Precision Piezo Z-probe Now available.

Got Piezo 20 on my Anet A8 but i wanna put this sensor to Ender 3... anyone got mount for this sensor or tell me how to mount on Ender 3??

Quote
Moriquendi
@Ohmarinus,

That's not actually a piezo sensor at all, it's an FSR, a force sensitive resistor, they are much more delicate than piezo disks and not as accurate.

Idris

I have received the sensor today. Will try it out first thing this evening. I'll let you know how it went.

---

http://www.marinusdebeer.nl/

@Krakow2000,

Not yet but I've just bought an Ender3 myself so I will be working on sensors for it soon.

Idris

---

{Precision Piezo} Accurate, repeatable, versatile z-probe plus piezo discs, endstop cables, pt100, 50w heaters.

first of all, i like and use the reprap forum quite often and it's a gread pool of ideas.
I'm really impressed of some ideas.


A few weeks ago, i got my precision piezo v2.85 and i'm impressed how sensitiv it is and, how could it be, i've two questions to Moriquendi.

1.) Is it possible to use piezo rings instead of the piezo discs. (I searched for it, but i didn't get a right answer)

2.) For a new Printhead i want to implement the electronic directly to a pcb layout. The latest circuit version i found was 2.11, a few pages before.
Is there a way to get the scheme of the current version.

Thank you very much.

Hi Snowwolf,

I did reply to your email but perhaps it got lost (assuming you're the same person, the questions are the same)

As far as I'm aware the commonly available piezo ring elements are quite different from the piezo buzzer disks that the circuit was designed for. I have no experience of the rings and no way to know whether they would work or what might need to be changed.

The current generation of products are not open source and therefore I do not make the schematic available. I'm happy to sell you a pcb without the pins soldered if you want to mount it direct to your PCB via straight pinheaders.

Idris

---

{Precision Piezo} Accurate, repeatable, versatile z-probe plus piezo discs, endstop cables, pt100, 50w heaters.

Hi Moriquendi

you are right, i wrote you an mail direct from your homepage.
Thank you for your answers.
I ordered some rings and I'll test it.
I'll, see what happen.

Thank you for your offer, but my target is only one pcb where all is included.
I think i'll test it with the v2.11 version.

Is this version also working fine, or are there some bigger problems with it?

Many thanks

Bit of an odd situation here. Getting setup on Marlin and it seems I am getting a ton of false triggering when my Z axis first starts moving..having several Piezos I figured ah thats alright I will turn the sensitivity down a bit so its not doing that. However, while that solves the issue of not triggering from Z-axis movement it then requires the probe to dig into my bed quite a bit before triggering..

G1 Z-10 F200 is an example of the movement speed that can cause triggering. Any recommendations? Making my probe more sensitive makes it more prone to Z axis noise while not making it sensitive enough makes it dig right into my bed which defeats the purpose.

Here is my config file for Marlin 2.0: [pastebin.com]

Did you reduce your acceleration and jerk to absolute minimum for the probing run ?
Sorry, not familiar with the Marlin configuration stuff.

Hey Jens53, thanks for the feedback.

I actually realized my custom Jerk values were not being used as I forgot to enable that code block

Z values now
-My Z jerk is now 0.2 so I will try that in an upcoming test
-Accel has been 500mm/sec


Ideally the situation would be when Z starts to probe its not a sudden jerky movement and then its able to reach a higher speed as needed for the probe to trigger properly. I guess if anyone knows values that people use for something like a CR10 that would be useful. Normally I havent had to adjust the sensitivity on my orions from stock.

Open to any other suggestions.

The one other bad thing about lowering my jerk and accel for Z is that they will be like this for the entire print unless maybe I change them in my startup Gcode only aroudn the G29 command.

On a side note is there any benefit to use 5V over 3.3V for the probe?

There's no practical difference.

Idris

---

{Precision Piezo} Accurate, repeatable, versatile z-probe plus piezo discs, endstop cables, pt100, 50w heaters.

Okay so after being all stubborn trying to make leveling work in other ways I am considering getting piezo's for my delta printer.

Because I want the effector to be as small as possible I have chosen for the underbed solution. I am asking myself, will the nozzle damage the kapton surface when probing? I can imagine that the leveling process only has to be correctly done once because the machine is solid and will not move after leveling everything with UBL.

I haven't been able to find much about designing a mount for piezo bed mounts. Does the bed depress the piezo's or do the piezo's measure 'shock'/vibration of the nozzle hitting the bed? Or should the mount deflect and buildup pressure on the piezo?

What's the differenc between the 20mm and 27mm disc besides the diameter? Any difference in sensitivity? I would assume smaller is easier to design a mount for.

Edited 1 time(s). Last edit at 12/26/2019 05:40AM by Ohmarinus.

---

http://www.marinusdebeer.nl/

I have recently been doing some considerable research on underbed piezos to try to find out why some people find thme to be the best and others find them unusable. Although I have used these on all of my own printers I had not discovered some of the problems that they can have. In brief, having piezos mounted too far in from the edge of the bed can give bad results and 27mm are better than 20mm piezos. There are other things worth noting but I will include them when I report what I have found.

In answer to your specific questions, the piezos respond to pressure rather than vibration. There is a limit to how much pressure piezos before they become less sensitive, Under direct pressure (not bending) the 27mm piezos work well with up to 1.6kg per piezo while the 20mm with up to about 0.75kg - I am less sure in the bending case.

Edit: The contact force and speed for a good piezo setup is so low that it won't mark any surface. I use a Z speed of 2mm per second and get contact pressure of less than 25 grams in the worst case - less than 10 grams is normal.

I am away from base at the moment and am also using a Chromebook on my lap while surrounded by overexcitable children and relatives but hope to put the results of my research in this forum by early next week.

Mike

Edited 1 time(s). Last edit at 12/27/2019 05:14AM by leadinglights.

Quote
leadinglights
I have recently been doing some considerable research on underbed piezos to try to find out why some people find thme to be the best and others find them unusable. Although I have used these on all of my own printers I had not discovered some of the problems that they can have. In brief, having piezos mounted too far in from the edge of the bed can give bad results and 27mm are better than 20mm piezos. There are other things worth noting but I will include them when I report what I have found.

In answer to your specific questions, the piezos respond to pressure rather than vibration. There is a limit to how much pressure piezos before they become less sensitive, Under direct pressure (not bending) the 27mm piezos work well with up to 1.6kg per piezo while the 20mm with up to about 0.75kg - I am less sure in the bending case.

Edit: The contact force and speed for a good piezo setup is so low that it won't mark any surface. I use a Z speed of 2mm per second and get contact pressure of less than 25 grams in the worst case - less than 10 grams is normal.

I am away from base at the moment and am also using a Chromebook on my lap while surrounded by overexcitable children and relatives but hope to put the results of my research in this forum by early next week.

Mike

This sounds really good! Thanks for your answers!! Right now what I notice is that the FSR I'm currently using needs a lot of pressure causing the readings to be somewhat inaccurate.

Did you use this kit?
[www.precisionpiezo.co.uk]

25 grams sounds exactly like what I'm looking for. Is it so that the piezo elements are only used during probing? Because I'm afraid my machine might vibrate a lot and trigger the piezo's during printing. Of course, during a print I don't need the piezo's to work. I could easily design new bed mounts for the piezo setup. I'll have a special holder made with the CNC machine at the art academy where I work. Or maybe only a plate with 3D-printed mounts on top that allow the round bed to rest on those. With 25 grams I wouldn't worry about deflection during probing.

---

http://www.marinusdebeer.nl/

I don't use the Precisionpiezo conditioner although I did make one identical to earlier Precisionpiezo models and it worked very well. I have been trying to come up with an MCU based conditioner which would allow a piezo to be used without tuning. Unfortunately, this looks a little like a pipe-dream at the moment as I found that the fine details of the output varied too much across the build surface to simply replace a microswitch.

For the moment, and until I can put together a fuller report, I would recommend the following: Mount the 3 piezos at the tips of a triangle that either fully encloses the circular print bed or comes close to doing that. The exact details of the mounting are limited only by your imagination - as long as the pressure of the nozzle contacting is converted to pressure or bending of the piezo element

If the input to the controller is dedicated to the Z probe then there will be no problems if the probe gives spurious contact signals during non-probe operations - they will simply be ignored. An allied problem that can give problems is that an XY movement between each probing operation can cause false triggering. The answer in Repetier is to have a delay between the end of an XY movement and the beginning of a Z probe operation (Zpop). I think this is called DELAY_BEFORE_Z_PROBE in Repetier firmware. I believe all other firmwares have equivalent settings.

Mike

I have two deltas with Piezos as under bed sensors. Work perfectly since more than two years
[reprap.org]

Quote
Dancer
I have two deltas with Piezos as under bed sensors. Work perfectly since more than two years
[reprap.org]

Interesting. It looks really simple. So the round part with the cutout is mounted underneath the heated bed and rests on the piezo disc? The cutout part and the piezo disc go inside the holder?



Edited 2 time(s). Last edit at 12/29/2019 02:30PM by Ohmarinus.

---

http://www.marinusdebeer.nl/

Yes.

Some pictures:
[reprap.org]

I call the lower part "cups". They are mounted, so that the top parts (inserts) are not completely loose but little bit rubbing on one of the inside wall of the cups.
Config is still the same as on the pictures taken in August 2017 - and many rolls of filament have gone down on this printer

Edited 1 time(s). Last edit at 12/29/2019 02:57PM by Dancer.

Quote
Dancer
Yes.

Some pictures:
[reprap.org]

I call the lower part "cups". They are mounted, so that the top parts (inserts) are not completely loose but little bit rubbing on one of the inside wall of the cups.
Config is still the same as on the pictures taken in August 2017 - and many rolls of filament have gone down on this printer

Ahhh I see. Now I understand the cutouts are for the soldered wires. I was already wondering about that.

I assume you are Dutch, where did you order your parts? I am thinking of ordering the board from Precision piezo, it's probably easiest to get the piezo's there as well.

---

http://www.marinusdebeer.nl/

Sorry for my late reply.

No, I'm not Dutch, I'm Austrian

The piezo discs are from ebay - don't know which exact - listing does not exist anymore, but nothing special about them.
Th circuit was built on breadbord according to a link I found in one of the first entries in this forum.

Should be this one:
[www.thingiverse.com]

Original full Circuit + with my notes attached to this message.

Continuing from a few weeks ago, when I was not in a position to give any answers, I have done a bit more of a write-up on my piezo investigations and reported on these in the Tech-Talk subforum [reprap.org]. In brief though, I have found that there are some dragons lurking in the underbed sensors of any type. While these dragons can be slain – or at least kept at bay, this is not an area for the unwary. For a taster of the problems, I have put a diagram below graphing the analogue input to the piezo conditioner at 8 positions on my build stage.



a summary of what I have found that works to get good sensing: Avoid electrical and particularly mechanical noise so that the trip point can be as low as possible. Do not place the sensors too far in from the edge - If the bed can be wholely within the triangle formed by the sensors that is better. Avoid too much compliance in the system - Although I use piezos in pressure mode my latest sensors have little hinge springs that create more compliance than if I had used sensors in a bending mode.

Mike

As written years ago, the simplest and best way is to integrate one SINGLE sensor (any kind of, not just piezo) in the hotend/carriage.

---

"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)

As heard many times over the years, and just as WRONG now as then. The best sensor depends on the use: Multiple interchangeable heads, ultra-thin layers etc.. Saying there is only one best way is like saying there is only one king of road vehicle (well O.K., only the 1938 SS Jaguar need apply)

edit: That was supposed to be "some kind of road vehicle" but I will leave king as it is.

Mike

Edited 1 time(s). Last edit at 01/11/2020 01:03PM by leadinglights.

Quote
leadinglights
As heard many times over the years, and just as WRONG now as then. The best sensor depends on the use: Multiple interchangeable heads, ultra-thin layers etc.. Saying there is only one best way is like saying there is only one king of road vehicle (well O.K., only the 1938 SS Jaguar need apply)

edit: That was supposed to be "some kind of road vehicle" but I will leave king as it is.

Mike

There are certainly exceptions, but not multiple heads or ultra thin layers (???), may be the etc...

BTW, you document pretty well all the issues with under bed sensors.

Any way, these 3D printers just spit blobs of molten plastic and to speak micron is preposterous.

---

"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)

FFF printer don't really spit plastic, they really sort of schmear plastic onto something - which is mostly already schmeared plastic. As far as I know, there is not a lower limit to the thickness; people have been printing 50 and even 25 micron layers are not out of the ballpark.

Multiple print heads may have had nozzles changed etc., so need to have the new nozzle height set. See [www.youtube.com] This is a printer which is really not amenable to nozzle contact sensors in the heads

Ultra-thin films are more sensitive to shifts caused by temperature changes and I have used under-bed piezos around the 1 to 3 gram range. I have never had that sort of sensitivity on hot end sensors.

BTW, because I am getting disappointing results on my new printer I will probably use a single underbed sensor for setting the nozzle heights backed up with an optical sensor using two colours and a confocal lens to do the bed mapping/leveling - I rather like DC42s IR sensor but hopefully even more immune to problems with different surfaces.

Mike

Edited 1 time(s). Last edit at 01/11/2020 02:48PM by leadinglights.

My 3rd 3D-Printer.
A Prusa Mendel i2 with Under-bed-piezos in the "cups" shown in the middle of this page: Standard Deviation: 0.001034

Quote
Dancer
My 3rd 3D-Printer.
A Prusa Mendel i2 with Under-bed-piezos in the "cups" shown in the middle of this page: Standard Deviation: 0.001034

So you get similar if not better results with three simple piezo in a cup and a few op amps than leadinglights far more elaborate set up ? How comes ?

---

"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)

I do not know.

But there are 4 Piezos... (Standard 220 x 220 Heatbed, 4 Holes on the Edges)
Most important is to wait 1 second before each probing, because the piezos trigger from the Y-movement.

Pictures of the Printer to come (SD-Reader not working in PC...)

It certainly is possible to get excellent results with a simple setup - or bad results as well. When something has got the possibility of dynamic behavior it becomes a bit of a lottery - rather like building a bridge and finding out it wobbles (millennium bridge) or falls down (Tacoma Narrows bridge) even though you have used the best design techniques. My Delta printer is almost spectacular but my newest printer has problems that need some very lateral thinking.

Mike

As promised - pictures of the printer and setup.

I'm absolutely happy to have the Piezo Z-probe, because especially on this machine, if the two Z-Steppers are not perfectly aligned, leveling with the piezos compensates for it.