Tagged: Laser pin
May 4, 2017 at 4:09 pm #32892
Today was released the final 1.1.0 marlin version. I was doing some tests with my prusa i3 clone 3 or 4 weeks ago and it’s working good. It seems that Z issues (resets) related to high speed with 1/32 microstepping is still there by some issues that I can read.
I’ll try it in a new printer I’m building (very interested in UBL new feature) in the next weeks.May 4, 2017 at 4:13 pm #32893
Been messing with it for a few hours. So far so good here.May 4, 2017 at 5:33 pm #32894
I’m having SD card issues. Apparently this might have been fixed in the new 1.1.0 Bugfix…Dang itMay 4, 2017 at 7:17 pm #32895
I was getting reboot looping when trying to run from S3D and then endstop activated issues (no endstops!!) running fro the SD.May 4, 2017 at 7:28 pm #32897
So far so good with the 1.1Bugfix for me. Reading through it it was a float error, that was fixed.May 4, 2017 at 7:39 pm #32900
Dui, ni shuo de duiParticipant
What changed?May 4, 2017 at 8:06 pm #32908
For us we get all the RC8 and 1.1 updates, since we skipped RC8 from all the weird issues that popped up. I think the most stuff for us will be LCD improvements in RC8, after that they were just optimizing code and working on bed leveling stuff. For the end user If your system works, no need to update. They did just add a cool trigger that would be good for printer cooling fans, it turns on when the steppers get activated and off when they are released, good for print cooling fans and board fans. That way when your print stops you can stop the fans and seriously cut down on dust collection, and fan noise.
RC8 implemented a lot of things but slowed down the system, then they optimized it back to where it needed to be with 1.1.
So for us anything above rc7, https://github.com/MarlinFirmware/Marlin/releases
1 user thanked author for this post.May 4, 2017 at 8:08 pm #32909
Oh and it might automatically pull in the ug8lib now, I am not clear on that part.May 4, 2017 at 10:59 pm #32912
Dui, ni shuo de duiParticipant
I saw this thing on the Marlin website:
As you may have seen on my MPCNC thread, I have small issues with the corners of my prints being too big, in comparison with the rest of the print.
It seems like it is possible to activate a function called “linear advance” in Marlin and set up a “K factor” to adjust the extruder output.
Do you know if this is available on the new release? Have you tried this already? I cannot find much explanation on how to enable it (but I did not even try to look in the code yet).
I believe this is less of an issue for smaller nozzles, but it could potentially make a pretty big difference on a big one…May 5, 2017 at 1:35 am #32913
I’m using linear advance in my printer from some time ago and it is a great improvement.May 7, 2017 at 8:20 pm #33134
I gave it a try on my i2 clone printer this weekend. Still had to install ug8lib manually (switched computers since the last time I upgraded Marlin.) I added in the 1.1.x bugfix stuff since it looked pretty important (i.e. looked like stuff that should NOT have been overlooked on an official release that’s been over a year in the works!)
Seems to be working fine on my printer. The LCD menus are definitely improved a bit…but I haven’t upgraded my printer since RC6 – so probably nothing new for those of you on a more recent RC 😉 In fact the few test prints I did are some of the best I’ve gotten off my printer.
Then I made the mistake of trying to figure out UBL 😛 The docs are a mess. I had to read through the code and a bunch of PR’s in github to figure out what was going on. I eventually managed to get it to generate a mesh, and print a G26 mesh test print…which was mostly ok but the back rear corner was really a mess…and I’m too tired to figure out the G29 P>3 stuff for refining the mesh now. So…project for later this week.
I may dig in on the linear advance stuff tonight…been wanting to mess with that for awhile.
Haven’t bothered to try it on my MPCNC yet though…heck I’m still on RC6 there as well because it’s been good enough for what I use my MPCNC for. But with an official release out I’ll probably be upgrading once things settle down. Right now though it sure looks to me like 1.1.0 will have some teething pains. But when you keep adding features to your RC’s that’s not unexpected 😉May 7, 2017 at 8:29 pm #33135
Yeah I’m not a developer so I don’t understand the logic behind the releases. For a few weeks prior everything worked great, I try them out all the time. Then they had a statement saying “feature freeze” just make sure everything works. That all makes sense to me. Then all the sudden right before release they changed a bunch of big things. Shouldn’t that all get added after the fact for the next build?
Either way I have learned to not jump on it right away, the first bug fix seems to be perfect for my uses on all the machines so that is probably what I will set out as the next update so we can all use a more fluid LCD. any other bug fixed will probably be UBL related and you all know I am not a software leveling fan, But I would love to try out the Linear advance stuff.May 8, 2017 at 3:01 am #33139
I would like to test UBL but my I3 clon has a sanguinololu board and don’t fit in the flash with this feature enabled.
I’ll test it in my new p3steel I’m building with arduino / ramps.May 8, 2017 at 3:00 pm #33175
It seems version 1.1.1 will have some cnc specific support out of the box
https://github.com/MarlinFirmware/Marlin/pull/6268May 14, 2017 at 9:14 am #33603
I’m doing a laser run with the new MPCNC Version of Marlin 1.1 and it’s going well enough. The machine does stutter steps on the refresh interval of the LCD but it doesn’t seem to negatively affect the burn so far. I’ll give it another go in a bit and report back.May 14, 2017 at 9:24 am #33604
Dam. Dam. better than RC7?
Well we can turn off the progress bar, not sure if that would help because the screen still refreshes but maybe it will be a faster refresh. And/or go down to 16th stepping I am not really sure what exactly causes this stuttering but this would be less commands sent. I have the laser machine apart to put on the T8 but maybe I should go finish that up and do some tests.May 14, 2017 at 9:53 am #33605
Do you know if there’s a way to just turn off the refresh altogether? Once I start something it’s either go til it’s done or emerg stop if it’s failing so the LCD really is only a file loader and not needed after it starts.May 14, 2017 at 10:06 am #33606
Not sure about that I don’t see anything obvious for it.
You are sure it is when the screen is refreshing and not when the acceleration changes from rapid to burning? Jerk settings and acceleration can affect this, if you set the rapid rate to within 20 of the burn rate it should not stutter (de/accelerate).May 14, 2017 at 10:07 am #33608
I guess I don’t really need the laser on to test this stuff. I’ll see what I come up with.May 14, 2017 at 10:48 am #33609
Well, I see what your saying only when the screen updates does it stutter, much better than it was before though. I reflashed without the lcd…no stutter.
So now I will try some things and see what happens. Strange.May 14, 2017 at 11:00 am #33610
I just finished the one pic and going to reflash RC8 and do another to see if it’s as pronounced. And intervals were exactly every second corresponding with the refresh.
Edit: No stuttering on RC8 at all. I’ll likely run a few more tests but prefer the non-grindy movements. That said, the picture I did with V1.1 turned out pretty nice with no evidence of issue from the weird movements.May 14, 2017 at 11:27 am #33613
What resolution are you using?
I have flashed the board at least 8 times now with different settings. Turns out the resolution is so high, commands are .18mm apart it is flooding the controller and our max speed is about 40-50mm/s I think. I see no difference above there at least and the messing with the buffer seems to confirm, larger buffer fast movements, smaller buffer drastically slower. If the buffer is less that half full slowdown kicks in.
So many variables, but the .18 resolution seems to be the biggest issue because of the shear volume of commands.May 14, 2017 at 11:33 am #33616
The min for our lasers is about .18mm, typical is about .23. Just changing the resolution to .23mm gives about half the file size and a much less pronounced stutter.May 14, 2017 at 11:49 am #33617
Well….Here is a new one. The feedrate doesn’t work in that program.
The first F command needs to be at the end of a g0 or g1 command to work.
So far an f1000, works perfect not stuttering at all with a resolution of .23mm
Still testing more.May 14, 2017 at 11:56 am #33618
F1300 seems to be max. I am going to put this in the img2gcode thread.May 14, 2017 at 12:11 pm #33625May 14, 2017 at 12:20 pm #33627
I’m using 0.1mm resolution and running a M140 2W laser right now. I also bought PEP5 after testing it and the filesize is dramatically smaller. The current burn is going smooth and looks perfect as well under RC8. It’s almost done so I’ll give a better look at the results soon.May 15, 2017 at 8:03 am #33687
I wound up going back to RC6 on my printer. Just too many goofy things in 1.1.0 even with the bugfix-1.1.x branch 🙁
When it worked it worked great and I got some of the best print quality I’ve seen off my machine.
But it was just doing too many goofy things. Even with UBL disabled it was doing stuff like randomly not printing the 1st layer at all. Same gcode, but would get different results when I’d run it 3 times in a row 🙁
And with UBL things went all kinds of crazy, trying to print 20mm above the bed, or slamming the print head into the bed.
After a week of messing with it and applying every fix that came across I finally decided enough was enough and went back to my working reliable RC6 config for now.May 15, 2017 at 8:09 am #33688
Maybe you missed a setting or something. I have this on 4 of my printers and 1 of the MPCNC’s that is printing and it hasn’t had an issue yet. Rc6 is pretty far back some of the movement equations have been fixed since then.May 15, 2017 at 9:22 am #33690
Went over all the settings several times. Like I said it worked great when it worked. I suspect most of the issues are with auto level which I don’t use on my MPCNC and I believe you don’t use on your printers…but on my printer I rely on because the bed shifts far too often ( working on some mechanical upgrades to solve that though.) Most of the issues I had were with the new UBL which I have to say really doesn’t seem to be release ready. But even with the older leveling code things were just not right.
The way it would sometimes skip the first layer and sometimes not from the same gcode really scares me. I can’t think of any setting that should cause that kind of thing to happen.
There does seem to be a lot of improvement on the basic movement stuff. But there were enough rough edges that I’m stepping back from 1.1 for now.
And yeah…RC6 is pretty old…but it’s a case of “if it ain’t broke don’t fix it” 🙂 RC7 had too many issues and it scared me from trying RC8. The printer works and is reliable on RC6 and after a week of fighting with 1.1.x variants and getting fed up it was just a matter of going back to what was working fine before so I can finish printing the parts I need to rebuild my Y axis.
Once that’s done I’ll probably dig back in on 1.1.x…but watching the development the past week or so I’m a bit wary now. There doesn’t seem to be much vetting on new code before it gets pushed into the main branch. The way they kept adding new features to RC’s I’m not surprised that the “stable” release isn’t that stable. They really need to get better about respecting the RC’s and doing feature freezes to get stable releases.
You must be logged in to reply to this topic.