Cool. I noticed that your dial is not the M5 dial from M5Stack. I hope that the firmware will still work, despite the fact that yours is apparently made by a different manufacturer, and may not have the same guts in it.
Hey, does the “other” dial you got have an ESP32-S3 inside it? If so, and it has a “button click” function in its rim, then perhaps the existing firmware could work for it, even though it’s a different brand than the one spec’d in the docs.
I screwed something up. I had a perfectly working LR3 using the jackpot controller before I set it up to use the FluidDial. I now can’t home my Z axis. I updated the FluidNC firmware to 3.7.13 and used the latest FluidDial firmware. I updated my yaml file to add in the uart settings as instructed.
What happens now is X and Y home and move as expected. Z on the other hand moves correctly until I try to home it. It moves down a bit and then stops. After that it won’t go back up until I restart the jackpot.
This happens whether or not the pendant is connected.
On the original config file from V1E, the “C” channel (axis) is commented out, but yours is uncommented and active. If that is intentional, this may not be a concern, if unintended, it may be a cause of issue.
These trailing spaces have been known to cause issues. They may or may not be a cause of issue for you.
On the newly added UART code, you are missing a space between “report_interval_ms:” and “75” — you have “report_interval_ms:75” instead of “report_interval_ms: 75” — and since spaces have been known to be important, this may or may not be a cause of issue.
The above should look like this:
Note: I used the “find” feature in VS Code to check for presence of spaces. That’s why the spaces show up highlighted in these screen shots.
Thanks Doug. I’ll correct those issues and see what happens. I don’t know how they got there other than the missing space in the UART section. I’ve only used the original file and made none of those changes you show. I use Notepad ++ for an editor.
If you would find it convenient, Notepad++ also has a convenient option to show whitespace characters, although spaces don’t show up as well as with the find feature.
Early on, there was not yet much awareness of the possible issues of trailing spaces or missing spaces. They may have been present from the start (here), and… they still may not be in a place that causes an issue (if they are from the original file, they may not be harming anyone). I unintentionally added a trailing space that caused one of my Z motors to move twice as fast as the other one!
It does have an ESP32 and rotary dial with push-switch. However it is also 2.8 inch screen, with a 3 inch overall diameter. It just woun’t work with any of the cases anyone else is using.
I may repurpose it for another project (home assistant controller) and just order the M5 dial.
It’s very low on the project priority at the moment. I need to finish LittleLR3 first.
Coming back to the issue of coordinates not displaying correctly on the DRO. The only issue I had was the Z axis showing strange numbers in-between polls. It didn’t effect usability.
Today I’ve mounted my Kobalt router and I can’t use the dial at all while it is running. The screen goes full glitch. The router power cord is no where near the controller, M5 dial, or any of the cables for the dial.
I noticed that too. I think it’s showing the machine coordinates not the workspace coordinates so when you set zero it’s for the workspace and the machine zero is unchanged
Such a strange occurrence that I’ve never experienced. I could understand if the router was wired into a relay off the controller but its completely separate. I even have a ferrite ring around the usb cord as well. Super bummed but I really only need the dial for initial setup before I turn the router on. I just don’t like unsolved things.
Would a shielded USB cord apply. I thought the same thing but that’s what I’m using.
That M5 dial is using a separate four pin cable, right? Not the USB you were referring to? I would be inclined to check that cable to see if it is shielded.
If I remember correctly, in the kobalt thread you mentioned that your router is surging a lot and the brushes smell. That probably means the router is also generating a lot of EMF which would mess with your low voltage signals.
Well, I found the problem. I fixed the yaml and tried it again with the same results. I then noticed a message in the terminal window after I saw that it was going into alarm. It indicated that it did not detect and endstop pulloff so I suspected a limit switch and sure enough one of the z endstop switches had pulled out just enough to lose contact at the jackpot. I am now printing your clamp block. It’s been on my to-do list for a while. And a bonus is that I got my yaml file cleaned up, lol. Everything seems to be working correctly now. Thanks for your help Doug.