Marlin CoreXY functionality?
Posted by crispy1
|
Marlin CoreXY functionality? October 28, 2012 08:03PM |
Registered: 12 years ago Posts: 374 |
Has anyone played with the corexy functionality in Marlin?
I'm building an h-bot prototype. I uncommented the corexy line in the config, flashed to board with no problems. But when I home X or Y, only one motor turns, not both as it should.
I have the motors plugged into the X and Y outputs on the board. I'm assuming that's where the motors should be connected but there's basically no documentation on how to set up or use corexy other than uncommenting the line in the config file.
I'm building an h-bot prototype. I uncommented the corexy line in the config, flashed to board with no problems. But when I home X or Y, only one motor turns, not both as it should.
I have the motors plugged into the X and Y outputs on the board. I'm assuming that's where the motors should be connected but there's basically no documentation on how to set up or use corexy other than uncommenting the line in the config file.
|
Re: Marlin CoreXY functionality? October 28, 2012 08:34PM |
Registered: 12 years ago Posts: 939 |
There is a bug in it.
If you have it set to disable motors after they are idle for a while, it only enables one of the two motors when you do either a X or Y move.
You can manually jog both motors before doing anything, or I assume disable the idle code.
Edited 1 time(s). Last edit at 10/28/2012 08:35PM by Polygonhell.
If you have it set to disable motors after they are idle for a while, it only enables one of the two motors when you do either a X or Y move.
You can manually jog both motors before doing anything, or I assume disable the idle code.
Edited 1 time(s). Last edit at 10/28/2012 08:35PM by Polygonhell.
|
Re: Marlin CoreXY functionality? October 28, 2012 09:48PM |
Registered: 12 years ago Posts: 374 |
|
Re: Marlin CoreXY functionality? August 17, 2013 02:05PM |
|
Re: Marlin CoreXY functionality? August 22, 2013 03:17PM |
Registered: 12 years ago Posts: 544 |
you know now that ive been using the corexy, at the begining i was having the issue of only one motor turning on but i havent had it happen since i updated to the latest marlin firmware, btw my corexy printer is working great, the speed is limited though because my hot end/ extruder mount is a little flimsy, and also its a direct extruder not bowden, i did try the bowden, it just has too much oozing.
|
Re: Marlin CoreXY functionality? August 22, 2013 04:17PM |
Registered: 11 years ago Posts: 162 |
Good to hear aduy. Myself and a buddy of mine over at terawattindustries submitted the marlin patch. I've been using it on my own h-bot and I haven't run into any bugs yet, but if you do find any, let me know.
Vinicius, I do not understand what you mean by "have not found where the line of the idle-motor". Could you clarify your question?
Vinicius, I do not understand what you mean by "have not found where the line of the idle-motor". Could you clarify your question?
|
Re: Marlin CoreXY functionality? August 24, 2013 07:12PM |
Registered: 11 years ago Posts: 490 |
So is it safe to assume that the output quality is the same with CoreXY now?
From what I have read, there were two bugs: one, in which subsequent layers were shifted to one side and another that asymmetric objects didn't come out right (one side was mirrored or something along those lines).
I, too am thinking of building a (second) printer and am looking at the various designs with static motors. The Ultimaker style, while being less than ideal for experiments due to the constraints of the carriage was the only viable option so far, but if the CoreXY functionality is finally implemented correctly, this might be another interesting option because the carriage allows for much more room to experiment with. Also, you can use non-round linear rails (thinking of square aluminium extrusions from the hardware store) while with the Ultimaker-style, you have to use round rails.
From what I have read, there were two bugs: one, in which subsequent layers were shifted to one side and another that asymmetric objects didn't come out right (one side was mirrored or something along those lines).
I, too am thinking of building a (second) printer and am looking at the various designs with static motors. The Ultimaker style, while being less than ideal for experiments due to the constraints of the carriage was the only viable option so far, but if the CoreXY functionality is finally implemented correctly, this might be another interesting option because the carriage allows for much more room to experiment with. Also, you can use non-round linear rails (thinking of square aluminium extrusions from the hardware store) while with the Ultimaker-style, you have to use round rails.
|
Re: Marlin CoreXY functionality? August 24, 2013 07:34PM |
Registered: 11 years ago Posts: 162 |
All of the old corexy code was gutted and replaced with a higher-level implementation in the step planner. There's no loss of accuracy that stems strictly from the corexy code.
octopus print
I printed that using corexy Marlin on my UConduit hbot with bowden extrusion.
octopus print
I printed that using corexy Marlin on my UConduit hbot with bowden extrusion.
|
Re: Marlin CoreXY functionality? August 24, 2013 07:47PM |
Registered: 11 years ago Posts: 490 |
|
Re: Marlin CoreXY functionality? August 24, 2013 07:55PM |
Registered: 11 years ago Posts: 162 |
Yes -- here's the summary of the pull request:
https://github.com/ErikZalm/Marlin/pull/388
and here's the commit:
https://github.com/ErikZalm/Marlin/commit/ff6fa09ecf188272761f8d4a08df9749a70c1d7e
Versions of Marlin from up to 3 months ago should already have the corexy fixes.
https://github.com/ErikZalm/Marlin/pull/388
and here's the commit:
https://github.com/ErikZalm/Marlin/commit/ff6fa09ecf188272761f8d4a08df9749a70c1d7e
Versions of Marlin from up to 3 months ago should already have the corexy fixes.
|
Re: Marlin CoreXY functionality? August 25, 2013 12:14PM |
Registered: 11 years ago Posts: 490 |
|
Re: Marlin CoreXY functionality? August 26, 2013 01:59AM |
Registered: 12 years ago Posts: 544 |
I found an issue, when you start each print one of the motors does not turn on for 1mm, then if you pause the print, home it, it will be truly homed but it will start printing slightly off. It happens when you are heating up the bed and nozzle after it has homed. I just printed some polycarbonate ill show you the timelapse in a second here
|
Re: Marlin CoreXY functionality? August 26, 2013 02:00AM |
Registered: 12 years ago Posts: 544 |
[www.youtube.com]
Actually I figured out what went wrong, its not the firmware, well it is sort of, I set the steps per mm as "1600/15" because it was easier to modify, however when you pull up the eeprom settings, it says 106.00 instead of 106.67 this accounts for the movement, because I went and changed a parameter in the eeprom during the print then it corrected itself. so the lesson here is dont use fractions in the firmware i guess.
One other thing, the timing is so off for the estimate on my machine, even with maximum acceleration and jerk the estimate is around half the actual time.
Edited 1 time(s). Last edit at 08/26/2013 02:34AM by aduy.
Actually I figured out what went wrong, its not the firmware, well it is sort of, I set the steps per mm as "1600/15" because it was easier to modify, however when you pull up the eeprom settings, it says 106.00 instead of 106.67 this accounts for the movement, because I went and changed a parameter in the eeprom during the print then it corrected itself. so the lesson here is dont use fractions in the firmware i guess.
One other thing, the timing is so off for the estimate on my machine, even with maximum acceleration and jerk the estimate is around half the actual time.
Edited 1 time(s). Last edit at 08/26/2013 02:34AM by aduy.
|
Re: Marlin CoreXY functionality? August 26, 2013 07:57AM |
Registered: 11 years ago Posts: 490 |
Nice, maybe we will see more H-Bot RepRaps in the future. The Mendel design may have been a great solution for the limited possibilities back then, but I think we should boldly take steps forward in machine design.
by the way, how is polycarbonate? I have heard great things like that it is warp-free, high temperature resistant and extremely strong, so it might be suitable for Tantillus-like printed frames that also serve as heat retention chambers.
by the way, how is polycarbonate? I have heard great things like that it is warp-free, high temperature resistant and extremely strong, so it might be suitable for Tantillus-like printed frames that also serve as heat retention chambers.
|
Re: Marlin CoreXY functionality? August 26, 2013 01:47PM |
Registered: 12 years ago Posts: 544 |
it it is hard to make it stick to plain kapton but the abs juice worked awesome, basically anything can warp though if it is not stuck down to the bed. I am only running at 265c currently, im waiting on getting my prototype hot end back that can do up to 600c before I start ramping it up. I was able to print a nice 1 perimeter cylinder at 150mm/s with 100mm/s jerk and 10000 acceleration, the layer height was 0.1mm and i was using a 0.6mm nozzle. I think this is my favorite material to print, its oddly flexible, yet extremely tough. and the optical quality is amazing, you can see through the perimeters vertically, and if you have nice thick layer heights a lot of light can pass through, I think with a higher temperature it will be even better.
|
Re: Marlin CoreXY functionality? August 28, 2013 05:11PM |
Registered: 12 years ago Posts: 544 |
|
Re: Marlin CoreXY functionality? August 28, 2013 05:35PM |
Registered: 11 years ago Posts: 162 |
|
Re: Marlin CoreXY functionality? August 28, 2013 07:18PM |
Registered: 12 years ago Posts: 544 |
|
Re: Marlin CoreXY functionality? August 28, 2013 09:19PM |
Registered: 11 years ago Posts: 162 |
|
Re: Marlin CoreXY functionality? August 28, 2013 11:05PM |
Registered: 12 years ago Posts: 544 |
how can you change it to a float, it kind of makes a huge difference. oh wait now i see, in the configuration.cpp it is a float already, but when you pull it up through repetier host in the eeprom editor it shows 106.00 still even though it is set as 106.667 in the configuration.h file. Ok so I cannot get it to save it as 106.67 using the arduino environment, however i can edit and save it to the eeprom using repetier host. something is going on here and I dont like it.
Edited 2 time(s). Last edit at 08/28/2013 11:16PM by aduy.
Edited 2 time(s). Last edit at 08/28/2013 11:16PM by aduy.
|
Re: Marlin CoreXY functionality? August 28, 2013 11:54PM |
Registered: 11 years ago Posts: 162 |
|
Re: Marlin CoreXY functionality? August 28, 2013 11:57PM |
Registered: 12 years ago Posts: 544 |
|
Re: Marlin CoreXY functionality? September 07, 2013 10:52AM |
Registered: 11 years ago Posts: 1 |
|
Re: Marlin CoreXY functionality? September 09, 2013 12:59AM |
Registered: 12 years ago Posts: 544 |
What kinf of problems? also I figured it out, you need to add a .0 to every number in the line to make them all floating point, the z steps was just 2560 so after i replaced it with 2560.0 it came up correctly with out any integer division error. this in my opinion is a huge deal it should at least be commented in the firmware that they all need the .0 behind. anyways i got a shit ton of carbon activated filter from amazon for cheap and some big ass fans to run them on as filters, time to print some poly carb at maximum speed and temperature.
|
Re: Marlin CoreXY functionality? October 07, 2013 06:50PM |
Registered: 12 years ago Posts: 253 |
Just got my machine together today, and second print was great.
As good as a stratasys. A short, half dollar sized washer.
Third part was a tall tower with a base flange, but it shows a diagonal issue.
I've got all floating point.
The X and Y steps are the same number, at 43.6.
I've changed Z to 400.0 and E is 471.3
An M503 reports all the same numbers that I want.
The diagonal print goes off in a 45 degree versus the X/Y axis.
Am using the Marlin from an August download.
I've used a dial indicator to verify the X and Y movements are within .01mm from any desired move.
As good as a stratasys. A short, half dollar sized washer.
Third part was a tall tower with a base flange, but it shows a diagonal issue.
I've got all floating point.
The X and Y steps are the same number, at 43.6.
I've changed Z to 400.0 and E is 471.3
An M503 reports all the same numbers that I want.
The diagonal print goes off in a 45 degree versus the X/Y axis.
Am using the Marlin from an August download.
I've used a dial indicator to verify the X and Y movements are within .01mm from any desired move.
|
Re: Marlin CoreXY functionality? October 07, 2013 08:38PM |
Registered: 11 years ago Posts: 162 |
|
Re: Marlin CoreXY functionality? October 08, 2013 12:20AM |
Registered: 12 years ago Posts: 544 |
|
Re: Marlin CoreXY functionality? October 08, 2013 08:30AM |
Registered: 12 years ago Posts: 253 |
I'll look for that. I was thinking that the top was straight.
That section and the top few layers of the base were run at 200% in 3DCreator, meaning 200mm/s.
It was moving pretty well, and maybe the quickness was helping it not loose steps??
Maybe I'll turn up the Rambo amps to a bit higher also.
Maybe check the connectors also.
On a large box print, the 45 degree infill was using only the lefthand motor and it had issues and made some sounds.
I'll continue looking before I resort back to standard cartesian belts.
Thanks for the tips.
That section and the top few layers of the base were run at 200% in 3DCreator, meaning 200mm/s.
It was moving pretty well, and maybe the quickness was helping it not loose steps??
Maybe I'll turn up the Rambo amps to a bit higher also.
Maybe check the connectors also.
On a large box print, the 45 degree infill was using only the lefthand motor and it had issues and made some sounds.
I'll continue looking before I resort back to standard cartesian belts.
Thanks for the tips.
|
Re: Marlin CoreXY functionality? October 08, 2013 03:47PM |
Registered: 12 years ago Posts: 544 |
You know, on my printer im actually running two motors for each axis, in the corexy configuration, I have the motors both wired exactly the same, I wonder if I switched the polarity of one of the motors, so it energized opposite coils in each motor if that would make it stall less, even with two 4.5kg motors on each axis I can only get it to run reliably up to 120mm/s, although i am running 15 tooth gt2 pulleys, perhaps thats too small.
|
Re: Marlin CoreXY functionality? October 08, 2013 04:40PM |
Registered: 11 years ago Posts: 162 |
You might be right about the pulleys being too small. I use 40T pulleys and can reliably go much faster. At 15T, you're only moving 30mm/revolution. So that means the motor could potentially be turning at 4 rev/s when you go 120mm/s. When translated to step frequencies, that's 4*200*16 = 12.8kHz, which is pretty high. For a test, you could try going down to x4 microstepping to see if it helps.
Sorry, only registered users may post in this forum.