape and fiducials video functions settings as well as some screenshots with and without processing. Also, please review pages http://www.liteplacer.com/camera-setup/ and http://www.liteplacer.com/setup-downloo ... rocessing/.

[soren]

we measured the resistor of the coils and we can see that the motor is broken. to find the components we use functions treshold and invert, do we copy this to components or to paper tape. cause we are still having problems finding components.

[mrandt]

we measured the resistor of the coils and we can see that the motor is broken. to find the components we use functions treshold and invert, do we copy this to components or to paper tape. cause we are still having problems finding components.

OK, so I assume you will replace the motor?

What exactly is the problem with the part detection? Are you trying to pick parts from cut tape strips or individual, loose components from the table?

For picking parts from tape, the software detects the sprocket holes in the tape - not the parts. It then uses an offset defined by the tape config to find the component center. As tapes come in white / grey paper, clear plastic and black plastic, there are different filters to be configured for each sort.

Assuming you are picking from cut tape:

Make sure you have sufficient contrast on the holes. If using black tape, it might help to put a white paper strip or similar underneath the holes. If using white paper tapes, put something dark below.

Move the camera to the tape location, switch to camera configuration and enable the "show circles" checkbox. Next, configure filters so the holes get detected (orange circle marks). Once satisfied, copy the filters to "Paper Tape" (or whatever tape you are using) by pressing the "Copy here" button in the respective section.

Finally, set up a tape location, choose the tape type (thereby referencing the correct filter settings) and set 1st hole position.

Should work then - if not, copy the log here for further debugging.

[JuKu]

we measured the resistor of the coils and we can see that the motor is broken.

Send me a PM or email with your contact info, I'll replace the motor.

[soren]

we did everything lik you said and it works, thanks a lot for the help ! much appreciated just ons small matter. If we test it with 1206,0805 packages it picks it up. Buth with 0603 it misses the component with a few mm. What can we do to make this perfect ?