Piezo Probe for MK3
Ive been reading a number of posts that list the PINDA probe a source of error for proper bed leveling. Mine seems to be working fine, but i do know that inductive probes aren't perfect by any means. I have read good things about the Piezo probe by precision piezo, and bought one for my project printer. I haven't had any first hand experience with the probe yet but the science seems sound so I'm wondering if anyone at Prusa Research has had a chance to look at this probe yet? Obviously it would lower the hot end a bit and require different printed parts for the extruder assembly, but It seems like a worthwhile optional upgrade. I would be interested in any Z probe upgrades that come in the future as well.
piezo sounds like a physical probe, that would touch the bed and create a signal through deformation of a piezo crystal, if so, I suspect it would need a rise and fall mechanism to lift it out of the way for printing. similar to the servo operated microswitch solutions of yesteryear...
achieveable but extra complication,
Yea, I think Area 51 is correct, though I haven't researched the tech. I bought one of the probes from Precision Piezo, and its a PCB than mounts between the E3D hot end and the rest of the extruder assembly. I have read a few posts from other forums about the probe and it seems to have pretty positive results. The downside is that because the probe mounts between the hot end and extruder, it lowers the hot end a bit causing misalignment between the nozzle and stock fan cooling duct. It not a plug and play solution by any means, but that seems like an easy fix with some redesign work.
The real advantages I see are:
-the elimination of the probe offset
-the fact that it will theoretically work with any bed surface
-the quoted probe accuracy.
Obvious negatives are:
-required redesign of the extruder assembly
-reduced print volume
-cost (PINDA=$14.11 Vs. Precision Piezo=$31.26
If the probe really is as good as quoted, an increase in cost doesn't seem that bad.
Sounds interesting, keep us posted! please
inductive probes aren't perfect by any means
I was very surprised how well the pinda performed compared to others, I'm convinced that most if not all bed/mesh leveling issues we have on the mk3 are either software related (temp compensation) or environmental (temp/humidity or not using preheat, etc). its definitely not the probes fault, as clearly shown.
There are actualy 4 different ways to place a piezo:
1. under the bed. As the mk3 do not have screws with springs (no manual bed leveling) this would not work on mk3 without big modifications
2. between coldend and extruder. This would require an modified printhead
3. between printhead and x-carriage. If the nozzle touches the printbed, the preasure leads to an momentum around x-axis. So in a small housing (about 4-5mm thick) between extruder head and carriage, the preasure-change can be detected https://www.thingiverse.com/thing:2526978
That thing is not for the mk3 but the principle should be clear.
I would call that the best solution, only design the small housing and use 5mm longer screws to attach the printhead.
4. A piezo to be put onto the nozzle. In this case you don't do mesh bed leveling before every print but only if something changed. It is a quite simple solution often used in deltas.
The probe do not work as an endstop, as it is only attached for mesh bed leveling. You need an classic z-endstop. That would not be a problem I think; you may add an hall-effect z-endstop and an connector for the piezo probe.
As I remember you can read out a mesh and write it via gcode (on a duet its definitly so, on an einsy I not tested it yet). So you can store meshes for different spring steels and temperatures as gcodes. After changing the spring steel or the beds temperature, simply "print" the mesh changing gcode.
This is the method I want to try on some chineese clones adding mesh bed leviling; but as my pinda works fine, no need to change mk3 now.
Be aware mesh bed leveling with piezo is not an good idea with a hot or dirty nozzle. A dirty nozzle will give false results; a hot nozzle may damage your bed. The heatbed of course should have the desired printing temperature, as bending may ocure with increasing heat. Always a good idea to wait 10 minutes after preheat the bed to let it get warm evenly (and not only at the point the termistor sits)
For option 1-3 there is of course the problem of homing also. If your nozzle oozes, the z homing would not be correct. So you have to home with a clean nozzle in cold state (but bed heated) or use a seperate z-endstop.
Piezo is a nice way for mesh bed leveling. But with the need to have a clean nozzle mesh bed leveling before every print seems not to be a good idea. Also use a z-endstop for homing avoiding any errors due to not clean nozzle is always a good idea. meaning changing firmware, as mesh leveling sensor and homing use different inputs.
Also: Auto Mesh Bed Leveling on a prusa mk3 looks also for "strange" results. Thei are coused mostly by some plastic got between heatbed and spring steel. So actualy it has a reason why prusa checks for every print...
Altogether, as nice the idea of piezo is, it has side effects that make it only interesting for advanced users who want the extra precission and take time to prepare before every print. For the common user Pinda definitly is the better solution.