Maslow Home Maslow Community Garden Newsletter

After updating firmware and gc to 1.27 my squares are now parallelograms. see pic

First time I cut a horizontal line it dropped 1/4 in over 4"

Then it dropped 3/8" over 4"

Then it skewered on the Y axis as well almost an whole inch off x and y!

After this it erased my ini file (which I did not think to make copy of), So I manually entered my config data and it gives unexpected sled movement error.

If I have to manually reconfigure this it will have been the 6th time. I am beginning to get frustrated as I have been trying to get this to work since 7/15.

This started out as me just trying to erase a unwanted line on a sign that kept appearing. Now it is unusable at all.

MxLinux
ver 1.27 gc and firmware

I just noticed that v1.27 incorporated a new kinematics method. I think it requires a recalibration to make it work.

1 Like

Ok. Doesnt please me that I need to recalibrate. But I appreciate they continue to try and make this better.

1 Like

I think if something is changed that force people to recalibrate, then it should be noted. Also, I’m a big proponent of giving people a means to not use the change until they are ready to recalibrate. We had a discussion regarding this in the holey triangular calibration thread but I was under the impression that the kinematics equations hadn’t been incorporated into a release. But I looked just today and saw a release was made last week and those changes were included. So, I can’t guarantee this is what cause the problem, but I suspect it is.

1 Like

I just looked at the change log. My pull request is not included. Not sure how that got in there.

I think the sled weight and chain elasticity pr was backed out.

I just looked. The kinematics updates are from @c0depr1sm, merged on Jan 29.

I am pretty sure the kinematics updates should not cause this much change. At most, there would be about 6 mm change in the corners and at the top center. However, there are differences between this and the Holey Calibration implementation. I have not validated this.

Oh boy. I just remembered. There is a pretty big issue that I had to fix a while back, not in the main fork (the issue is in the main fork, but the fix is not). The chain stretch parameter is communicated from GC in scientific notation, but the firmware is not setup to read scientific notation. Because of this, everything gets messed up. This issue came with the pull request from @c0depr1sm.

Can you point me to the specific PR?

Here are the Firmware changes.

So with the PR removed, this isn’t an issue>?

You would have to remove the PR on the GC side, as well.

Sorry for being slow… which specific PR in ground control needs to be removed? I’m just not following.

I couldn’t find it either. I thought there would be a corresponding GC commit, but I guess not.