FT300 path recording

jianlan_

Hi everyone,

I started using FT300 with ActiveDrive URCaps several weeks ago. There two issues quite often.

I am using FT/300 + two finger 85mm gripper + UR5.

1.   even when the gripper is not holding anything or in contact with objects,  ActiveDrive throws out erros "excessive force is applied to the tool"

2.   in the path recording mode,  the robot fights against me when I try to move it, often resulting in singularity that quits force mode.  Also, even I apply every little force, the robot tries to move to some singularity points.    

I wonder any of you encountered this, am I expected to do additional setups before using path recording? Thank you so much!

David_Levasseur

Hi @jianlan_, I opened a support ticket to help you with that. I will post my findings here afterwards.

David_Levasseur

Thanks to the video @jianlan_ sent, we noticed the sensor was not properly aligned. For the sensor to work as expected, it absolutely needs to be mounted it in the exact same orientation as the example on the picture below. If the provided dowel pin is installed, there is no other possible orientation. Then, it is required to set the COG, TCP and payload. Once this is done, the sensor needs to be calibrated.

jianlan_

@David_Levasseur  Sorry for the delay,  thanks for the ticket.  I answered the ticket by replying your email. Please let me know what you think, thanks!

Martin_B

I'm having an issue where the UR is fighting against me when I'm trying to do a path recording.
I have double checked that the dowel pin is fitted (as described previously), I have also re calibrated many times.
I have also double checked we're using the latest version of the software etc.
This still hasn't helped me. Sometimes the robot will drift, sometimes it will fight against me, it seems to be very inconsistent.
Is there a setting somewhere that I'm possibly missing?
Any help would be appreciated.

David_Levasseur

Hello @Martin_B, Could you post a video of the issue you are describing please? Also, what is the serial number of you sensor, version of URCap and version of Polyscope?

Martin_B

Hi David,
Here's a Dropbox link to all the information you requested: https://www.dropbox.com/sh/71u262psmzd2jz9/AABYyQRWs4jNddqqco7A4lCja?dl=0
I also have another question about ActiveDrive - is it possible to 'fix' the Z height whilst recording a path?
I would like to ideally limit motion to X and Y and have the Z at a set value, just wondering if the FT300 has this option.

Annick_Mottard

Hello @Martin_B
I looked at your video and the drift that you experience is normal. When in ActiveDrive mode, if you stop applying a force, we don't want the robot to go in a sudden stop. We programmed the ActiveDrive so it would decelerate slowly to avoid any faults. 

In order to fix the height in Z, you can use ActiveDrive's plane mode:

 
It will force the robot to move on a plane (x-y). 

Martin_B

Annick_Mottard said:

Hello @Martin_B
I looked at your video and the drift that you experience is normal. When in ActiveDrive mode, if you stop applying a force, we don't want the robot to go in a sudden stop. We programmed the ActiveDrive so it would decelerate slowly to avoid any faults. 

In order to fix the height in Z, you can use ActiveDrive's plane mode:

 
It will force the robot to move on a plane (x-y). 

Hi Annick,
What you can't see in that video (and I'm not sure how to show it), is that whenever I'm trying to pull the robot in a particular direction, it is fighting me, sometimes quite severely. To the point where it will actually pull in the opposite direction to the one I'm trying to get to.
We simply wanted to trace out a basic square, but can never get the robot to follow the path because it fights to go in another direction.

Regarding my other question; I have selected the Plane function but I am still able to move the robot in the Z axis, which is an even bigger concern at this point in time. I have a customer demo on Monday morning and I need to get this resolved before then if it is possible. Are there any other settings that can affect the locking of the Z axis?

David_Levasseur

Hello @Martin_B, Does the sensor "fight" you only in the X axis by any chance?

Martin_B

David_Levasseur said:

Hello @Martin_B, Does the sensor "fight" you only in the X axis by any chance?

Hi David, I'm going to try and make a video tomorrow morning, showing the exact problems we are having. I think I now have a way of showing it correctly. It is mostly in the X axis, but not totally. We are also getting unexpected reactions in the Y axis.

Martin_B

@David_Levasseur Here is a link to the latest video. When we try to move in active drive, it is making some very strange moves when we try to move in X and Y only (it seems to combine both). In addition to that it is much harder to move in the +X direction than the -X direction.
Note, we performed a brand new calibration immediately before shooting the video.
https://www.dropbox.com/s/vuq854auwr3yfiv/Robotiq_FT300_Issue.MOV?dl=0

David_Levasseur

Thank you for the video @Martin_B, I sent you a firmware update in private. Please give it a try and let me know how it goes.

Martin_B

Thanks for that @David_Levasseur we managed to perform the firmware update but now have an issue with calibration.
We are unable to successfully validate the calibration for some reason. We get an error message that reads "Could not retrieve or parse calibration results."
We have turned the system off and on several times to reset everything and have tried to calibrate dozens of times with the same result. Any idea what we're missing?

David_Levasseur

Hi @Martin_B , I wish I would have thought of that earlier but since you are now running a previous version of the firmware, it is not fully compatible with the latest URCap. The last compatible version is UCS 1.1.1. Unfortunately, this older version does not include the calibration procedure and all the new features of the latest release. The calibration has to be done like before by loading the calib_tool.urp program. The only other way around would be to send the sensor back to Robotiq for a major update. Let me know if you want to proceed with the return and I will issue a RMA number for you.

Martin_B

David_Levasseur said:

Hi @Martin_B , I wish I would have thought of that earlier but since you are now running a previous version of the firmware, it is not fully compatible with the latest URCap. The last compatible version is UCS 1.1.1. Unfortunately, this older version does not include the calibration procedure and all the new features of the latest release. The calibration has to be done like before by loading the calib_tool.urp program. The only other way around would be to send the sensor back to Robotiq for a major update. Let me know if you want to proceed with the return and I will issue a RMA number for you.

So just to be clear....
If we run the calib_tool.urp program then everything should work fine?
If that is the case, we would like to keep the unit and wait until you release a new URCap.

If, however, you think it would be better to just send the sensor back anyway, then we can do that too.
We will comply with whatever you advise is the best course of action.

David_Levasseur

That should work for now but as soon as we release a new URCap with a firmware update prompt, the issue will come back. In my opinion, the best course of action is to send this sensor back for a long term solution. I will send you the paper work in private.

Martin_B

OK, thank you.