Duet Wifi (2.02RC4): AnyCubic Kossel Calibration

  • Morning

    I recently installed a Duet Wifi (Bought end of 2017) in my Anycubic Linear Plus.

    It should be mechanically sound. I changed the steppers to 0.9 degrees and made the estep adjustments, and the X,Y and Z probe offsets have been set (I use a BLTouch)

    I cannot do a proper calibration print (a circle). I run the G28 command, and then the G32 command

    It probes the bed (without "skipping" an area)

    When I start the print, it doesn't appear to do any compensation (even if i have included a G29 command)

    The print at the Z tower is squashed

    The print at the X tower is squashed

    The print at the Y tower has very poor bed adhesion

    The end stop screws are the same length. I understand that using the auto calibrate should remove the manual calibration, but I am currently struggling.

    If anyone can offer some advice / pointers please?



  • You have the same model I have, but it was the 'kit' version, it arrived in about 50 bags. I understand the 'new' versions are fully assembled. When I contacted the factory in China in an attempt procure some technical help, their reply was, "we have a new model, it fixes all the probles of that model". I kid you not.

    I had to practically change-out virtually every component due to the poor quality of components provided in the kit (the rod-end bearings were absolute crap). The only parts that my AnyPubic Linear Crapper + handi-wipes has that are from the original kit are the four stepper motors (had to lengthen the cables as they were not long enough), the end-stops, the linear carriages, the 2020 'T-Slot' profile and the belts. Everything else including the so-called Tri-Gorilla controller (caught fire), has been replaced.

    When I was trying to calibrate, the printer changed its mechanical geometry so much that each printing session required a new calibration. I never was able to print anything.

    There are folks who have been successful to cause their AnyPubic's to work. Good luck...

    The parts that I changed?

    • Controller (to DC42's DUET-E controller)
    • diagonal rods (to carbon fiber with mag-balls)
    • effector + hot-end (to DC42's Smart-Effector, very slick design)
    • E3D-V6 hot-end (superior design)
    • power-supply (to 350 W switcher)
    • the plastic corners (to metal 'vertex' corners, to prevent flexing)

    had to purchase a heated bed, as my kit did not provide one,
    and I added a solid-state relay to handle the high current.

    So? Likely not the answer you'd like to hear, and there are many inexpensive deltas that don't work well or not at all, irregardless of the their glowing positive reviews.

    In retrospect, I should have bought DC42's delta. I'd be happily printing right now instead of, after 7 months fiddling around and replacing parts (it was a learning curve), I am just now causing the printer to actually print reasonably accurate objects. I am having some type of configuration issue that I'm working now that I (magically) resolved the heat rise timeouts, etc.

    good luck


  • @3mm said in Duet Wifi (2.02RC4): AnyCubic Kossel Calibration:

    DC42's delta

    I have solved it.

    Patience and play with the Z offset as well as the offsets in the bed.g file

    Just took much longer than i anticipated.

  • Looks like you solved it...

    Still, see if any of this helps:


  • administrators

    @mac_za said in Duet Wifi (2.02RC4): AnyCubic Kossel Calibration:

    The print at the Z tower is squashed
    The print at the X tower is squashed
    The print at the Y tower has very poor bed adhesion

    On a delta printer with an offset Z probe, that's a classic sign of effector tilt. https://duet3d.dozuki.com/Wiki/Calibrating_a_delta_printer

  • @dc42 said in Duet Wifi (2.02RC4): AnyCubic Kossel Calibration:

    @mac_za said in Duet Wifi (2.02RC4): AnyCubic Kossel Calibration:

    The print at the Z tower is squashed
    The print at the X tower is squashed
    The print at the Y tower has very poor bed adhesion

    On a delta printer with an offset Z probe, that's a classic sign of effector tilt. https://duet3d.dozuki.com/Wiki/Calibrating_a_delta_printer

    Yeah, and the "out of the box" Anycubic rods are NOTORIOUS for not being the same length... I do like these printers, (I have two), but the very first upgrade is to change the arms to Blue Eagle/Hayden Huntley magnetic end rods.

    With due respect to Duet... I'd upgrade the rods before I'd upgrade the controller. 🙂

  • And do you still have the crappy screws in the Carrier that hits the endstops? If yes...throw them away..they have slop and it seems that regualry not hitt the switch well and without a proper endstop setting sometimes there will be a good calibration..and sometimes not. and get new endstops...the anycubic gets pressed by a simple piece of metal...too soft...get some cheap switches with a roller at the end of the switch.

  • @mac_za

    I should point out, that my AKL was very likely an early pre-production 'prototype' model, as there were so many mechanically deficit engineering problems, that...? It is unbelievable to me that any engineer with any personal pride would lay claim to having been involved in the engineering of that product!!

    I mentioned that I changed out the rod-end bearings...the mechanical gadgets on the end of the diagonal rods. One of the problems I found was that distance between the center to center of the rod-end bearings affixed to the ends of the diagonal rods were ALL different lengths.

    The miniature rod-end bearings also exhibited there own problems, one being that the ball was swaged by mashing (mallable) brass around the lip holding the ball captive, but then the linearity of the circumference of the ball was also altered, meaning that the dynamic center position changed as the ball the rotated, along with exhibiting rotational 'hard-spots'. I replaced those diagonal rods with the Hayden-mag-balls, which are an excellent solution, however moderately expensive. That fellow works dilligently to ensure that should there be different lengths, that at least they are paired, but in anycase the lengths are just tiny fractions of milimeter differences.

    Re; the end-stops. Those switches exhibit extremely unreliable position triggering. As one fellow here in the forum alluded that the end-stop position trigger point 'moves-around'. Quite true. I machined on my mill an optical end-stop adapter, and that works very well.

    Another problem I resolved involved the plastic vertex corners which flexed, causing the towers to move around a bit. I replaced them with aluminum vertex 'corners', but then a new problem arose, the top idler belt pulley did not line up with the toothed pulley below. That took a bit of head scratching to fix.

    When I replaced the AKL+ effector with the Smart-Effector (very clever engineering IMO), I used the AKL+ plastic belt-clamp/carriage assembled to the linear-bearing block to mount the Smart-Effector PCB carriage plate on which the upper mag-balls are affixed. Its ok, but eventually I'll get off my butt to machine a replacement.

    I am becoming unhappy with the filament extruder drive assembly. It is a very simple, inexpensive, functional assembly. However, I have to unscrew the top Bowden tube push-on fast-connector to manually load the filament into the tube through the connector, else it frequently jams during filament load operations.

    Since the AKL+ I received did not include a heat-bed, I had to buy a third party part. Because of that the bed clamps were not properly engineered for it and I had to not employ the borosilicate glass. In fact the clamps are just not engineered properly for mounting on the 2020 T-Slot profiles. The step on the bottom of the clamps through which the fastener (Allen screw) passes to reach the 'nut' held in the profile, need to be 'notched'. I did so using a rasp-file so that bottom of the step doesn't impinge on the nut inside the profile, preventing the clamp to properly tighten.

    Apparently AC has resolved all of the problems I encountered. It was a fun kit to build, but it certainly was big pile of un-engineered hacked together junk. AnyCubic in China, I wrote them asking for technical help, and as I said, they answered me saying that they had a new model that "fixes all the problems, buy that, you will be very happy!" So I wasn't just bad mouthing them for no reason. Maybe they have their acts together now, what ever may be the case, they screwed me. I bought mine in 2015.

    ps, I am finally printing, got a little problem I am working but its actually working!!

  • I too started with an Anycubic Linear plus. The only original parts on the printer now are the linear rails. I was actually considering putting the leftover parts back together to have a third printer, but then I realized I'd just want to re-upgrade that into a duplicate of my second printer.

    eventually I'll get off my butt to machine a replacement.

    PRINT a replacement. 🙂

    unhappy with the filament extruder drive assembly

    Look at Zesty Nimble. High initial cost, but dead reliable.

    For your bed issues, I find clamps printed in PETG work quite well up to ~100C.

Log in to reply