OX CNC build and doing my first cuts. I can't seem to figure out how to complete a cut. Its randomly may going through the code and stopping mid job. Varies at what point s a me job. When it does the spdir light flashes red. I don't understand the console enough to interpret the issue. I don't see it physically hitting the limit switches my test cuts are only roughly 6"x6" yet it seems like it.
What you describe is commonly caused by parameter setting issues and/or electrical noise (what Zootalaws is referring to "induction")
A common source of electrical noise is the spindle motor.
You could try "running your job in air" first with the spindle off, then with the spindle on.
Zero your Z axis above the work surface, then with the spindle disconnected try jogging each axis from the command line or the CNC.js jog interface.
If work OK, try running your Gcode in air, first with the spindle off, then if runs try with spindle enabled(but still in air).
A parameter set would really help this discussion move forward.
Run $$ from the command console, copy the results to a file, post the file to a cloud drive and provide a URL
Also on cnc.js my machine doesn't 100% of the time set the x and y machine position. Stil seems to love but it just shows 0.00 in the location fields before I ever zero anything out.
I am about to lose my ever loving mind. I have yet to get a 100% successful cut out of this thing.
It stopped mid cut again in the air. I made the change as requested loaded some stock in and got to cutting and then boom midway. Is there any way to recover the job once this happens?
You asked "Will [setting $gppa=1} change require me to recalibrate all axis? Ans:No, but you likely figured that out already
Suggestion: From console set $p1pof=0.00. This will turn off the PWM LED when you machine is at rest. Mostly cosmetic, but might help with the next question.
When the machine stops midway thru your job, do you get flashing LEDs or other indicator?
From staring at your board wiring diagram, I believe I see an issue.
It appears that your switch connections are single wires, not shielded twisted pairs. But I can see the whole run, so it might be OK but out of view.
Do you have a separate power supply for the spindle?
Limit switch interfaces are very sensitive to noise spikes, even when using NC switches.
A symptom of a limit switch activation would be a flashing LED.
Could be solid or stranded conductors, stranded more flexible.
Make no shield connection at the switch end (machine end).
At tinyG end, connect all shield drain wires together and connect to the " - " Vmot terminal.
Use the red wire for the input terminal (e.g. Xmin, Ymin, etc) and the black wire connects to the gnd terminal on J8 terminal block
Try to route the switch wires separately from the stepper cables and the spindle drives.
I just posted a SpDir flashing support request on another thread and I saw this one. So I am running an air test without the spindle running and this was the first 100% run with the SpDir led is dark.
Just FYI. all limit/home switches are run in shielded cable, and I re-flashed the TinyG to 440.20 and all went well but I also tried to re-flash to 449.19(Edge) version of the fw and the verification failed(just sayin).
If the spindle is the source of the issue and all my limits are shielded what would you recommend if it does turn out to be that if the spindle is running the TinyG faults?
Also I have a separate 48V spindle power supply.
CMCGraph,
Just and update I ran another air test (spindle off) and about 70% though the run and all was well. At that point I turned on the spindle ( it is run manually at this point without any connection to the PWM inputs) and about 5 minutes later the TinyG faulted and the SpDir led started to flash.
Any Thoughts?
SteveRich
This topic is from the original forum (April 2019). Your reply starts a new discussion linked to it.