Testing Resolution / accuracy , Your results wanted! Z Probe Results
-
DjDemonD and dc42
can you run the test 2 times ( please post one more run of each. I need more data to do the equally with the rest of the data!)
also David, can you run the IR probe you have built, do you have it installed on something you can test? that one is a real mystery still
know one has given data on it anywhere! Thanks!Jrlederer,
That was kind of you to gift a Bord!
Can I ask you to trade an old version 0.8.5 for a Duet Wifi? I have a good never used ( was for a project that i haven't finished yet) 0.8.5 that i did some testing with. then i upgraded that new Delta to a duet wifi. I would just sell it and get a Duet Wifi but no funds for that…
( hey mamma always said its never a bad idea to ask)and for some reason you say yes but dont want the 0.8.5 ill do a give a give away on my YT channel.. ( sponcerd by you )
ok.
Here is the data from AndreS
still waiting on he second run from you other guys! ( please)your data added to the workbook.
Higher rezhttp://open-source-energy.org/rwg42985/russ/Projects/3D%20printer/micro%20switch%20Z%20probe%20Data/Bed%20Probes%20accuracy%20others%20added.jpglooking for more! I want to make a video on this some time so if there are more users why haven't contributed please do so soon!
Thanks!!! -
Not heard of that Russ please do post the links?
ok i posted it here so we dont mess up this thread
https://www.duet3d.com/forum/thread.php?pid=15107
~Russ -
Okay so second run for Piezo20 Hotend Z-Probe www.precisionpiezo.co.uk
Stationary:
Bed probe heights: -0.111 -0.116 -0.116 -0.121 -0.116 -0.116 -0.116 -0.116 -0.116 -0.116 -0.116 -0.116 -0.111 -0.116 -0.111 -0.116 -0.111 -0.111 -0.116 -0.116 -0.116 -0.116 -0.111 -0.116 -0.111 -0.121 -0.111 -0.116 -0.111 -0.116 -0.116 -0.116 -0.111 -0.116 -0.116 -0.116, mean -0.115, deviation from mean 0.003
Done printing fileMoving:
Bed probe heights: 0.004 0.017 0.013 0.019 0.010 0.013 0.013 0.014 0.017 0.014 0.013 0.014 0.013 0.017 0.019 0.022 0.019 0.019 0.013 0.018 0.010 0.010 0.017 0.017 0.019, mean 0.015, deviation from mean 0.004
Done printing file -
ok updated Charts.
I found a few errors on my data selection so I corrected those, That change a few results. but , all for the better.
its all correct now…
also, i added Davids data but would like to have the other data set. all ways better to have a pool of data as a one time run can make you look to good
heheyour data added to the workbook.
Higher rezhttp://open-source-energy.org/rwg42985/russ/Projects/3D%20printer/micro%20switch%20Z%20probe%20Data/Bed%20Probes%20accuracy%20others%20added.jpg -
Russ, if you have a Duet 0.8.5 that you don't want, you can probably find a buyer for it on the reprap.org forum.
My delta is down for maintenance right now, I'll generate another set of date when it is running again
-
New Data, Precision Piezo Titan/Aero Piezo Z-probe bracket https://www.thingiverse.com/thing:2404904
15:48:16Bed probe heights: 0.005 0.005 0.000 0.000 0.000 0.005 0.000 0.005 0.005 -0.005 -0.005 0.000 0.000 0.005 0.000 0.005 0.005 0.000 -0.005 0.000 0.005 0.000 -0.005 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 -0.005 0.000 0.000 0.000 0.000, mean 0.001, deviation from mean 0.003
Done printing file
15:47:10M32 0:/gcodes/cartesian_Probe_Testing_Single_Point_36_probes_P_values.gcode
File 0:/gcodes/cartesian_Probe_Testing_Single_Point_36_probes_P_values.gcode selected for printing
15:46:55Bed probe heights: -0.015 0.000 0.005 -0.005 0.000 -0.005 0.000 0.005 0.005 0.005 0.000 0.000 -0.005 0.005 -0.005 0.000 -0.005 0.000 0.005 -0.005 -0.005 0.000 -0.005 0.000 -0.005 0.000 -0.005 0.005 -0.005 0.000 0.000 0.000 -0.005 0.005 -0.005 0.000, mean -0.001, deviation from mean 0.004
Done printing file
15:45:47M32 0:/gcodes/cartesian_Probe_Testing_Single_Point_36_probes_P_values.gcode
File 0:/gcodes/cartesian_Probe_Testing_Single_Point_36_probes_P_values.gcode selected for printing -
Tevo Little Monster with BL-Touch converted to Duet WiFi and 0.9 degree steppers on X, Y, Z.
7:50:34 PM: Bed probe heights: 0.149 0.161 0.155 0.158 0.149 0.155 0.158 0.155 0.152 0.143 0.136 0.146 0.115 0.102 0.090 0.036 0.021 -0.003 -0.020 -0.026 -0.023 0.018 0.061 0.121 0.130, mean 0.098, deviation from mean 0.066
7:56:14 PM: Bed probe heights: 0.149 0.183 0.186 0.155 0.152 0.083 0.049 0.021 0.005 -0.014 -0.026 0.002 0.058 0.118 0.155 0.177 0.186 0.186 0.149 0.146 0.130 0.111 0.086 0.043 0.018, mean 0.100, deviation from mean 0.070
7:58:47 PM: Bed probe heights: 0.015 0.015 0.008 0.015 0.018 0.018 0.021 0.015 0.015 0.018 0.018 0.015 0.021 0.015 0.015 0.018 0.018 0.018 0.015 0.011 0.015 0.011 0.015 0.018 0.005 0.008 0.008 0.011 0.018 0.018 0.018 0.018 0.015 0.015 -0.004 -0.010, mean 0.014, deviation from mean 0.006
8:01:26 PM: Bed probe heights: -0.020 -0.026 -0.029 -0.029 -0.032 -0.035 -0.032 -0.035 -0.029 -0.035 -0.039 -0.039 -0.039 -0.035 -0.042 -0.042 -0.042 -0.035 -0.035 -0.023 -0.017 -0.001 0.024 0.046 0.068 0.080 0.102 0.108 0.118 0.115 0.118 0.124 0.127 0.127 0.136 0.140, mean 0.021, deviation from mean 0.069
Probably explains my troubles getting a consistent first layer.
-
I still have not figured out how to get the kind of results you all are getting.
I am using the dc43 IR unit and it works well but something must be wrong with my build.
One thing that might be important is that repeated calibration runs get better after the 2nd or 3rd but then begin to get worse.
Have no idea that that means.
Frederick
-
IMO both the BLtouch and IR sensor have their limitations, BL touch is not as accurate as it really needs to be, the notion that if a probe is within 50 microns its good enough is clearly not correct. IR works fine if you have an evenly reflective surface to IR, and on a delta, no effector tilt. In practice neither of these conditions are easy to meet. IR on a cartesian with an evenly reflective surface is a good prospect. Join the piezo revolution guys
-
…or wait for our delta effector, which is close to availability. See https://duet3d.com/wiki/Smart_effector_and_carriage_adapters_for_delta_printer.
-
I have seen on rare occasions a .00* with my IR probe. Most of the time I am between .010-.020
-
Hi,
To remove any issues with my Z-probe can I use the "paper strip" method on the nozzle touching to bed to perform a calibration?
Thanks.
Frederick
-
…or wait for our delta effector, which is close to availability. See https://duet3d.com/wiki/Smart_effector_and_carriage_adapters_for_delta_printer.
I am certainly willing to give it a try if it helps me locate my problems.
As a long term solution I don't know if it will help me as I am attempting to equip my printers with Diamond hotends which, of course, need a different mount.
Thanks.
Frederick
-
As you may have already noticed I ordered a Piezo from you right after I posted those results
I was getting decent results until I foolishly decided to have it do a quick calibration before each print like David talked about in his blog. That's when the BL-Touch inconsistency truly bit me. I started getting first layer adhesion problems that shifted from spot to spot on the build plate. I even tried adding in the G29 mesh leveling but I still could get good adhesion at all spots on the bed using PrintBite. Hopefully I can get deviations like other posted in the 0.005 or less range as opposed to my current close to 0.1 range with BL-Touch. BTW the Tevo Little Monster comes with the BL-Touch from the factory and I had hoped it would work well because people at MRRF 2017 seem to like it on their printers. I'm not impressed with it though.
IMO both the BLtouch and IR sensor have their limitations, BL touch is not as accurate as it really needs to be, the notion that if a probe is within 50 microns its good enough is clearly not correct. IR works fine if you have an evenly reflective surface to IR, and on a delta, no effector tilt. In practice neither of these conditions are easy to meet. IR on a cartesian with an evenly reflective surface is a good prospect. Join the piezo revolution guys
-
Or you can paint the bottom of the glass black.
-
You can but it's still an offset probe though, which can mean nothing or can be a source of frustration. With our Piezo system, and David's new system, the nozzle is the probe, so you swap from buildtak to glass and just hit calibrate/level and print. Sure you might need to babystep 50-100 microns if the surface hardness is quite different, but the only variable is z offset and we are talking 0.1mm between soft and hard surfaces.
I have huge respect for the ir probe, compared to the inductive and capacitative probes it was originally designed to better it was a big step forwards, in almost every way, and is a lovely unit. But times have moved on, especially where deltas are involved, my IR sensor now lives in my parts bin. I like the mod using it indirectly with a rod as the probe but it's still offset in x and y, and a few graphics on your bed surface and your entering height differences at each probing point, which is a tedious workaround.
Leverless high quality microswitches are as accurate and repeatable (slightly better in some cases) than piezo, but that's within one probing run/one deployment, if you test with deploy, probe, stow and repeat you'll see repeatability start to drift.
I've only just started to play with underbed Piezos but so far the sensitivity is amazing. I reckon you could blow on the bed and it would register a trigger, maybe not quite that sensitive but it's very sensitive. This is a plausible solution for deltas and diamond hotends at present.
-
interesting, would bed sensing work for cartesan?
I was going to redesign the mount so i do not loose so much z height but the bed might be a better option if i can get it to work.
-
Yes it can work. As long as the bed is supported on mounts that can squeeze or bend piezos https://www.thingiverse.com/thing:2446460 then sure it can work, those would need a little redesigning with corner shaped recesses in the top parts, and the mounting screw maybe moved in the lower part to keep the mass over the middle of the piezo, but why not? Good for complex hotends like diamond where mounting options are limited.
The mounts though have to be thick enough to prevent the piezos going above 60 deg C or they behave erratically. So you're still talking 15mm.
I haven't done it myself on a cartesian.
BTW I just tried blowing on the bed of my microdelta and it triggers the z-probe
(and I turned down the sensitivity as it was triggering when anything moved in any direction at first.Just put my jewellery scales on the bed and it registers a trigger at 3-4g of force.
-
The only problem I see with underbed sensing is that you need multiple sensors and that the sensitivity may differ depending on where the nozzle will touch the bed. I am currently designing a new mount for my i3 clone that integrates the piezo without loosing too much Z height.
-
You'd think that but my experience so far suggests that it isnt a real-world problem. Sure if you had an arrangement of sensors in such a way that pressure on one part of the bed caused cancelling of the sensor outputs, i.e. one sensor was pushed and another pulled in equal measure then yes I could see that being a problem, but this doesn't happen with 3 sensors equally spaced around a circle, I have tapped across the bed in any conceivable problem areas such as the perimeter between two sensors without issue, and it wouldn't happen with 4 sensors placed at the corners of a square.
Perhaps with a 3 point bed mount, with two sensors at corners and the other side with one sensor in the middle of that side, in the same arrangement as 3 point bed levelling screws, there might be a spot on the bed where this effect comes into play. Someone more mathematically gifted might care to try to work it out, where is Ian/Deckingman when you need him? But I'd be willing to say just a slight tweak to one of the mounts, for example the one supporting the middle of the side of the bed, to make it slightly asymmetrical and it wouldn't be a problem.
