English
Login

P.I.N.D.A./SuperPINDA Sensor testing

Relevant for

MK2.5
MK2.5S
MK3
MK3S
MK2S
+
41 comments
Article is also available in following languages
English
Čeština
Polski
Deutsch
Français
Español
Italiano
日本語

P.I.N.D.A. stands for Prusa INDuction Autoleveling sensor and is a vital component facilitating the calibration of our printers. There are 3 different versions:

  • P.I.N.D.A. V1 - used on MK2/S and MINI.
  • P.I.N.D.A. V2 - used on MK2.5, MK2.5S, MK3, and MK3S. 
  • SuperPINDA - Compatible/used with MK2.5/S, MK3/S/+, and MINI+.

The difference between P.I.N.D.A V1 and V2 is that the P.I.N.D.A V2 has an inbuilt thermistor to compensate for the difference in cold and hot measurements. V1 and V2 are not interchangeable. However, SuperPINDA and PINDA V2 are interchangeable but may require a firmware update. The SuperPINDA does not have a thermistor, as its high-quality components and manufacturing remove the need for one.

For MK2.5/MK2.5S printers that have the SuperPINDA installed, you MUST use firmware version 3.9.3, or above! In addition to this, you must let the printer know you have the SuperPINDA installed, and not the P.I.N.D.A. V2, by navigating to LCD Menu -> Settings -> HW settings -> SuperPINDA -> ON. MK3S & MK3S+ owners using the SuperPINDA must use firmware version 3.9.2, or above!

Testing the P.I.N.D.A./SuperPINDA sensor

Firstly, we need to check if the sensor works at all. For this test, you need a small and thin metal object (spatula, scissors, pliers). Your printer should also be running on the latest firmware version for your printer model.

  1. Press the knob next to the LCD for 3 sec and turn it to raise the extruder approximately 10 cm above the heatbed. This is to distance it from any metal.
  2. If you have firmware version 3.5.3 and later (unless you have an MK2/S), go to the LCD Menu -> Support -> Sensor info and see the sensor value ('PINDA:' 1 or 0)
Before firmware version 3.5.3, the P.I.N.D.A. sensor value was found in LCD Menu -> Calibration -> Show end stops. If you have an MK2/S, you can find the Z-value there.
 
  1. Place the metal object (spatula, screwdriver, etc.) underneath the P.I.N.D.A./SuperPINDA sensor. The value should change from 0 to 1, and the red light on top of the sensor should turn off.
  2. Move the extruder by hand all along the X-axis and wiggle with the extruder's cable-bundle. The value should still be 1, as long as there is something metallic under the sensor.
One of the most frequent issues is an overtightened zip-tie, which causes the P.I.N.D.A. sensor cable to fail when it reaches a specific position. Double-check all the zip-ties and make sure they are not too tight.

Testing the P.I.N.D.A. sensor-readings in Pronterface  

There is another way to check whether or not the P.I.N.D.A. sensor works correctly. Please perform the following procedure 3 times. For the sake of consistent measurements, please make sure the tests are done under identical conditions. 

Pronterface may not be supported on newer versions of iOS (Apple Laptops) after they dropped the support of 32-bit apps with Catalina. However, there are some workarounds. See this Github thread.

  1. Connect the printer to your computer following the Pronterface and USB cable guide.
Make sure the nozzle is clean, and the print surface is clear of any obstructions.
 
 
  1. Issue a G80-command into the command-line. It triggers the mesh bed leveling process.
  2. Once the process is complete, issue a G81-command which gives you a report of the interpolated Z values for a 7 x 7 array of values.

Compare the data arrays. Keep in mind that every number in the matrix has their pre-defined positions, always compare the numbers in the same position. You want to make sure there are no significant deviations (>0.2) between the 3 measurements.

After adjusting the P.I.N.D.A. sensor or any other hardware, we recommend running the Calibration Wizard from the beginning.
 

19 comments

Log in  to post a comment
hitjim
Is the ">0.2" meant to directly apply to the resulting standard deviation calculation's bare numerical result or the percentage calculation?  For example when I use Google Sheets' STDEV do I evaluate the resulting bare value from my 3 measurements (0.003470778011) or the % conversion (0.35%)?
plaTime
I have an MK2S and the PINDA probe seems to have failed.  It doesn't trigger when doing the endstop testing.  Can I replace the PINDA V1 with a V3?   
Giuliano - Official Prusa CS
Hello. The SuperPINDA is meant for the magnetig heatbed, so the printer model should be MK2.5 or newer.
CD_FER
Sorry to be that guy but this line:"newer versions of iOS (Apple Laptops"
should say:
"newer versions of macOS (Apple Laptops"
Cassius Duke
Hello,
I have had a major failure, where the hotend termister wires got so hot they melted and the printer screamed.  I couldnt see why and the fuses were not blown.
I have replaced the V6 with the new E3D Revo and the issue looks like its rsesolved.  the board powers up and the hotend will heat, extrude, cool, XYZ move and hte PID callibration worked.
BUT the PINDA sensor is no longer working.  when i try to do a 1st layer calibration the nozzel hits the bed.
If I follow the instructions above I can see the red light, and I can see if go off when i place a metal object under the PINDA, but the Printer shows a zero the whole time, even when the lights is off.
I have a Mk3 that had the MK3s upgrade.  I have flashed the latest 3S firmware and  down flashed the previose one incase and it dosnt work.  Ive tried the non-s firmware and it has the same behavior.
Question: given that the light is on so the PINDA has power is it more likly that there is a broken cable or that the board is damaged?
Giuliano - Official Prusa CS
Hello. If there is no visible damage around the port for the PINDA in the board, it would be hard to establish what is broken. Certainly replacin the PINDA is less expensive, so I would try to replace that first. Also, in case it was not about the PINDA, you would still have gotten the newest version, which is more reliable.
Anachronist
I came across a nozzle-cleaning brush that is mounted by sticking it to the underside of the right edge of the build plate with magnets, for which you add a bit of gcode to pass the nozzle over the brush. Are there any magnets along the right edge of the build plate that are ignored in a 7x7 bed-leveling calibration? That might be a good place to position an extra magnet for mounting that cleaning brush.
David

Hi! Would the left side work? The right one is being probed by PINDA all the way to the edge.