Joint of the diagonal rod kept falling one after another. It really pisses me off and I’m paying high price for innocent, naive mistake I did to use a cheap plastic glue for the diagonal rod assembly. The metal epoxy didn’t last long either approximately three weeks.
When there is slight free play on effector related parts – six diagonal rod, effector assembly, hotend – setting up the even level is impossible and it creates real printing problem which is building the first and second layer imperfect or cause excessive stress to the joints.
In my case upper left side was always 0.1 to 0.2 mm higher than the bed center. So whenever I started printing medium to large size model, the nozzle literally grinded the aluminium plate. To remedy it I set up the Z height on taped bed surface but the loose diagonal joints made big scratch marks on the aluminium bed.
The major weakness of Delta 3D Printers that uses three or six diagonal rods for movement of all three directions – X, Y, and Z axis in space – is the compound error effect. A tiny error in any mating part, i.e. 0.2 free play at upper side joint of a diagonal rod, produces big error in printing.
The Marlin firmware assumes perfect set up of mechanical parts. Auto level probing could n ot hasn’t remedied the titled bed problem caused by uneven length of diagonal rods. Constant change of effector, which dictates the Z height – the distance between print bed and upper three end stops that limits the carriage movement in three axis, makes impossible to set up the base plane. Each calibration produces new base plane 0.1 mm to 0.2 mm off because of the slight length variation in diagonal rods.
I think this is one of the reasons why many Delta 3D Printer owners, who started using a Kit that uses cheap mechanical parts, eventually upgrade their 3D Printer with high precision, quality metal parts mostly in related to linear / rotational movement.
I’ll write an article about these kind of error once I sort current problem out.
I have to decide new set of diagonal rod and ball joint either metal ball joint and threaded tube or homemade magnetic ball joint and hollow tube. It will takes one or two weeks to get the parts and materials.
What would I do for the waiting period? This madness of 3D Printer drove me to do intermediate solution. As long as there is no freeplay in diagonal joints and make them keep in glued position, Bulbul Junior will print objects fine.
So I bought a 200g of plaster, two sheets of fabric yesterday to reinforce the diagonal rods. The plaster wouldn’t stick on bare carbon fiber tube in the first trial. The second time I applied wood glue on the tube then wrapped band of a fabric it. It seemed to hold the plaster.
How will the plaster coating handle the tiny, continuous impact and vibration while the 3D Printer in operation? I’m uncertain. I only hope it will hold for one or two weeks so that I’ll have printed plastic objects, mastered the art of calibration.
Current Delta geometry values:
X (Endstop Adj): -6.60
Y (Endstop Adj): -5.61
Z (Endstop Adj): -4.86
P (Z-Probe Offset): X21.90 Y-8.60 Z-6.20
A (Tower A Position Correction): 0.00
B (Tower B Position Correction): 0.00
C (Tower C Position Correction): 0.00
I (Tower A Radius Correction): 0.00
J (Tower B Radius Correction): 0.00
K (Tower C Radius Correction): 0.00
R (Delta Radius): 110.06
D (Diagonal Rod Length): 212.31
H (Z-Height): 211.50
-0.630 -0.630 -0.630 -0.690 -0.830 -0.830 -0.830
-0.290 -0.290 -0.440 -0.600 -0.730 -0.820 -0.820
0.160 -0.050 -0.260 -0.460 -0.620 -0.770 -0.890
0.400 0.140 -0.130 -0.360 -0.590 -0.760 -0.860
0.600 0.340 0.080 -0.170 -0.410 -0.590 -0.740
0.470 0.470 0.160 -0.090 -0.330 -0.520 -0.520
0.300 0.300 0.300 0.080 -0.150 -0.150 -0.150