I am getting my M4 setup after building the frame. I ran into some small issues because I place my spoil board vertically off-center (18 in from the bottom) and stopped the machine mid calibration.
Now I get errors when opening the web interface that are asking if the FW is correct. After a few retries it will move on, but it takes a few minutes. I have updated the firmware repeatedly and the index.html.gz and Maslow.yaml files. Rebooted, left unplugged for 5 minutes, the whole thing has been done a few times over.
Iām not sure what might be going on. Did you get through a successful calibration?
If not, you might start with a fresh (stock) maslow.yaml, (edited with your frame anchors size / calibration grid width ā note the grid width is from the ācenterā of the frame, not the spoilboard) and try that again. Note too that during this calibration, donāt connect more than one browser to the maslow as this has caused issues for me. The actual calculations for the calibration are done in the browser, so if you have a faster machine you can use for the browser wherever your maslow is (just for the calibrationā¦) that is best.
I would suggest connecting directly to the maslow. Itās possible that something
on your home wifi is causing you grief (either a poor connection, a DNS issue, a
conflicting IP address, etc)
you havenāt mentioned disconnections or heartbeat failures, but a networking
problem is the only thing that I can think of that would cause long delays in
responding to anything with a freshly updated machine.
Thanks for the suggestions. I had already started down the road of erasing and reinstalling this morning, so I finished that, but kept it off my home WiFi. It has been much better so far and I think the WiFi must have been the issue.
Thanks for the suggestions. I had already started down the road of erasing
and reinstalling this morning, so I finished that, but kept it off my home
WiFi. It has been much better so far and I think the WiFi must have been the
issue.
from the reports, it seems as if a combination of the small wifi antenna on the
esp32 board, and the noise from the electronics (not even requiring the router
to be turned on) is resulting in very short range for the wifi link.
@bar, it may be interesting to test a version with a ESP32 module that supports
an external wifi antenna and see if thatās better (depending on the noise
source, it may or may not be) and if it is, offer it as an option
Ok, so the wifi issue is now resolved. If I want to have it on wifi, I will move the router closer.
After getting through calibration everything was great, I was able to move the machine around.
However, I then loaded up a simple file that I put together and was going to watch the machine move. I hit the SD not initialized error at that point.
If I try to move the machine using the jog buttons, it does the same thing, it spins up the fan, acts like it is moving (I see the x and y values changing), but nothing moves.
I tried g-code generated by kiri:moto and Carbide Create to try to rule out a bad file.
You said a power off fixed the sd card errorā¦ so after that did you take it down, retract/extend, re-hang and apply tension?
Unfortunately, this process must happen every time you power down, but Iām hoping once I learn more I can contribute belt saving to flash so if you are purposely powering down, you can resume where you were. Iām not there yet, and I donāt know where on the list of things to do this is for Bar
Iām hoping once I learn more I can contribute belt saving to flash so if you
are purposely powering down, you can resume where you were.
As a point of reference, the original maslow would save itās current location
after being idle for 2 seconds to avoid recalibration (which was even worse on
that version, faster, but much more work)
I think a prompt on power-up to accept the current belt lengths (or two buttons
to turn off the alarm, one of which being that the current position in correct)
would work.
I believe that grbl/fluidNC also has a āthe system has not been homed since
power upā state internally. if thatās the case, then a homing cycle that
prompted āI think I am here, accept or notā with ānotā doing a retract/extend
would be a good option.
(also, while you are mucking in this area, letās see about being able to add a
configurable extra length to each belt to allow for belt extensions, or for
stops to be placed on the belts so a full retraction would not be needed)
Thank you both for your help with this. I was able to get through everything and make my first successful test cuts a minute ago.
I am now on my way to executing a project that my wife has wanted for a while but I was not able to do with conventional woodworking (a scallop tile backsplash done in birch plywood).
