Autocalibration for deltas
-
So when I run G32 and auto-calibrate my delta, the printer generates a solution, if I send M665 - I get back the values computed. Are these outputted by the firmware purely for information at that point? Or does the firmware now run using these values until rebooted? I understand I should enter them into config.g to make them permanent.
So for example would it be valid to run a G32 as part of my start g-code? I have to wait for my nozzle to warm up - so it's dead time anyway?
Also I have X-Y probe offsets entered in config.g, but I read in the cartesian printer setup guide on the wiki, that these are not required for a delta. If so it might make sense to put this into the delta section on the wiki. I assume they aren't required for a delta as they aren't used for auto bed levelling so the relative position of probe and nozzle are unimportant, so long as the the probe reads sufficient points on the bed to calculate a solution. What if I chose to do ABL using a second bed.g with 3 points and S0 parameter, then the offsets are required?
-
M665 and M666 with no parameters display the current values of the appropriate parameters. This is generally true of M commands in RRF although sometimes you need to specify a heater or motor number. After running G32 the firmware will use the calculated values from auto calibration until you reboot.
I run G32 just before my first print, after heating the bed and hot end to operating temperature. Some people put G32 in their slicer start gcode. It's best not to run G28 after G32, that way the G32 will compensate for any jitter in your homing switches.
-
Many thanks for the fast response.
-
I'm one of those who puts G32 in my start code. I print these really persnickety parts - like the white spool plate "open mesh" on this fly fishing reel:
This requires nearly perfect calibration - zoom into the photo and look at the consistency of the small openings in the mesh - that is the first layer. -
So is it possible to auto-calibrate, then run ABL (using a second bed.g file and S0 parameter with 3 points - sufficient as I am using a tooling plate). I presume if so, I would not include M561 in my ABL bed.g file, as I would want to retain any bed compensation applied by auto calibration?
I accept this might not be necessary with all mechanical issues ironed out but I'd like to try it anyway.
-
Currently, RRF doesn't support bed compensation on deltas. I intend to add support for it when I implement grid based auto bed compensation.
-
Currently, RRF doesn't support bed compensation on deltas. I intend to add support for it when I implement grid based auto bed compensation.
-
May I ask who's filament you are using? If I had to guess I'd suspect it's Protoparadigm Quartz and Emerald….
I'm one of those who puts G32 in my start code. I print these really persnickety parts - like the white spool plate "open mesh" on this fly fishing reel:
http://mhackney.zenfolio.com/img/s6/v149/p214006467-3.jpg
This requires nearly perfect calibration - zoom into the photo and look at the consistency of the small openings in the mesh - that is the first layer. -
Nope, el cheapo eBay stuff. PLA
-
Thanks David, I'll go back to checking the mechanicals.
mhackney - loving the 3d print, would you consider sharing the model, I'd like to have a try, its something to aim for.
-
Wow, it seems like it's getting harder to find "bad" PLA these days. What brand does it sell under so I can take a look?
By the way I used your guide to get the FSR's setup on my Duet Wifi–it was a tremendous help! I had all my configuration ready to go ahead of time so the minute my Duet WiFi showed up all I had to do was swap in the controller, stick my SD card in, and hit the ground running.
Do you have any issues with the hot-end nozzle marking up your PEI during auto-calibration? 20mil PEI isn't exactly cheap--nor is the 250mm round piece of boro it's permanently adhered to, so I've been very careful not to let the hot nozzle come in contact with the PEI and as a precaution I actually flip my glass upside down so the PEI is facing down against the aluminum heat spreader when doing a G32. It's sort of a pain so I was curious what you're method is. Or do you find letting the hot nozzle "bump" the surface doesn't leave any marks...
-
@DjDemonD the files and instructions are available for non-commercial use on my website: http://www.eclecticangler.com/3d-printed-reel-plans/
@W3DRK, I've been printing all brands of PLA for 5+ years and have never had a problem with any of it - but it took a lot of work to get to that point. This material is called Sea Green and available from eBay vendor faveprices. I typically buy my PLA based on color and not brand! I have customers who want special colors to match a fly rod, or a college colors, or whatever.
Thanks, glad that helped. The Duet WiFi with David's firmware and Christian's Duet Web Control is the perfect 3D printer setup in my opinion - especially for delta printers.
-
After running G32 the firmware will use the calculated values from auto calibration until you reboot.
is there a way currently to save those calibrations so on a reboot it will hold those cal values? such as saving them to the eprom / Bed.g file?
in my case i find it not necessary to do a auto level every time i want to print something after I turn the system off. ?
-
@Russ:
After running G32 the firmware will use the calculated values from auto calibration until you reboot.
is there a way currently to save those calibrations so on a reboot it will hold those cal values? such as saving them to the eprom / Bed.g file?
in my case i find it not necessary to do a auto level every time i want to print something after I turn the system off. ?
Currently you have to send M665 and M666 with no parameters to find the current values after you ran G32, then edit the M665 and M666 commands in config.g to use those values.
-
It's very well calibrated now deviation 0.050, and the bed level after calibration is within 0.05mm from one tower to the next.
What I had to do was change all the probing points so that they positioned the sensor not the nozzle in the optimum positions to make the calculation, my sensor offset is -13,34. I presumed if the firmware knew the sensor offset it would do this automatically but of course it can't as it would potentially move the probe out of the bed.
-
It's very well calibrated now deviation 0.050, and the bed level after calibration is within 0.05mm from one tower to the next.
What I had to do was change all the probing points so that they positioned the sensor not the nozzle in the optimum positions to make the calculation, my sensor offset is -13,34. I presumed if the firmware knew the sensor offset it would do this automatically but of course it can't as it would potentially move the probe out of the bed.
My bed file generator at escher3d.com does allow for the probe offset, and I believe the bed file generator in the RRF configuration tool does as well.
I have considered adding a mechanism in firmware to generate the probe points automatically, but then of course you would not be able to add the H corrections that most delta printers seem to need.
-
Yeah I started there but when I asked for 16 points it wouldn't generate the points, "suggest probing points" was greyed out. In the end since my maths has utterly left me, I just moved the head around manually and entered the coordinates of some suitable points.
Also nowhere on the page does it ask for Probe offset. It is a very useful tool, I might run some of the numbers on my Kossel Mini through it.
