Skip STEP calibrator
I purchased an upgrade kit to take my 3d printer to a MK2 .
In my original Prusa I had a problem of the motors SKIPPING their steps . This would essentially destroy the print and would result in stopping the print for the scrap and starting again . Unfortunately, not all prints need to be perfect so if you had a system of re-calibrating mid way in a print it would /could potentially save the print. I attempted to play with calibrated encoders to make the print process automatic but this calibrated heater bed may prove to be a simpler solution and has further potential.
METHOD1- A button pushed on the Prusa tells it to stop printing and recalibrate on a calibration point zone that has no print body. This would then align the extruder back to the proper co ordinates.
METHOD2 - If you suspect your print skips STEPS a function in the software should be set so that after N steps the printer automatically checks itself .
METHOD3 - By having an inductive pickup on each AXIS a calibrated markers PCB strip can be used and the software check itself every time it passes such an inductive test point. This means that the printer would not need to stop its print mode and move to a calibrated point on the heated bed for calibration as per above. Although ,it would not replace the heated bed calibration it would reduce the step skip and this may lead to cheaper motors being possible or less failed prints. It also may improve prints that are far away from the heated bed (High Z).
Re: Skip STEP calibrator
You are trying to fix the consequence instead of elimination of the cause.
Find out why the stepper skips steps and fix this.
Re: Skip STEP calibrator
I take your point but in the event you have a large print and a few steps drop out due to a weak motor or bias of driver the last thing you need is a trashed print. Also the printer setups stretches over time and heat and temperature so a encoded location helps keep the print very tight .
Re: Skip STEP calibrator
I don't like your method anyway.
First, this way of position calibration has terrible repeatibility. Every time it calibrates it will get slightly different result.
Second, especially when printing large parts, you will get into trouble trying to reposition the extruder back down to calibration pattern/device without collision with printed part.
Third, if X or Y stepper skips steps in the middle of printing whole layer, the rest of the layer will be ruined anyway. It will most likely ruin all print as well.
You need some type of skipped steps detection capable stepper drivers to be able to see lost steps just at the time when it happens. There are such drivers, they are based on Back-EMF stall detection, but they are not present on current printer control boards.
High precision optical encoders could be another solution. But you need something with resolution of 10 microns (single step on X/Y axis) or better, along with device capable reading their output fast enough and provide feedback to control board.
Re: Skip STEP calibrator
David is once again correct 
Prevention is better than cure...
Peter
Please note: I do not have any affiliation with Prusa Research. Any advices given are offered in good faith. It is your responsibility to ensure that by following my advice you do not suffer or cause injury, damage…
Re: Skip STEP calibrator
Thanks for input David .
a) If the repeatability is poor with this method would this not mean that the Prusa MKll bed alignment approach is an ineffective method ?
b) These crude encooder strips can be designed to be located every 10mm as they are only used to cross check extruder location. As a result their pcb surface area can be designed to give a consistant trigger at a predesgned interval economically.
c) If these crude encoder like devices above ( or something similar ) run near the XYZ belts the Extruder would not need moved to a "calibrate point " and hit the object being printed because the extruder location error ( a skip ) will be immediately picked up as soon as STEPS are skipped. So no need for a skip detect. The whole point for this is a skip detect and recalibrate.
Possibly a small portion Z layer will be missprinted but at 0.5mm is likely to not affect the overall print.
TY
Re: Skip STEP calibrator
a) If the repeatability is poor with this method would this not mean that the Prusa MKll bed alignment approach is an ineffective method ?
Not at all.
Inductive probe is quite accurate measuring distance from metal object. But you are trying to use it detect in horizontal plane, it is much more inaccurate this way.
PR uses tha same method in XYZ calibration, but it is quite different application. The purpose is to get as accurate position of probe points as possible, but if you run XYZ calibration repeatedly, it gets a bit different results every time, but it is OK because difference of 50 microns actually doesn't matter. Moreover, XYZ calibration does every measure many times and averages the results, that's why it takes so long. Do you really want to wait for "lost steps detection" that long after every single printed layer? I don't.
Possibly a small portion Z layer will be missprinted but at 0.5mm is likely to not affect the overall print.
Either I don't fully understand you or you don't fully understand me. I'm sorry if my wording is not clear.
If you print anything (size doesn't matter) and printer skips steps, you get the rest of the layer shifted. Shift distance may vary from few microns to many milimeters. Even if you in next layer correct the shift of the extruder to (almost) the same previous position, next printed layer will be printed above previous shifted one. They will not match. Even if resulting look doesn't matter, you will get significant loss in strength and layers coherence. Especially ABS will suffer from that and probably crack just at that failed layer.
To be honest, I think your proposed way is completely wrong, no offense.
You need either high precision optical encoders (on motor pulleys or belt idlers) or Trinamic stepper drivers with stallGuard activated. And massive changes to the firmware, of course. And probably 32bit CPU.
Re: Skip STEP calibrator
Just to update and close this issue. The problem was interesting. I examined the power supply and noted the voltage was below 12V (If I recall about 11.8) . This caused a number of issues namely, the motors overheated and skip steps started to occur. This was a very easy fix adjusting the voltage on the pSU recovered these problems. I note today Prusa has incorporated recovery as per this blogs ideas .