Duet3D PCB delta printer effector sneak preview
-
The reproducibility you get will depend in part on the microstepping you select and how friction-free your machine is. The resolution of the probe can't be better than the microstep resolution, and will be worse if friction causes the effector not to move on each microstep - which is more likely if you use smaller microsteps.
There are the Z resolutions you get with typical configurations:
1.8deg motors, 20 tooth GT2 pulleys = 12.5um
1.8deg motors, 16 tooth GT2 pulleys = 10um
0.9deg motors, 20 tooth GT2 pulleys = 6.25um
0.9deg motors, 16 tooth GT2 pulleys = 5umIf you increase microstepping to x32 then you might halve the resolution, but only if friction is low enough and you have enough motor current.
-
I have 0.9 deg motors 16 tooth pullys and have 16 microstepping with interpolation set so I am pretty pleased with how it is coming out (Haydn Mag ball arms Delrin ends)
-
do you think you will have black color pcb instead bleu(it s for better looking )
need an advice from expert
i would like to use 550mm length of horizontal OpenBeam (triangle)
and 1200mm for vertical
for you what will be the best size for length arm
i can t arrive to know what size i need to order on haydn s shop -
I use 500 mm Horizontals and 450mm arms which are a bit longer than is actually need so maybe 480 or so would be ok for you.
I can't see the pcb's being in anything other than Blue to be honest and the Beta hotends are also anodised Blue whether the production ones will be the same is to be seen
-
on this calculator i can see for my setup 440mm..
what is you verticale size? -
Vertical makes no difference to arm length but mine is 1000 you may well get away with 440 but they will be very close to the limit What size bed are you aiming for mine is 360mm Dia
-
The production PCBs and heatsinks are both blue. I have both fitted to my delta now
-
Did any beta tester experienced "ringing" due to the custom mounting with the sensor ?
-
This certainly seems the optimum delta configuration at present, nozzle contact sensor, linear rails, Haydn magnets, stiff effector and carriages (PCB/metal). About time for someone to blow it wide open with a radical new design
I do like the Tevo Little Monster frame that's a nice idea, a lot of the rest of the stock machine is so so but the frame is a good idea. -
This certainly seems the optimum delta configuration at present, nozzle contact sensor, linear rails, Haydn magnets, stiff effector and carriages (PCB/metal).
I agree, I think it's state-of-the-art at this time.
About time for someone to blow it wide open with a radical new design
I do like the Tevo Little Monster frame that's a nice idea, a lot of the rest of the stock machine is so so but the frame is a good idea.I've not seen a Tevo Delta up close, but from the photos on their web site I wonder whether the base plate and top plate are sufficiently rigid to match the verticals.
-
do you know where haydn share his Machined Delrin Sockets file?
i would like to print some.
dc42: pleaaaaase take pictures
-
There are pictures at https://duet3d.com/wiki/Smart_effector_and_carriage_adapters_for_delta_printer. These are of the second prototype, but the changes to the production effector are small.
-
looks brilliant.
quick editing error,
"Support for 2-wire thermistor, or 2- PT100 temperature sensor"should be "2-wire PT100"
-
Thanks, I've corrected that.
-
do you know where haydn share his Machined Delrin Sockets file?
i would like to print some.
dc42: pleaaaaase take pictureshttps://www.youmagine.com/designs/magnetic-ball-joint-rod-ends
Any good?
-
I've not seen a Tevo Delta up close, but from the photos on their web site I wonder whether the base plate and top plate are sufficiently rigid to match the verticals.
The cbeams are fairly substantial 8040. If the end plates are stiff and precisely drilled, and the extrusions tapped for long bolts, I can't see it being a problem. Mounting rails is tricky as the mid point in the cbeam inner aspect is between two grooves in the extrusion. Double rail per tower has been proposed.
If I were going to rework the kossel Xl, which would be mad as it's printing brilliantly, I'm tempted to buy the cbeams which are cheap from china and have end plates water jet cut from thick aluminium. Maybe mount the rails on plates spanning the two inner slots.
-
I like the base plate / top plate idea but why not use 40x60 verticals instead and then you have a center groove to mount rails on.
-
It's too late for me to read through all of the comments.
This looks good, I do not fully understand why you did this, I think I remember something about it on your blog. No effector tilt is definitely appealing though, I also would like a camera or two (just to get a closer look or explore AI potential) or at least a light in the bottom that would look good when it's all enclosed in a heated chamber. ?
RGB? Haha I only care about white light there. Really like the no effector tilt though, mine is….significant. That's what I get for buying a cheap kit....
-
how do you see if you are tilt or not?when you move the head on corners with bubble level?
-
Yes you can see tilt on the spirit level, you can also (if you have an offset sensor, presume there is tilt, depending on the trigger height changing relative to the nozzle depending on where on the bed you are probing (this presumes the bed will not affect the sensor like the graphics on some build surfaces do to the IR sensor).
Another method is strapping a laser pointer to your effector firmly, then marking on the wall the dot from the laser, then move to anther coordinate (at the same z level) and see if the dot is still on the same horizontal plane.
Generally with linear rails, and high quality magnetic arm systems, plus stiff, precise effectors, there is little/no tilt measurable.
