Interesting. I'm not getting the error with .162. Can you double check that you dont have two Controls running. (silly question, but I've had that happen to me) Edit: Yes, you can go back to .158 until Peter gets eyes on this.
No, just one controller open. Thank you for the reply. I have the latest Jave version on my Windows 10 computer.
Peter is in a different time zone than us, so might be a bit before he sees this. Just to be clear, I'm not talking about two windows open. I'm talking about 2 different processes running. Do you mind pasting a screen shot in here? I'm not sure I'll be able to fix any of this, but I'm am curious. If you dont want to, then just wait for Peter.
Just hit a snag. Everything working great, ran a gcode drill program ran good, ran next gcode engraving program it ran good, then when I loaded the next gcode program it said Gcode Opened: in black letters in the Serial Console and nothing came up. I mean no program displayed. So switched to 3d view and saw a small black bock with a frowny face in it. This was in the top left hand corner. I have reset the controller and the cnc, and have tried to load other gcode programs that have worked in the past. same result. Openbuilds control seems to be functioning, it connects, disconnects, homes, macros execute ok, z probe is ok. Any hints ? I have tried to load old programs that have worked before to no avail. Programs were made with aspire, running win 10 with all updates, Openbuilds V160, cnc is a sphinx 1050 with xpro v3
Further to previous post: downloaded v162 and attemted to load it, it wouldnt install. Came up with a "java script error occurred in the main process". So I tried to find my original download of v160, and couldnt find it, so I reinstalled my V 156. It loaded and is running correctly. Homings ok, Macros ok, Z probe ok, gcode programs load and execute ok. At least I can finsh my project. Know where I can get a copy of V160?
I was opening one of four tiling files in the contain software and had no problem. When I went to open the second one it opened really small and with no graph. The same happened for the other two files. I then tried to open a file I worked on the other day which had the graph but it didn't today. Has anyone else has this problem? Should I be concerned? Maybe I do an air run tomorrow to make sure everything is okay.
Busy night So let me group the replies: @Michael Laley: You should be fine with 1.0.162 but I think your PC hasnt done the update right... The bug that was in 1.0.161 was definately fixed in 1.0.162. How sure are you - That you did update to 1.0.162? Maybe your PC didnt remove 1.0.161 and thats what came up after boot? Maybe needs a uninstall then reinstall? - That is the same error? Please screenshot the error (try catch the whole window in the screenshot, theres always clues in the background behind the error too) CONTROL is not Java based. Javascript is a totally seperate language, in this case running under NodeJS/Electron. No need to install Java or anything else, the application is 100% selfcontained @Steveathome - Yours might be an issue with either the GCODE you load, or WebGL drivers. Glad to help you troubleshoot further but gonna need some screenshots I've never seen a Frowney Face? (didnt code one in...), so please catch a screenshot? If the gcode fails to load its usually a problem in the gcode itself (not compable with the GCODE spec, in particular the Grbl based dialect) The issue you see was software based. Try a reboot or Quit and restart CONTROL (System tray icon -> Right Click -> Quit, then start it again) Resetting the controller wont affect the software side. The gcode you loaded (please attach - would need to see the file to tell you what you have in there thats not to spec) messed up the 3D viewer (in some unrecoverable way) so subsequent files wont load until you reboot or restart control (NB restart properly from the Quit menu as detailed above). After a reboot they'll load fine again, but do find which file "breaks" is so we can check it out and make it "break less destructively" in future and lastly @jamin35008 - Please upload your gcode files that act funny Please post your gcode files, and screenshots of anything weird you see: Without files best I can offer is guesswork (; CONTROL is just a "sender" it takes your gcode and sends it to Grbl on the controller. In between it tries to give you a preview and help you with jogging, but at the end of the day its between your GCODE and GRBL - if you used a Grbl compatible post processor to create the GCODE it will be fine
Just installed .162 Works just fine for me under Xubuntu 18 though the old one (think it was 142) did not prompt for an upgrade. This is normal on this machine, I usually have to download and install it manually. Continuous jog for the win! gotoXYZ for the fail. the tool was home, Z high, X+ and Y+, and that macro dived the tool down to Z5 and then went to X0Y0, thus colliding with some stuff on the table! want some video? I would still prefer G53 G0 Z0 G0 X0 Y0 G0 Z[safeheight] or a user selectable Z height, or not move Z at all (assume user has it safely high enough). I have done at least one job where moving to Z5 would have crashed Z since the workpiece was so high Z could move 2mm down to Z0, and cut 1mm deep from there. For that job I very carefully did not use any macros at all (-: oh one other thing, the continuous jog had X+ mapped to the 'rightarrow' key and I got no movement until I remapped it. I might have mapped that at some point when installing my USB keypad that I use for jogging. Now also have a cheap USB gamepad to try out. a utility called qjoypad (linux!) lets you map any button or stick motion to a key or the mouse.
Automated updates are only for the Windows folks at this time, sorry Will look into adding it for mac/linux someday soon Yay! I added the one you wanted me to add a couple months ago: $('#gotozeroMPos').on('click', function(ev) { sendGcode('G53 G0 Z0'); sendGcode('G0 X0 Y0'); sendGcode('G0 Z0'); }); -> SetzeroXYZ -> Dropdown Menu -> Bottom item?
Exactly, so hard to the "right" thing, when you got to keep it easy I edited my reply a little though, just reread, re gotozero
I tried to calibrate number of steps using the fine-tune tool but I don't know exactly the size of ball screws in my CNC router. The ball screws may be 1/2" in diameter. Because the machine is inside a cabinet, it was hard for me to read the ruler. Any suggestions?
Even if you know the size of the ball screw, you'll still need to calibrate it to be accurate. If you cant get into the machine to read a ruler, maybe try laying down a strip of masking tape. Bring a V bit down to put a small hole in the tape. Raise the bit and move 100mm or whatever. Lower the bit down again and make another hole. Raise the bit, remove the tape and measure.
The principle is sound, but I don't recommend using masking tape if you have to take it out to measure it - it stretches. Use any rigid scrap you can tape in place. Alex.
Follow up: I was able to load V156 and run the same gcode programs as before. Tried to upgrade to V162 as it popped up. Failed again with Java error. Going to load V156 again
Found V160, loaded and run. No Java error. Gcode programs run and so do macros. Will see how long it stays up this time. Since I figured out how to screen shot and post.
I get the exact same error. Uninstalled control software, re-installed and same error. I removed.161. v.162 did not fix the problem. I guess I'll go back to v.158 or something because it works. Had no problems until v.161.
I still haven't figured out the screen shot part yet. Print screen doesn't seem to do anything. Windows Snip just gives me text of the Java error instead of a picture.
You don't need to know anything to calibrate the axes, but it does mean you will have to repeat the calibration at least 3 times. Take an unknown system, we don't know the steps per rev, nor the microsteps, nor the leadscrew pitch. We set the steps/mm to 1 and command a move. With steps/mm = 1 the machine will move 1 step for every mm we command, but some unknown number of mm, usually very short compared to the commanded motion. Now if we can see the motor turn we can command a 200 move and it will turn (most motors are 200 steps per rev) 1 turn if microsteps are 1x 0.5 turn if microsteps are 2x 0.25 turn if microsteps are 4x 0.125 turn if microsteps are 8x and so on. This can give us an idea of how far to command to get a useful distance moved that we can measure. If we assume a common setup of 200 steps per rev * 8 microsteps on a 8mm pitch leadscrew then moving 200*8 = 1600 will move the axis 8mm (one revolution). Thus we can quickly get a feel for how far to command for a useful distance for calibration. We want to move at least 100mm for the first calibration, so if commanding G0 X1600 does move 8mm then 100/8*200*8 = 20000 should give us 100mm. Measure that actual movement as well as you can and enter the values in the calibrator (but first move G0 X0, I find GRBL works better if you do the changes after moving back to 0). So we are entering a commanded move of 20000mm and an actual movement of 100. The formula (for completeness we have to do the math, but the calibrator does exactly this) newstep/mm = currentstep/mm * commanddist / actualdist gives us our new steps/mm setting newstep/mm = 1 * 20000 / 100 = 200 steps/mm (this is a sensible value for a 8mm pitch leadscrew at 8x microstepping, the default for Blackboxes) Now repeat this (NOTE that movement will now be very close to actual millimeters so entering a 20000 move is not correct! that is actually 20 meters), using longer moves if you can. Each time you do it the real movement will become more accurate. You can of course start with a higher step/mm, maybe 10 or 100. An OX belt drive system is around 40 step/mm so setting to 100 would move way too far (and fast), but an initial setting of 100 for leadscrews may be quite useful. Start with small moves and see what happens. And yes, I have actually done this process on physical hardware and it does work. (-: note to self, make a video about this (-:
Hi there seems to be something wrong with the new update - I get the error too. I can open the program from the task bar as it is running in the background. I can connect to my machine (C-Beam using the smoothie brd) but none of the controls work - when I try to move or even enter a g-code (ie M114) nothing happens? where can I find the earlier versions? I have done the uninstall / delete the user data folder. Running on Windows 10
Sweet, thanks @Kevin Brewer with good screenshots that does prove a little more than what I was having to work from before. Will dig in a little later, I cannot reproduce the error here (everything works fine here) so not sure whats different about your PCs yet @Michael Laley @Steveathome - Will check and release a new version soon Smoothieware - hard one, think you might be the first and only user here that use it. With the new keyboard and jog controls I started eliminating Smoothie support. The smoothieware userbase is very small around here, and Smoothieware doesnt even have the continuous Jog and stuff we implemented lately. Gonna be hard to keep maintaining support for it without a large userbase. Try CNCJS it still supports Smoothieware? bCNC too, plenty of Smoothieware compatible hosts.
So OpenBuilds is not supporting Smoothie anymore? You guys sent me it when I bought the C-Beam kit, you were out of the other controller at the time and substituted it. I'm guessing the "black box" is the replacement? When will Smoothie be dead in the water, with the OpenBuilds SW?
Well, not officially yes or no, but might be no soon. In the newer versions I implemented the much needed Continuos Jog - which Smoothieware doesnt have! gnea/grbl Smoothieboard can run Grbl though, so your hardware can still work : cprezzi/grbl-LPC Also, the beauty of open source is that you have freedom. OpenBuildsCONTROL focusses on the current needs and tries to keep up with the times. But there are amazing alternatives that do cater for smoothieware still, like CNCJS, bCNC, etc so all good. Same principle as with our PlanetCNC controllers we also sell, they work with TNG, not CONTROL Hard to say, as time goes on and Smoothieware falls behind. But you can always find older builds on OpenBuilds/SW-Machine-Drivers - up to v1.0.153 should work with Smoothie. After that we started to modernise around the better Jogging Grbl offers I may get time to fix Smoothieware Jogging, but due to the small user count (only you that I know of) it might be a couple days
v1.0.164 is in your updates from today - New Feature: Surfacing/Flattening wizard! - UI Change: Keyboard, Jog Widget etc have been consolidated under the new Wizards and Tools button (Cleaned up the top toolbar) - UI Change: New switch below jog buttons to enable/disable continuous jog (Was in the dropdown under the old Keyboard Shortcuts Button on the top menu, but now its easier, just one click away) - Improved behaviour of 3D viewer (Some cases that caused viewer to go full white, should now fail elegantly to a blank grid instead) - Improved GCODE Editor (Added Update Preview and Save buttons) - Fix for the double instance JS error from above - OpenBuildsCAM '.obc'-files can now just be doubleclicked, OpenBuildsCONTROL will catch the event, load CAM and Import the workspace file for you. Some screenshots : The new Toolbar button / Jog Continuous/Incremental Switch: The Surfacing Wizard
Okay dumb question time. I never updated my BlackBox to 1.1g for Blackbox 4x. If I were to flash that to my BB would I lose all of my step settings and such? I'm assuming yes. Edit: Answering my own question. It does not.
Upgrading from .162 to .164 throws and error. It doesnt seem to effect the upgrade though. Just click ok to dismiss and continue with upgrade.
