MiniKossel Upgrade and Duet 0.85 firmware issues
-
[h]MiniKossel Upgrade and Duet 0.85 firmware issues[/h]
Background
In November 2014 I ordered a MiniKossel from Think3DPrint3D for a gift for my daughter. We had previously built a Prusa and were looking for something with better auto calibration (e.g. the mechanical probe) and quality. The MiniKossel was functional, but over time the deployment of the probe was intermittent. Having noticed the IR probe availability in November 2016 we ordered an upgrade which included: IR probe, New print head and E3D V6 hot end, and Duet electronics (v0.85).
We were guided by: https://miscsolutions.wordpress.com/2015/01/04/upgrading-the-mini-kossel-to-duet-electronics-part-1-hardware/
The last print before the upgraded parts were added was a box to hold the Duet electronics (from: https://www.thingiverse.com/thing:441075)). This is mounted on the Z tower (I have a photo but don't know how to attach it to this message).The Duet firmware used was that which shipped with the v0.85 Duet board. I no longer have information about what this version was but it would date to before December 2016. I can see from a backup that the web interface was v1.11.
The usability and quality of the subsequent prints from the upgraded MiniKossel was much improved over the original Arduino Mega and RAMPS board. By the middle of the year we were designing and printing prototype cases for a commercial electronics product.
The quest for improved layer 1 adhesion
The main issue was bed adhesion. Printing with PLA, I had been using glue (Pritt) to get good bed adhesion but wanted a high quality surface finish for the product prototypes from the bed. With some advice I installed a borosilicate glass bed and adjusted print/bed temperatures (and re-calibrated). I made a simple 30x30x1mm shape and got quite a good print (I have a photo but don't know how to attach it). This was OK, but there still seemed to be some variation in adhesion depending on the position on the bed, and as the printing area required for our product was about 80% of the 180 mm diameter base I was still keen to get more accurate first layers (without glue).
I then noticed a firmware upgrade was available and it contained improved levelling and bed mapping.
I applied the upgrade and it is now:
Software Information
Firmware Name: RepRapFirmware for Duet
Firmware Electronics: Duet 0.85
Firmware Version: 1.19.2 (2017-09-01)
Web Interface Version: 1.19I would like to attache the the bed heightmap images from the duet web ui – but I still don't know how to attached figures. I am not sure how to interpret them.
Even after repeated calibrations, I am unable to get a successful print. The best test object print of the 30x30x1mm shape I have done in over a month is a twisted and torn shape (I don't know how to attach an image on this forum). The head seems to scrape the surface of the bed.
I am not sure what to do next. I suspect that going back to the original firmware will take me back to a usable printer.
Any advice appreciated.
-- Steve
-
The images for the above text can be found here: https://imgur.com/a/7WcMK
The upgraded kossel
https://imgur.com/OYdl8bBThe 30x30x1mm calibration object: LHS before firmware upgrade; RHS using Firmware Version: 1.19.2 (2017-09-01)
https://imgur.com/5DIvOo5Three views of the heightmap
-
To attach images, host them on a public server e.g. Dropbox or imgur, copy the link, then use the Image button in the menu bar to paste the link between image tags.
The G29 mesh bed compensation will only work well if the effector tilt doesn't vary significantly with XY position. Otherwise the tilt will alter the relative height of the IR sensor and the nozzle when the sensor triggers. G29 must be used after the usual G32 delta calibration, not before or instead of it.
I suggest you stop using G29 mesh bed compensation for now and use the same calibration settings that you used before. The printer should then behave as it did with the older firmware. On a Mini Kossel you should not need mesh bed compensation anyway. It's larger printers that benefit most from it.
Regarding print adhesion to the bed, probably the best thing you can do is to change the bed surface to PEI or PrintBite. Note that if you use PEI, then it must be painted black on the underside in order to work well with the IR sensor. If you choose PrintBite, choose the opaque version.
There are now other types of Z probe available that use the nozzle itself as the probe, so they are not sensitive to tilt as the IR sensor and other offset-mounted Z probes are.
HTH David
