Capacitive proximity sensor defect? influenced by heat?

Hey everyone,

I hooked up a capacitive proximity sensor (C18A3-H-Z/BZ, cheap 10$ sensor from china) to my CoreXY for auto leveling purposes. I lowered the heatbed in small increments, till the sensor gets hit and then measure the distance between nozzle and bed, the value is then uploaded to repetier firmware.

The first print turns out really well, the auto bed leveling is doing a good job. But whenever I start a second print job, the proximity sensor is always waaaay of. If the sensor gets triggered, when the bed is 0.8mm from the nozzle at the first printjob, it gets triggered like 2mm off or more at the second print job and I always have to recalibrate the whole z probing.

The inbuilt poti seems to be a problem. I always have to readjust the poti at the second print, the achieve the trigger height, that I set up at the first print.

May the sensor be influenced by the heat of the heatbed? The sensor is rated for 65°C, but I guess it gets hotter because of the nearby 100°C heatbed.

The sensor is rated for 6-36V, I'm running it with 12V.

Does anybody have an idea, what could be wrong?

quite a few people have tried these capacitive sensors with varying results, in practice they have a pretty wide margin of error usually 10% , i swapped mine out for inductive sensors initially and tried using aluminium tape on the glass bed but i found there was a bit of variance in that as well,
eventually i gave up on auto bedleveling and simply stuck the inductive sensor on the frame and had a screw adjuster on the carriage and now my sensors detect the top of a M8 bolt, since then it's been flawless. induction sensors if you decide to swap tend to work best if they have a nice chunky solid piece of steel setting them off

---

-=( blog )=- -=( thingiverse )=- -=( 3Dindustries )=- -=( Aluhotend - mostly metal hotend)=--=( Facebook )=-

Capacitive proximity sensors are affected by temperature and humidity. The cheap ones are intended to detect whether an object is present or not, not to measure distance with repeatable accuracy.

Inductive probes give more reproducible results, but the low-cost ones need either a steel target, or an aluminium target at very close range (so not underneath glass).

I tried lots of different approaches to height sensing, and ended up using differential IR. I now manufacture and sell proximity sensor boards for 3D printers using this approach. You can find details in the ForSale section of these forums. Whether or not one of these would be suitable for you depends on the bed surface you use.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].