Question Tap vs Klicky - thermal expansion of nozzle for probing vs printing... confused

[Leight616]

Printer Model

Siboor 350 Voron 2.4 R2 - Build in Progress

Hi all... first post here.

I've decided to take the plunge and ordered the Siboor 350 kit. The recent upgrade comes with Tap as standard, which (in principal) seems excellent as i like the idea of the probe being the nozzle.

However... i am somewhat perplexed about the workflow for levelling where the thermal melting limit of the build plate means one would be limited to 150oC for probing, and then needing to heat up to the ABS (or other) printing temperature.

the question... is there any issue with this? is there any significant thermal expansion in the 100oC change between probe and printing???

Klicky seems that it would resolve this problem, but introduces potential errors related to probe vs nozzle location and Z offset. I am intending on printing with a 0.6mm nozzle but like the idea of super simple swaps without needing to faff around with resetting offsets etc.

I've been bed-slinging printing for years, but this will be my first CoreXY.... i really really don't like inductive probes as I've had many many issues with them on various cartesian printers. I did actually get a Klicky pcb kit too so i could switch over.

I did search but didn't get a satisfactory answer... thanks in advance for your council

 

[Sandman50]

I have a 2.4 350 with TAP.

In practice whatever thermal expansion there is doesn't seam to matter.

Klipper assumes a cold bed and head for initial Z adjustment. (the paper test). What this does is tell klipper the distance from the nozzle to the distance the nozzle needs to travel for the TAP opto sensor to trip.

Then the basic process would be to do a 30 minute heat soak at 100*C bed temp and the nozzle at 150*C, then do a first layer squish test with your chosen filament.

The variation is adjusted out for the 1st layer squish. So it doesn't really matter if the head grows by 0.01mm as my squish Z adjustment compensates for that growth.

For ABS I like a 50*C chamber temp, which takes about 30-40 minutes to achieve. Whatever you do, you'll want to come as practically close to your real print conditions for your Z adjustment. (note than the nozzle can't be hotter than 150*C when it contacts a PEI bed.)

 

[MobileMcC]

I generally do the probe calibrate macro once, then adjust it manually in the config file after a couple test prints to get a first layer I'm happy with. I'm horrible with anything analog (a 'little' friction on the paper, 'barely' touching something, etc), so if anything I pick a less squishy base setting and then make it squishier until I like it and save it in the config file that says not to change it.

 

[ChrisA]

the question... is there any issue with this? is there any significant thermal expansion in the 100oC change between probe and printing???

I just looked it up. Brass will expand about 20 parts per million per degree C. So if you heat it from 150C to 250C you would expect the nozzle to be longer (length of nozzle)*(2000/1000000)

Let's guess the hot part of the brass is 20mm long. Then if I did the math right, the nozzle is 0.04 mm longer. Is this worth worrying about? It affects the thickness of the first layer.

A fun experiment would be to lay a high-quality 123 block on the bed and home to that with a 150C nozzle and then heat to 250C and home again and see if my above prediction is accurate. The 123 block is made of hardened steel ground to very high precision (0.0001 inch) and would not be harmed by the hot nozzle.

Footnote: Thermal expansion is not linear and I could not find tables that start at 150C, They seem to start at 20C. This is why I say "about 20 PPM". and it depends on the exact kind of metal the nozzle and other parts of the hot end are made from. Some heat blocks are copper and some are aluminum. You would just have to do a test with a 123 block.

 

[lrissman]

I concur what @MobileMcC said. I have an earlier siboor 350 kit that I changed out the microswitch at the back of the bed with a cnc tap. I never even thought of the expansion of my nozzle, having used copper, brass and hardened steel, and my zoffset is good. I find the QGL a much greater impact -- that can be a bit of pain to get dialed in with a backwards bowden applying pressure to the stealthburner if you have a stiff filament.

The bed mesh has a big impact on the perfect print too.. be sure to do your bed mesh.

Another thing I've noticed, when your print does go wrong... say overextruding.. the tap allows the head to kinda bounce over the nub of extra plastic instead of knocking your print off the bed... good? bad? i dunno.. it is.

I also "break the rules a bit" because I frequently do back to back prints and waiting for a 350 bed to cool to ambient is... well... tests my patience. So I do my zeroing with the bed at print temperature... 70, 90, 110, etc. .

Here is how I handle that in my start gcode.. (variables are slightly different in cura... have that to if you want it:

print_start EXTRUDER=[first_layer_temperature[initial_tool]] BED=[first_layer_bed_temperature]
BED_MESH_CLEAR
M190 S[first_layer_bed_temperature]       ; wait for bed temperature so QGL is stable.                    
M109 S150 ; bring hotend to probe temperature
G28 ; Zero XYZ

QUAD_GANTRY_LEVEL            
BED_MESH_PROFILE LOAD=default
M109 S[first_layer_temperature[initial_tool]]       ; bring hotend to temp

; Nozzle priming
G90 ; use absolute coordinates
G1 X7 Y7 F3000; move to start position of row 1
G1 Z0.28 F240 ; move to .28 off bed
G92 E0 ; reset extruder counter
G1 X7.0 Y140 E10 F1500 ; move to end position of row 1
G1 X7.3 Y140 F5000 ; move to start position of row 2
G92 E0 ; reset extruder counter
G1 X7.3 Y7 E10 F1200 ; move to the end of row 2
G1 X7.6 Y7 F5000 ; move to start of row 3
G92 E0
G1 X7.6 Y140 E10 F1500 ; move to the end of row 3

Also note that my macro "print_start" is empty in printer.cfg in klipper so that first line does not do anything. I moved all my configs into the slicer. No real reason for this, just did it.