The maslow suddenly moved downward beyond the boundary when cutting. This was not what I expected. I only had time to turn off the power in an emergency. I want to know what causes this? Is anyone in a similar situation to me? I’m using estlcam and I’ll attach my gcode file for reference.
xiaocheche.nc (224.2 KB)
I’ve been cutting patterns in 4x8 sheets that go to ~3" from the edges in the form of 3-4 line segments that end at each edge.
I had the Maslow drive off the edge twice. I thought it was my g-code because each time it should have cut to the edge, then moved from there to another cut location. So I was wondering if it was missing the move command or similar. However, G-code looked fine in a web-based g-code simulator.
Realized that the variable that had changed was operating on the site Wi-fi instead of the Maslow hotspot. Connectivity had been fine on Monday, but wouldn’t connect on Tuesday. The range extender was moved and then I was able to connect. So wifi iffy to start with.
I switched back to using the Maslow wi-fi and was able to complete an entire sheet cut without issue (i.e. multiple excursions close to each edge).
TLDR:
=> Try connecting to Maslow when it is in AP mode (instead of it being client of your existing wifi).
=> Confirm your g-code is not doing something weird with a simulator/preview
=> We also edited the source file so that all the lines ‘ending’ at the edge are now ‘starting’ there, so the Maslow moves to the edge and then starts cutting inwards.
I didn’t look at your file, so your situation may be different! But if you are using home or other wifi, try switching back to have Maslow be the AP.
I’ll be doing some more today so will see if it happens again.
Thank you very much for your reply, it is very enlightening, it is late at night in China at this time, I will try cutting with ap mode tomorrow.
Derek Graham wrote:
TLDR:
=> Try connecting to Maslow when it is in AP mode (instead of it being client of your existing wifi).
=> Confirm your g-code is not doing something weird with a simulator/preview
=> We also edited the source file so that all the lines ‘ending’ at the edge are now ‘starting’ there, so the Maslow moves to the edge and then starts cutting inwards.
do you have some stock off the edge of the workpiece that matches the workpiece
thickness so that the sled is supported as you go off the edge?
what size frame/workpiece are you using (wanting to check if you are getting
into the area where the arms are hitting the uprights, and if so, ask if you
noticed any difference in that area)
thanks for the report
David Lang
I still don’t think we’ve got to the bottom line of whether the disconnect is causing it or the issue is causing the disconnect. In THEORY you should not need to be connected once the maslow is told to start a job.
Ron Lawrence wrote:
I still don’t think we’ve got to the bottom line of whether the disconnect is
causing it or the issue is causing the disconnect. In THEORY you should not
need to be connected once the maslow is told to start a job.
do we have a LED that we can blink as a heartbeat when the system is fully
running? something that would stop if it’s a matter of the control loop crashing
while the PWM driving the motors continues.
David Lang
Would the blue led that blinks the up address serve that function? Maybe the may control loop needs a “circuit breaker reset” that if not reset within a period of time signals a failure by no longer blinking?
That currently does blink in the main update loop for the maslow class, so if it is no longer blinking that would tell us definitively. That said, when it happened to me I was also way too focused on hitting my power off to watch for it, but if I see another I will watch for that before I hit the power
Ron Lawrence wrote:
That currently does blink in the main update loop for the maslow class, so if
it is no longer blinking that would tell us definitively. That said, when it
happened to me I was also way too focused on hitting my power off to watch for
it, but if I see another I will watch for that before I hit the power
this is a good reason why it would be useful to setup a camera to video cuts
right now. If something goes wrong we can look at the video to see if there are
clues like this (and so others can see what happened)
David Lang