No reply from the console, just a new line. It does show in the settings and in the (i) info → Controller Settings, but is back to 1 once i disconnect and reconnect.
Let me check, might have flashed the master from Webcontrol
Edit: I followed the link from the readme, GitHub - WebControlCNC/Firmware at release/holey
Interesting. I see what happened and fixed the bug. Nothing wrong with your firmware.
The latest Makerverse version, 1.1.1-c, just finished uploading. You should have it if you restart. It also lets you edit the settings from the UI. Here you can see I clicked the “Save” button. Notice the “ok” in the console.
2 Likes
@zaneclaes - I’m working through the calibration process. I have the maslow classic with the holey-51.28 firmware installed.
I have done several iterations and keep getting similar results. The calibration in the lower left and right corners is off by ~30 mm. Here are the results.
Wonderful to be able to edit and save the values directly!
Small issue after new connection:
The blunt reason to set height and width of the sled was i was hoping i could use my precise grid on my acrylic sled, that was cut on desktop-cnc nice and ruined by not making a jig to round the bottom corner.
Did not realize that a radius is required for the non destructive calibration. Will overcome this as i think the ends of the grid can be seen on all 4 sides and will use that as a radius, unless i’m missing something. 4 holes set at user preference around the bit? The bottom view will be blocked for some
To much to ask, but for the record:
- is the sled radius used other then the Edge calibration
- what about the other values like centre of gravity
- chain sag correction
What are the values that are used, so i don’t drag old values that might be incorrect to a new setup?
Thanks so much! Getting close to take this to the workshop!
Kind regards, Gero
Edit: One more… The first time i open the connection it tells me to upgrade firmware. I close and open again and it’s fine.
- Are you on version 1.1.1-c (see settings)? Make sure you have prereleases on.
- Did you press the “apply calibration results?”
- If both are true, look at the console after applying. Do you see the settings accepted (“ok”?)
Aside from the obvious (motor width/height, chain tolerance left, chain tolerance right, chain length)… sled weight and rotation disk radius are the only others I can think of. You look like you probably need to change the rotation disk radius, given your attachment points (instead of the steel ring).
Sled radius is only used for edge calibration. I like your idea of using the grid. As long as the grid is the same distance from the bit on all 4 sides.
As for the settings that are still not saving… did you see them in the console with an “ok?” I’ll try editing those specifically when I get back to a computer.
All works if sled is at X0/Y0. If I move my sled 10mm up and then store the 120mm my sled moves to Y+ 270.480 mm and after dis- and reconnecting gives me rotation radius -8.84902000.
The same happens if I move the sled 10mm down and store the the 120RR. My sled jumps to Y+ 246.230 this time and after reconnect the RR is -8.84902000.
As long as i’m on 0,0 it seems to work. All time it is confirmed with an OK after a
position loaded at:
-0.00
246.23
Somethig to chew on tonight my Friend 
Going to bed. Hope i can chatch-up with how fast you are creating the updates.
Hmm. I was able to successfully edit/save the rotation radius, and see it re-appear correctly after re-connecting. I did find one small bug… possibly related, fixed in build #198.
Also, if you have not done so, I would advise “wiping” your device’s memory via the calibration modal. This reduces the chances that it still has some setting stored from a different firmware.
Good night, I’ll keep thinking on it
Still here, poured a glass more. Did you try that or save at X0/Y0?
Edit: Will wipe or better go for a new image flash and start there.
I did both. The fact that the Y position moves when you change the rotation radius is normal. The math just changed, so the position calculation is now different. What is abnormal (obviously) is that this setting is not being saved.
That the RR changes to -8… if I store off the 0/0 is strange. Will flash the new .img tomorrow to make sure no bugs are carried. Thanks!
Very. A negative value should not even be possible. It’s definitely not the ‘default value’ I would expect. That’s why I’m hoping a Wipe Settings via the calibration UI would fix it. Maybe the memory got corrupted somehow.
Ok, you got me, have a few more minutes. Stand by 
I’ll wait a moment as the green led on the RPi indicates the is writing in the SD going on…
Edit1: Heavy update?
Patch sizes are usually 200-300MB. I swear I’ll slow down releasing them once things get stable 
1 Like
ok, not tonight, my baby cat Twinkey is demanding to go to bed an i obey.
Edit: the Pi4 with 4mb should arive 26th this month. Hope to be of more help then. Have a good day/night.
Sleep well!
I don’t think this particular issue is related to the Pi, but newer hardware never hurts.
HUGE fan of the direct save addition!!
what are the $xx setting numbers for
1 Like
You can disregard those 3. I’m going to remove those three as “required” in the next release. Nothing actually depends on those three in particular. It’s just identifying the settings it knows about on the Mega which it can’t find on the Due.
1 Like