Precision Piezo z probes guide for duet users reference

DjDemonD

We got in 3ders!

https://www.3ders.org/articles/20180206-fdm-3d-printer-bed-leveling-with-cheap-piezoelectric-elements.html

Lakko

So if i understand, if i have

M558 P5 I0 R0.4 F330
G31 X0 Y0 Z-0.1 P100

and i have always to do -0.15 babystep i have to change config.g into:

M558 P5 I0 R0.4 F330
G31 X0 Y0 Z0.05 P100

right?

whosrdaddy

yes that is correct.
If you have a recent firmware, you can use P8 instead of P5 (unfiltered vs filtered input)

Lakko

@whosrdaddy:

yes that is correct.
If you have a recent firmware, you can use P8 instead of P5 (unfiltered vs filtered input)

is this also valid for piezo20? firmware 1.20

whosrdaddy

Yes, and your firmware supports this mode

LumberjackEngineering

Silly question, but why is the red power LED so incredibly bright? I'm running a Duet, so Vin is 3.3V. It's nearly impossible to see the blue trigger LED without covering the always-on power LED!

DjDemonD

Well look into that it might be a batch of leds of different output to spec.

Lakko

@Lakko:

So if i understand, if i have

M558 P5 I0 R0.4 F330
G31 X0 Y0 Z-0.1 P100

and i have always to do -0.15 babystep i have to change config.g into:

M558 P5 I0 R0.4 F330
G31 X0 Y0 Z0.05 P100

right?

i noticed in my config.g file in the epilogue section a [c]M208 S1 Z-0.2 ;set minimum Z[/c]
this can afflict the z height?
so removing the M208 I can go back and insert again this?

M558 P5 I0 R0.4 F330
G31 X0 Y0 Z-0.1 P100

DjDemonD

No I just recommend setting z min to just below the bed so you can probe points that are below the nozzle when it thinks it's at z=0.

But you are right to change your G31 as you state in the first example to 0.05mm.

rmriggin

@dc42:

It sounds to me that you have a short in the cable or the sensor between the +3.3V and ground pins.

Of all things it ended up being the 5V/INPUT VOLTAGE jumper not being seated/making connection

Sensor works now but the only issue I am having is that there is no LED whatsoever. In the web control I can see the z probe change values when probing but the LED indicator on the piezo board does not illuminate whatsoever. I have attempted to provide 5v and ground to the LED pins on the board to no avail.

DjDemonD

The piezo board LEDs should illuminate if you just supply +5v and GND directly to it. If nothing happens then they may have failed. If the board works despite this it's up to you if you want to proceed, otherwise please contact us via our website (in my signature) we can arrange to replace it.

Agniusm

Ordered the module with 2 discs to test. I have E3D titan aero but its mounted on aluminium bracket and then bolted to rail which is vertical. Its very rigid.
What needs to be implemented for this to work? Is it deformation of the disc? Which disc is more sensitive, 27 or 20mm?

DjDemonD

There not much in it in terms of sensitivity. Deformation, compression or if mechanically feasible just vibration can generate enough signal to trigger the circuit.

Some sort of plate between the two with the piezo embedded in it can work well. Try compression first if that doesn't work support a disc on one side on its periphery and push into the centre on the other side with something narrower.

This is a working prototype https://www.thingiverse.com/thing:2526978

Agniusm

I wander if i could mill out my bracket to acomodate piezo disc and if so how deep do i need it to be?

DjDemonD

They are very inconsistent especially bought from ebay. The most consistent ones we use are Murata 7bb series either 27mm or 20mm. Try rs components or mouser for a datasheet. The brass disc is very consistent, the ceramic is usually slightly off centre and the solder for the wires can be fairly variable. From memory the 20mm one is 0.4mm thick and the solder is 1mm thick. You need to allow generous relief for the solder especially if its going to be in a metal bracket, maybe even some kapton tape over it. You need to allow your bracket to move by about 0.1mm when the nozzle hits the bed, but this amount of vertical compliance does not affect print quality.

PlasticMetal

I've installed a Precision Piezo kit on my SeeMeCNC Rostock v2 delta, using kraeger's piezo mount on top of a 713Maker mount, using the stock effector. I also have a SmartEffector for this machine which works very nicely, but wanted to try something different for a second hot-end. I've got a couple questions while trying to get the piezo running.

The piezo board blue LED lights up when tapped, but is tripping as soon as I issue a G30 to try a test probe. It seems I can adjust this in two ways - by adjusting the piezo board, or in config.g with acceleration (M201) and jerk (M566). However, being a delta, there isn't a dedicated Z motor, and so any changes I make have to be applied to all motors, which will affect normal printing. So, is adjusting the piezo board the way to go?
Per [https://duet3d.dozuki.com/Wiki/Gcode#Section_M558_Set_Z_probe_type]](https://duet3d.dozuki.com/Wiki/Gcode#Section_M558_Set_Z_probe_type]), M558 use of "X Y Z" is now deprecated (I'm running 1.20). What would the new M558 & M574 commands be?

Here are relevant bits of my config.g:

[c]M350 X16 Y16 Z16 E16:16 I1
M92 X80.0 Y80.0 Z80.0
M906 X500 Y500 Z500 E600:800 I50
M201 X1000 Y1000 Z1000 E1000
M203 X20000 Y20000 Z20000 E3600:1800
M566 X1200 Y1200 Z1200 E1200

; Z probe and compensation definition
M558 P5 I1 R0.4 F300 H5 X0 Y0 Z0 T2400 ; Piezo per duet3d.com/forum/thread.php?id=2769

; M558 P5 R0.4 X0 Y0 Z1 H5 F1000 T2400 ; P5=SmartEffector, R=recovery time in seconds, only set Z axis, H=dive height (height above trigger height probing starts), F=probe speed (was 1200), travel speed

G31 X0 Y0 Z-0.1 P100 ; Piezo per duet3d.com/forum/thread.php?id=2769
; G31 P100 X0 Y0 Z-.21 ; for SmartEffector

M208 S1 Z-0.2; set minimum Z (-0.2 normal operation, ~ -3 initial setup)[/c]

Thanks!

DjDemonD

Firstly which piezo board do you have?

Your point labelled 1) above you can set different parameters for z, unlike in older versions of marlin, in RRF if you specify z acceleration 100 it applies to z moves not specifically to the z motor. So your delta kinematics are not adversely affected (z moves are not actually that frequent after homing and probing).

My M558 looks much like yours lose X0 Y0 Z0 if you want but you are not using the probe as an endstop in a delta. F 1000 is probably a bit fast for piezo we recommend 4-7mm/s or 240-420mm/min.

My endstop config looks like this:

M574 X2 Y2 Z2 S1 ; set endstop configuration (all endstops at high end, active high)

Which is unchanged whether you have a probe or not, as on delta a probe is just that its a probe and is an optional extra. You still home to x, y, z max but then you can drop to z=10 and send G30 to set fine z height or use G32 to autocalibrate.

PlasticMetal

Thanks, DjD. Sorry - I had meant to include the board version - it's a 2.02, bought just a couple months ago.

Thanks for the clarification on XYZ acceleration. I had earlier dropped down to Z400 with no benefit, but then started to second guess if it was Z movement, or just the Z motor. I'll try 100 mm/s^2.

The "F1000" is actually in the commented out line for the SmartEffector - I've got the piezo set for F300.

I'll try these changes, and possibly drop M566 Z a bit too.

Thanks!

DjDemonD

Cool. Jerk isn't applied to probing moves in RRF so leave it set whereever you want, frankly in the z direction it doesn't need to be very high.

dc42

@DjDemonD:

Cool. Jerk isn't applied to probing moves in RRF so leave it set whereever you want, frankly in the z direction it doesn't need to be very high.

Jerk in the Z direction is used when applying mesh bed compensation. If it is set too low, the symptom is that straight XU moves will stuttrer.