3D printer - Piezo Z probing

[Gina]

Following on from another thread, I've decided to try piezo-electric sensing using the nozzle for Z probing the bed level and been looking at various methods of mounting the piezo-electric disc.  I'm planning to try this system on my most accurate (Mini) printer where I'm using a small nozzle for precision printing.  In conjunction with small nozzle comes thinner layers and the need for more accurate nozzle positioning and height. 

I realise that this system requires that the hotend has vertical movement to stress the piezo element to generate the signal whist printing accuracy requires maximum horizontal rigidity.  I seem to recall that other SGL members are using this method and wondered just how the conflicting needs of vertical detection yet rigid horizontal mounting were overcome.  I would appreciate details of the designs used, please.

Edited February 17, 2019 by Gina

[Gina]

There are several mounting ideas on Thingiverse.  eg.

1. Precision Piezo - Piezo20 Hotend Z Probe - the official Precision Piezo design.
2. Universal piezo bed leveling - a more compact version of the standard Precision Piezo design.
3. Piezo z-probe for E3D Titan Aero (mounted direct via motor) - interesting idea which would work just as well with the standard E3D Titan extruder which I have.

No.1 is a very bulky design.  No.2 is a "slimline" variation but not sure of the rigidity.  No.3 is a completely different way of doing it and looks solid.  However, I've just thought of a snag.  The whole of the mass of motor plus extruder will be slammed into the bed!  The extruder is quite light but even the slimline 20mm motor is no lightweight.

Edited February 17, 2019 by Gina

[Dr_Ju_ju]

Forget the slamming into the bed, when probing the speeds used are slow and gentle.

[Gina]

So no.3 above should be alright then?

[Dr_Ju_ju]

Yep, looks good, & it was developed by the PP designer....

The device is soo sensitive you can mount it just about anywhere on the extruder assembly. Mine even 'fires' when doing normal moves, but nothing gets upset, and is quite normal.

If I get a chance later, I'll take some pictures\movie of mine going, to give you an idea of what happens.

[Gina]

That would be great - thank you ?

[tekkydave]

I'd go for No3 too. The amount of movement needed is tiny and probing speeds quite low. This design is basically the same as how mine currently works. The upward force on the nozzle pivots the motor triggering the piezo.

Don't be tempted to go for under-bed mounted piezos on a corexy.  According to what I have read it only works reliably on systems where the bed isn't moving in z.

[Gina]

I shall definitely go for option No.3 above.  From experiences with a rather less than rigid nozzle with the standard E3D Titan extruder and V6 hotend I'm considering upgrading to the Titan Aero which is much more rigid.

[Dr_Ju_ju]

As promised, a 'homing' video & my current config.g file...

Hopefully it will play ok, I've down converted it to an avi file,  to fit on here, and tried my best with LED lights to reduce banding, which is difficult as they work @ silly frequencies.... 

Looking to the right of the E3D logo, you can just see a small red led, which is the PP off led, as it moves to the left (homing X\Y) you'll see it flicker, & can just see the blue led light com on (PP triggered) & note the speed I've set for Z homing, slooowww, but after homing, all speeds run as normal.

Config.g.txt

Edited February 18, 2019 by Dr_Ju_ju

[Gina]

I appreciate that, thank you.  Yes, the red LED is very clear, the blue one less so.  I've had an email to say my bits have been posted (I opted for 1st class post) so may arrive tomorrow - can't wait ?  I have the printer set up with the inductive probe so ready to print the parts.  I shall be glad to get rid of all the faff of setting up inductive probes.

[Gina]

Tried printing the file for the part in No.3 above in PETG but evidently my Mini printer isn't quite good enough - the gap in the top half... um... isn't... well... it partly is but won't separate fully.  Guess it needs a smaller nozzle than the 0.3mm I've been using.  No problem with sticking to bed with a 3mm brim.

Guess this shows the the upgrade I'm planning for this printer is definitely needed!

Edited February 18, 2019 by Gina

[Gina]

Printing again but with the thickness (Y axis) scaled up to 150%.  See if that works...

[Gina]

Yes, that's printed fine ?

[Gina]

Piezoelectric disc and PCB have arrived.

[tekkydave]

One thing I found is that the polarity of the piezo is not always the way you expect from the wire colours. Best thing is to follow the instructions on the PP website for setup. The blue led should trigger when pressure is applied upwards on the nozzle. If it triggers when you release the pressure then the piezo is reversed - just flip the dupont connector around and re-test.

[Gina]

Read the installation instructions on their web site.  Board now has just one pot which they said was set up before purchase but might possibly need adjusting.  Clockwise reduces sensitivity.

There was no polarity marking on the PCB for the sensor but I examined the electrical layout and determined which was ground which should be the substrate or negative wire from the sensor.  Added a connector to the bare wires of the sensor and plugged it onto the amplifier PCB, connected the power input to my bench PSU set to 4v and touched the sensor - red went OFF and blue LED flickered.  Tick.  Working ?  It's certainly very sensitive - barely had to breathe on the sensor to trigger it.

[tekkydave]

My config.g if it helps

config.g.txt

The M558 has loads of options - see the Duet wiki for all the possibilities.

Edited February 19, 2019 by tekkydave

[tekkydave]

The tricky bit is getting the Titan mounting screws just tight enough to hold securely but also leave a small amount of compliance to allow the piezo to be compressed

 

[Gina]

Yes, I can imagine that.  I may use a different mounting method.

Here are photos of my current X carriage.

[Gina]

I'm considering putting the sensor in the X carriage plate.  With the sensor in the slot in the printed spacer, I only have to tap the table to trigger it.  I couldn't manage to take a photo and tap the table at the same time and catch the response.  (Take no notice of the colours of the wires in the 3 pin plug - it's just a handy wired connector for testing - in this case red is ground and orange is +ve.)

[Gina]

Still thinking about how to arrange the piezoelectric sensor.   With the X carriage above, the main plate (which is PETG) bends very slightly if I push on the nozzle.  I should be able to use this movement to compress the sensor.  When I rebuild this printer as I plan to do for better rigidity, I shall use plates front and back of the X rail so a connection between top of front plate to top of back plate would compress.

[Gina]

It works but I think I may need to turn the sensitivity down as it triggers rather too easily.

[Dr_Ju_ju]

Try reversing the sensor connector, as that can affect the sensitivity.

[Gina]

It's currently set so that increase in pressure triggers it.  Pressure on the nozzle puts pressure on the sensor with this geometry.  I think it's just a matter of reducing the sensitivity setting.  Any movement of the print head causes triggering, lifting the bed only caused triggering once before the bed touched the nozzle.

I've tried reversing the sensor connector and it made no difference but I've also turned the pot slightly clockwise to reduce the sensitivity and it now only triggers when the bed touches the nozzle.  It still triggers with any XY movement but that doesn't matter, I gather, as probe triggering is ignored except when probing.

Only testing ATM, with bench PSU and watching the LEDs on the PCB.  Tomorrow I'll check out the Duet settings and connect the PCB output to the Z endstop on the Duet and remove the inductive probe.  Then I can do the final setup and adjust the sensitivity if needed.  I also need to attach the amplifier PCB to the X carriage.

Edited February 19, 2019 by Gina

[Gina]

Been reading the Duet Z probe/endstop information.  Although the Precision Piezo is mentioned it doesn't say which connector to use.  Since Mode 8 is being used (according to Julian and Dave) and it says Mode 8 is similar to Mode 5 and that Mode 5 uses the Z Probe connector I deduce that the 4 pin Z Probe connector is the one.

Edited February 19, 2019 by Gina