Underbed Piezo-electric sensors
This page refers to using inexpensive piezo discs for nozzle contact detection in FFF printers
Piezoelectric disks have a number of characteristics which make them very suitable for use as contact sensors but also some problems which have to be taken into account when using them.
Inexpensive piezoelectric discs have been used for touch sensing in the past on projects like maze running robots. There have also been several postings in 3D printing forums speculating that they may be usable as nozzle touch sensors for FFF (FDM) 3D printers. The first known use of piezoelectric discs in this manner was by Njål Brekke  who fitted piezo discs to respond to the bed pressure transmitted through bed leveling springs.
The piezoactive material in most piezo disks is a ceramic known as PZT which expands or contracts when a voltage is applied to it; it also generates a voltage when it is put under pressure or strain. The direction of this is determined by something called "polling" which is done in manufacture and in piezoelectric disks this is radial, pointing from the centre of the diameter outwards. Expansion or contraction of the ceramic in this direction will cause the brass disk to which the PZT is bonded to dome outward or inward. There are other directions of sensitivity, particularly that between the two surfaces of the ceramic disc: This is known as the thickness axis or sometimes as the axial axis.
The very high sensitivity of piezoelectric discs is the single greatest advantage, outputs of several volts in response to a light nozzle contact are typical while repeatability may be in the region of a few µm. The voltage output from a piezo disk using the radial axis, that is one allowed to flex, is about 3 times the output from one where pressure is applied to the surface, in the thickness axis. Conversely, the compliance of a sensor using the radial axis will be higher than one using the thickness axis. Although naked piezoelectric discs must be treated with care, they are exceptionally robust when installed. They are also very inexpensive, being perhaps only a few percent of the price of any cometing sensor (e.g. FSR) The signal conditioning requirements for the output of a piezoelectric disc to be used by a micro-controller are very undemanding, it is even possible to use the output directly with only overvoltage protection.
- Most mountings explored to date have some degree of cross talk, that is the piezo will respond to X or Y movements. Z probing operations should take place only after any mechanical shaking from X or Y movement have died down.
- Piezoelectric disks suffer loose sensitivity with increasing temperature, this is particularly evident in "no-name" units where a maximum usable temperature of 50ºC or 60ºC is the maximum working temperature.
- PZT suffers from a pyro-electric effect which causes a voltage to be generated with changing temperature. If the low end cutoff of the conditioning electronics is too low then changing temperature may be interpreted as a nozzle contact.
- Piezoelectric disks detect an event, that it the nozzle coming into contact with the bed. The do not detect a status and if the nozzle is already in contact with the bed when a call is made to perform a Z probe operation then that call may fail.
- There is some documentary evidence that PZT deteriorates with time. This has not been confirmed in practice although it may be more severe in a heated environment
None of the above problems have been particularly difficult to overcome and a number of sensors are in regular use without any problems being found.
Geometry of Contact Detect Sensors
It is not generally practical to install piezo discs under the print bed in the same way as FSR sensors pioneered by Johann as piezo disks have a significant decline in sensitivity at temperatures above 50ºC. Getting good thermal isolation is easiest done by moving the piezo element away from the heated bed.
To minimize the compliance, Leadinglights developed a piezo disk mounting that converted increasing pressure on the build stage to reducing pressure on the piezo element on the thickness axis. This was accomplished with a 3D printed parallel mechanism.
A later development of this eliminated the parallel mechanism by having a piezo with a hole in the center through which the build stage mounting operated. This method was found to be unsatisfactory when used directly across the piezo as the output had two peaks, possibly from distortion to the disk during drilling. A modification which allows the piezo disc to be stressed in the radial axis does work well although it does increase the total compliance of the bed mounting slightly.
Moriquendi has developed an underbed piezo system which uses piezo disks attached to stressed load members. The deflection required to generate a readable signal with a piezo sensor is so small that 3D printed (ABS) parts are adequate for this purpose. The stressed members are designed as compliant mechanisms arranged such that bending is concentrated in the region below the piezo sensor thus maximising the output signal for small deflections. This is similar to the way in which foil strain gauges are used and uses the piezo element in a semi-radial manner.
Note that Njål Brekke, the first person to use piezoelectric disks now also uses them as strain gauges.
Piezo Electric Hotend Probe
DjDemonD has developed a Piezo-electric hotend probe using a single piezo disc mounted to the effector of a Delta printer. This is a very lightweight unit when combined with a direct drive extruder such as the Zesty Nimble or Flex3Drive
pyr0ball has developed a Self-calibrating Universal Probe called the EasyPiezi that can be used with a single disk or an array of piezo elements in a number of different mounting schemes, including on-head, under-bed, or on-gantry mounts. The EasyPiezi also features I2C and UART communications inputs for on-the-fly sensitivity adjustments.