Hi, a little update. Looks like a complete, calibrated and cutting Maslow just appeared in a cellar in berlin…
We applied so many of your suggestions that we are not sure anymore where the problem was. Anyway, at the actual status we have ground control and firmware version 0.85. Short usb cable connecting to the arduino, no extension for power cable.
Thanks again to everybody for helping and hopefully next post will appear in the ‘projects’ section!
Safety first (find the error)
No shoes. What do I win?! 
Ride in an ambulance! Stiches!
Medic shift today 
Our good days are other people’s bad days
I am running firmware .86 and ground control .86
I did the complete 9 step calibration process under “actions, calibrate machine dimensions” and got my errors within the .5mm tolerance. I closed ground control and the next time open ground control it immediately gives me the cannot calculate with 0 chain length error and i close that and it pops up again. I still don’t know how to manually calibrate the chain length and to have to unhook the sled and re-calibrate the chain length every time i turn the machine on is a huge waste of time.
Any way to fix this problem?
Recalibrate the chain lengths the twirl it out way, and mark the links at the top of the sprockets. Next time manually position the chains and go to advanced/manual chain calibrate. It just gives a message, no movement. You’re done.
I appreciate the slightly faster solution but do you really want to do that every single time you start up ground control?
Just to confirm, when you start GC, you see versions 0.86 reported, as below, and you have successfully completed the chain calibration one time since applying these updates?
If you are still seeing the calibration warning after those steps, we have missed something when trying to solve it. If so, an issue should be opened to track it, it’s important to squash this one.
I am having this issue as well. I am on master for both firmware and GC.
I only see this on startup, not whenever I change settings, which is an improvement over before.
Every time I startup, I get the error message, but many times the machine is subsequently able to determine it’s position. But sometimes, it just completely loses its s#!t and the red target doesn’t even appear on the screen.
The machine I run GC on is not fast. Is it possible that the main loop runs before the initial settings are completely saved?
I can try and capture a log of it happening, maybe that will help.
THANK YOU! I didn’t know that this existed. I have been running automatic chain calibrate with no chains so I can walk away for the five minutes and then putting the chains back on using my prior marks.
Darn, I thought we had this fixed with last week’s update. I guess not.
When it when it happens what lengths does it give you for the chain lengths in the error message?
I will try and reproduce it now that I know it won’t waste time resetting everything.
I think it reports a length of 0,0 but I will try and test it thoroughly tonight.
Here I made it happen and saved the relevant log portions.
In this case, I got the error, but then it was able to recover and after I closed the popup box it knew its position.
I will note, and I don’t know if this is relevant, I was having a hard time causing the error. I remembered that this happened a lot when I didn’t have a gcode file loaded, so I cleared out the openfile line in the ini. This seems to cause an issue where the other settings are also not loaded because it keeps loading up with MM as the units even though I keep selecting inches on close.
NOTE we don’t allow uploading of TXT files, so I saved it as gcode. I think we should allow TXT as well.
failed at start then recovered.gcode (1.4 KB)
EDITED - I deleted repeated lines and replaced them with elipsises to make it easier to read.
I am showing the .86 versions of GC and firmware same as your picture and
those are the first and only versions I’ve ever installed since I just
started assembling my Maslow a couple nights ago.
Also yes I successfully completed the chain calibration after installing
version .86 of GC and .86 firmware
I’ll try to open an issue tonight
Alright, never been so happy to create such an irritating error.
Everytime I open GC if the openfile line is blank, I get the error that I posted before, but GC eventually recovers.
However, and this is a head scratcher for me. If i restart my computer, then when I load GC again, I get the same error but it never recovers. Attached is the log of the event. I will reset and see if this is reproducable again.
Yup restart and I get the same problem again. Log file looks identical too. The bad error only seems to happen after restarts. Once I reset everything I am good until the next restart.
Other relevant settings:
Triangular or Quadralater doesn’t seem to matter.
I have a zaxis installed.
Happy to test out anything if anyone has any ideas.
On the theory that you can never have too much data. I tried having an open file and restarting my machine. I still get the total loss of position error, so maybe the two are not related.
The log is attached as it looks slightly different.
So again, anytime I restart my computer, when I load up Ground control for the first time, I completely loose my position.
failed at start never recovered with openfile.gcode (1.6 KB)
OK I will stop talking so much now.
the lengths it gives are just 0.00 for the chain and I didn’t write down the exact text but the error mentions chain length.
It sounds like a repeatable issue that a lot of people are having. I can’t make my machine do it, so if it’s OK with you guys I’m going to have to ask you to test my solution when I come up with one.
Thank you for attaching the log file. I was just about to ask for it, but you already posted it!
Of course.
I may have some time to look into the code tomorrow.
@bar, does it matter how one exits GC? Is it possible that the valid settings aren’t getting saved on exit, and the old invalid ones are being sent on startup?
GC issue #400 refers to Kivy exit using the ‘Esc’ key, which currently bypasses the gracious exit.
I’m pretty in the dark on this one so far, but the issue seems like it might be on the Arduino side to me. When the machine reports that it can’t find it’s position with chain lengths of 0,0 that means the lengths weren’t saved last time the Arduino powered down or couldn’t be loaded from the EEPROM.
The values are written to the EEPROM whenever no command is sent to the machine for two seconds which is the same time you hear the motors power off. For example, normal behavior would be Get Command -> Move to new location -> Hold Position for two seconds -> Motors power off (machine goes quiet) -> Position is written to EEPROM
Under what circumstances that wouldn’t happen I’m not sure.
