I experience the error 2 every time I tried to enter something into FluidTerm.
This recent time it was because I was trying to enter the code that Dland had
sent regarding resetting it. I didn¢t get it on the first try, but then I got
it to reset the ESP32. I couldn¢t figure out how to use the tee filename
command, but I¢ll try again tomorrow.
what commands are you trying to type in via fluidterm? most of them should have
$ in front of them and if you enter them without the $ in front you will get
error:2
I was trying to type in: tee filename to try to get the whole report since it didn’t seem like it was all there. I didn’t know about the $. I’ll try that tomorrow when I take another crack at it.
Before I go dismantling the whole thing again, do you have any guesses as to which motor is shorting out? I really don’t have time tomorrow to take the whole thing apart to grind down the excess leads, but I could probably get 1 or 2 done.
Before I go dismantling the whole thing again, do you have any guesses as to which motor is shorting out? I really don¢t have time tomorrow to take the whole thing apart to grind down the excess leads, but I could probably get 1 or 2 done.
I spent a good chunk of today grinding down all of the leads on the encoder chips. I finally got it put back together and tried to run a test. Absolutely no change.
It managed to save a calibration. Then I tried to move the Z axis and it worked. I took it down, put in a bit, zeroed the Z axis and appeared to hold the bottom out z point as well as the Z home. Fingers crossed, I’m going to try a test cut tomorrow.
If anyone can offer any explanation as to why the calibration failed at noon but then worked when I tried again, I would appreciate it.
Excellent! Ish! That pretty much confirms that there is something going on in the school WiFi environment which is messing things up, but it doesn’t give us a solution just yet. I think that a cheap wifi repeater is a good place to start looking.
The math problem that we’re trying to solve is too large and complex to actually “solve” instead we are basically doing a lot of guessing and checking which makes it non-deterministic. It’s quite sensitive to the initial conditions of where the anchor points are so it can basically learn from a failed first attempt and do better next time.
The project that I’m working on right now is to get a better initial guess to basically help the system make a smarter first guess for where to start searching which makes things more reliable, but I’m still working on it.
I used my personal computer that worked yesterday. When I connected to the Maslow, its status was ready to cut. It remembered where home was from my last attempt. It moved over to it perfectly. I moved up the board about 100 mm and redefined the home point so that the cuts were not overlapping.
Then I started the cut. I noticed that it was rocking on the top edge of it instead of sliding around the surface. But then once I started filming it stopped. I looked and it had disconnected.
When I reconnected it said its status was extended. I started the cmd ping and clicked reconnect. It disconnected again. I remembered the only thing I changed on my computer was that Mcafee antivirus had asked to install a Chrome Extension. I disabled that and then moved on?
I then started the cmd ping, as well as fluid terminal. I extended and retracted the belts, applied tension. That was when I realized the bit was a bit below the surface. So I raised the bit, reset z home, released and applied tension because the bottom belts were pretty loose at that point. The status said it was ready to cut after that. So I uploaded the gcode, set a new home and hit play.
Immediately the cmd prompt registered a request time out. Seconds later, the disconnect message popped up.
The new control card was delivered to me. Should I install it and give it a shot? Or are there steps I should take here to try and fix this one? I read that some people are using Firefox, should I try a different browser with no extensions whatsoever?
Try the new card. If you still have the same problems it eliminates the existing card as the problem. You will have to load the software on it to the latest release. I suggest you back up your maslow.yaml file on the existing board, then move the power cord to the new board (before installing it) and use the USB connection to update the software to V1.17. It makes it much easier to do on the desk, especially if you have to hold down the reset button while starting. To load the software you only need the power and the USB connection. Run a full_install.bat. After installation load your maslow.yaml file onto the new control board.
I then started the cmd ping, as well as fluid terminal. I extended and
retracted the belts, applied tension. That was when I realized the bit was a
bit below the surface. So I raised the bit, reset z home,
this makes me wonder of z stop got messed up, you would need to have belt slack
or remove the bit to run the router all the way down and hit set z stop (in the
settings menu)
if it loses it’s x/y position, there’s a chance it lost it’s Z position as well.
the Z position affects how much belt is needed, so if it’s idea of the Z
position is off, that could cause slack belts.
The new control card was delivered to me. Should I install it and give it a
shot? Or are there steps I should take here to try and fix this one? I read
that some people are using Firefox, should I try a different browser with no
extensions whatsoever?
the key thing you need is to make sure the client computer isn’t going to
interfere with the connection for power savings or security reasons. That’s
partly in the OS and partly in browser extensions. It should not matter which
browser you use as long as it doesn’t interfere.
I would try setting z stop and then doing the reatract/extend and see if things
work then before replacing the controller.
Good afternoon.
I installed the 1.17 firmware on the new card. Updated the index, yaml files. I’m working on my personal laptop so the school Wifi is not interfering. But it will not stay connected. I can’t even extend all the belts to run a calibration. For a while it was disconnecting with every belt I tried to extend a belt. When I finally got the belts to retract, extend, and apply tension, it give the message failed to compute x/y lengths. So I still can’t run a calibration. Let alone cut a job.
Since I have some time tonight, I thought I would give it another go and record it.
Here is a link to the new video.
Here are the serial messages I announce in the video as well as the Fluidterm text:
If there is a resource on how to use or read the fluidterm program I would appreciate it.
At this point my only thought is it’s a bad install and to reflash and reinstall all the drivers and firmware. Any other troubleshooting techniques or recommendations would be appreciated.
Thank you,
I feel like we’ve tried all the things that come to mind as obvious fixes, I’m going to work on a new firmware release next week with some more advanced diagnostic for connection issues because we don’t have a great way to diagnose them right now. I want to add active connection monitoring in both directions (computer to machine and machine to computer) and also some other things like scanning to see if there are other wifi networks on the same channel in the area.
That doesn’t help you today, but I am working on the next firmware version to hopefully make it easier track this down and I should have something to try out next week.
one thing you can do to get some more visibility is to go to the fluidnc tab,
under the left buggle menu there is a logging option currently set to info, if
you change that to debug there will be a little more info in the log (including
about what url is being connected to)
note that under the config (right bubble) there is also a debug option, but that
is a very different thing.
I need to get a usb cable up here at tome to be able to anwer your question
about fluidterm.
Thank you for looking over the information. I appreciate you taking my issues into account with the update. I look forward to trying it out.
Is there a tutorial on how to use the erase tool that is packaged in fluidnc? I think I need to wipe the new card clean and start over with a fresh install. It is currently worse with the new card than before, so I can’t help but think I botched the install and update. That is the simplest explanation. I read the readme file in the package, I don’t think I used it right.
Thank you for letting me know the video is private, I have set it to unlisted so you should be able to see it now. I was getting a bit frustrated when I made the video, so I apologize in advance.
I set my belt length to 1780 mm this was based on a measurement from my anchor to the center of the cut board. I will use the measurement you suggested tomorrow when I try again. When I attempted it the final time (one more after the video) it had forgotten my settings again. So I will check and update them tomorrow.
I will look at the debug mode of fluidnc. Hopefully that will give us some more information.
I am pretty sure the full-install.bat replaces everything except any gcode files you have downloaded, which are in a separate directory.
Use the version in 1.17 that should take you up to the latest release.
Looking at the full-install.bat file it sequentially calls:
erase.bat
install-fs.bat
install-wifi_s3.bat
then starts win64\fluidterm.exe
unless something fails, then it stops.
Open a command prompt window, change to where you have unpacked fluid-maslow4-1.17-win64 and run the full-install.bat from there.
I didn’t have much time today, but I tested again at school with my personal laptop.
I believe I was able to find the diagnostic setting in fluidNC. I also reinstalled the 1.17 firmware as well as the Index file, as well as updated the spoil board and material thickness, as well as use a new measurement for the extension distance.
I tried to calibrate the Maslow after school today. It kept dropping signal when I would extend the belts again. Then it either wouldn’t apply tension, wouldn’t release tension, wouldn’t retract, or wouldn’t extend. I had to reset it 2x during my attempt to get to a calibration.
Then I took the Maslow home and tried it in my dining room to see if I could reproduce the errors. It dropped the signal 2x while I was extending the belts the first time. (serial reports for each are attached) I had to reset it to get it to retract the belts after they were all extended. Then I went ahead and extended them again. And it worked. No drop. Then I was able to retract them with no problem. The serial report for after that as well as the fluidNC report are attached below.
You are connecting via the Maslow network, not a LAN. From what’s happening at home it looks like it’s not a problem with your Maslow but a WiFi problem, At school it could be a configuration problem on your PC or WiFi interference. You could try loading a WiFi analyser on your phone to help work out what’s happening.
