Precision Piezo Z-probe Now available.

Quote
AlmostNeverAgain
Hi Simon and Idris,

Building a custom cartesian printer, I was about to buy your piezo20, but I see your multi ceram Orion, and I don't know the advantage buying this one instead of that one.
Could you please clarify by comparing both ?

On the cartesian side, the Z travel is more restricted than of a delta, and every Z centimeter counts ! So, your Z probe integration into my design is really critical,
trying to get the exact dimensioning of your module, based on your published STLs.
It is a little a nightmare to find the latest version of your STLs, there is no PDF blueprint, no dimensional description.

Familar with CAD conversion, I managed to skin your STL with a STEP envelope, and i am very surprised to see how bad is your part design!
One example, I splitted your STL into the 4 parts, trying to re-assemble them... the holes patterns of the 3 screw-able parts are not exactly the same, impossible to CAD-align. the holes axis are not perfectly perpendicular to the side, impossible to fully constraint.

To sum it up, I think that you should publish the exact and latest dimensional data, either by providing the STEP/IGES CAD files, or front/side/top schematics.
Eventually, I would be more than happy to contribute by designing quality CAD files, of course.

Thanks in advance.

EDIT:
- The probe20 assembly based on the parts draft from the STL measures 24mm approx between upper/lower groovemounts. (2mm gap between piezo plate and lower part).
- Assembling the probe20 to a BMG extruder seems not possible, the upper groove entering too much into the BMG extruder body.

The parts that Idris supplies fit perfectly with one another, and yes I do use a BMG and a Titan with the Full metal E3D V6 with the provided mounts, yes the first one I was sent was a bad print, Idris very quickly replaced it with a working one, I does fit very well into the BMG body, just don't forget your 5mm protrusion of PTFE otherwise it will give you issues. You say you skinned an STL to a Step file? would mind explaining what software you used. As to the holes, I am not sure where you measured your points from but the file I have from Idris and the printed product show that they do line up correctly. IMHO what you think they should or should not publish is down to Idris and Simon, after all, it is their design and the older gens are already in the far east market, I can fully understand why they have not just given out the Cad files without due consideration.

Edited 1 time(s). Last edit at 04/18/2018 08:13PM by Dunston-UK.

Leonardo DaVinci did the same and added some errors in his drawings to keep them from being copied by spies

The advantages of the Orion over the Piezo20 are:

Accuracy, the Orion module has shown in tests to improve upon the Piezo20 module and acheive 0.005mm accuracy (5um)

Ease of assembly, the Orion module has only three parts and five screws. Brass threaded inserts already in place.

Rigidity, the Orion module does not rely on the strength of the piezo disk to hold the clamp in place. Ease of tuning,no fiddling with assembly bolts during tuning, just screw it down tight and leave it alone. Also includes easy tune tech from the PP20 v1.1 circuit

Durability, no delicate piezo solder joints in the Orion module. Brass threaded inserts make for much more robust connections.

Flexibility, groovemount and screwmount each with the option of bowden or direct drive feed. Dedicated kits for the CR-10 family and TronXY X5s are on the way. Hotend clamps for E3D Chimera/Cyclops and Microswiss hotend under development.

Modularity, Hotend clamps and printer mounting adapters are modular, simply print a different mount/clamp if you change your hotend or printer

Integrated bowden fitting, bowden versions come with an integrated bowden collet allowing use of a wider range of hotends without issue.


The parts for the piezo20 module are admittedly a bit of a mess, they were first made at the beginning of the project in TinkerCAD which is not an ideal tool for this sort of thing. The problem is, they worked so they've not been cleaned up. The Orion parts that Dunston-UK is referring to don't have any of those problems.

The documentation and files part of the website does need overhauling, it's on the list of things to do but it's a long list.

Idris

---

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

I have used both the PP20 and Orion now and can definitely recommend the Orion.
I had issues with the initial groovemount supplied with the Orion as it was a bit loose in the Titan. This was speedily rectified by Idris with an updated STL. It is now nice and snug.

Quote
o_lampe
Leonardo DaVinci did the same and added some errors in his drawings to keep them from being copied by spies

If only this were the case. No I designed the Piezo20 on Tinkercad so it is not geometrically perfect.

I'm not a professional at CAD or engineering. Idris is however and AlmostNeverAgain you'll find the Orion unit, which is his design, much more mathematically pleasing,

Edited 2 time(s). Last edit at 04/19/2018 05:08AM by DjDemonD.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Leornado would have loved your designs, and the piezoelectricity, for sure !
Chinese are writing from right to left, as Leornardo did, to hide written words, but not for the same reasons !
I know, a lot of people are not playing the game of the OSH, making CAD files release like a real issue.

For sure, the Piezo20 parts seem like a quick and easy initial design, and I don't criticize this at all, please take no offense, but the least assembly need mat, flush and orthogonality. even 0,01° is an issue.
On the opposite, Orion STLs found on thingy are very well designed indeed, the STEPs I got from them are just perfect.

Am convinced by the Orion' clear advantages. Adopted. Does the Orion electronics bring something new compared to the Probe20?

About printer features, losing 19mm on Z on a cartesian is not a detail, thinking on an alternate solution where the PCB/piezos sandwich could be compressed by the frame, with really less height.

Just one question: If a frame-based (bracket, hinge,...) solution would allow to trigger the endstop when PCB/Piezos compression is released, do you think it is possible and efficient using the Orion?

Preload compression (release) > Endstop ON
More (adjustable) compression > Endstop OFF

Thanks for your feedbacks.

To AlmostNeverAgain:

Not familiar with your proposed design. Perhaps an under bed configuration may be more appropriate for your design if the build plate and the support carriage only moves in Y direction, the Z with effector/hot end moves vertically for your proposed design.

I have tested all except the early versions of the hot end. The Orion beta, the PP20 V 1.1, the V2 with/without mounting holes and the whole range of piezos from 12mm ,20mm,27mm, 41mm available from Mutate as used by Simon/Idris. I have found the different sizes do not matter as the sensitivity can be compensated for with the adjustement pot(s) provided. The main consideration may be the resonant frequency of the printer. The Rf for the piezo ranges from 9 kHz, 6.3 kHz, 4.6 kHz and 2.2 kHz from smallest to largest as published by Murata*ñ. Personally I am torn to use the under bed vs the hot end versions. I have tested on the Tevo Black Widow and the Tronxy 3. I can attest to the accuracy of the various designs with the Orion for its superior design, ease of mounting, accuracy of probing. I have NOT tested these on a delta, so no comments.

I have had not so good results with some far East products and applaud Simon/Idris/Group for their innovations .
Regards,
Stef

Edited 2 time(s). Last edit at 04/19/2018 09:20AM by Chowa.

The circuit implemented on the Orion is practically the same as that on the latest Piezo20 PCB, the advantages of the Orion are in robustness and ease of use though accuracy has also been improved.

If you really want to conserve z-height then you will need probably need to engineer something different to the Orion, unless you can mount it higher on the carriage to "hide" the height of the module. The Orion is designed to be a complete module, easy to use, but this does mean that it's not optimised for every situation. The Universal PCB is designed for people who need to do something different.

Piezo endstops are possible, I've been using them very effectively for a long time now.

Idris

---

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

To Chowa,

Indeed, I could go underbed for the piezo detection, but designing swappable beds, the Z detection would have to sit on the cradle (supporting the beds), why not.
Z detection on the printhead bracket is my favorite to avoid going this way.
I do appreciate your experience description and noise data.
Am convinced by the superior quality of Precision Piezo products, and I support european businesses. Am mainly focused on efficient integration. Lot of options!

To Idris,

Already considering the Universal Kit, or building around your Piezo20 PCB.
I appreciate the clever design of your Orion PCB/Piezo sandwich, and I was about to ask you, if it is possible to just buy the Orion PCB/Piezos.

By the way, even if I have CNC mill and diamond endmill, I settled very early to only use undrilled piezos.
Depending on the way I install my detection sandwich, and its size, I could try several piezo diameters. Based on chowa [edited] comment, it seems that they just differ by their resonating frequency, and that their signal is proportional to their respective sizes.
Before going this way, first, the groovemount Orion, but with modified external structural parts.


Thanks to both of you.

Edited 1 time(s). Last edit at 04/19/2018 04:15PM by AlmostNeverAgain.

I should have clarified-- the larger the piezo, the higher the output voltage per given flex.
The resonant frequency may play a role in noise pickup that a specific printer may generate. To illustrate, if your printer generates noise during probing in the 2.2 kHz range, the 41 mm piezo would not be a good candidate as a sensor in my opinion. This was observed in moving x, y, z axes during my tests. The 12 mm piezo did not produce sufficient output even after tuning to give reliable results. I found the 20mm and 27mm piezo disks when properly tuned to be most reliable for my under bed application. I use one disk mounted at the center of the carriage to which is mounted the heated work surface.
I also have a glass work surface on top of the heated work surface. The glass being held in place with printed corner brackets. The Orion has two 12mm piezo in parallel which give sufficient output to trigger. I have previously posted images of my rig. When I get home will repost several fyi.

Regards
Stef

Edited 1 time(s). Last edit at 04/19/2018 02:06PM by Chowa.

Almost ever Again, no offence taken, and I'll be happy to see if you can come up with some new ideas.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Here are some images of my rig. I should also note that double sided tape ca 1mm thick was used to attach the piezo pucks to the carrier. the work plate is attache to the carrier with spring loaded screws to the corners of the x carrier.The thick D/S tape may provide some mechanical noise isolation.

To Chowa,

Thanx. Very nice boxes with a "candy" inside, but what is on top the green and yellow boxes? looks like foam, or something?
I previously envisioned to use strain gauge/load cell, some have the same cylindrical aspect, but the piezo solution seems easier to integrate, to tune up and offering a better resolution.
Another factor is the heat chamber, as it seems that piezo signal is widely reduced above 60°.

You mentionned 2 piezo in parallel? You mean, Orion has them in //? Why not in serie to double the tension ?


Thanks Simon

Edited 1 time(s). Last edit at 04/19/2018 06:58PM by AlmostNeverAgain.

Those candy boxes were my first iteration for underbed using Simon's stl for his mini delta. The last image 2121.jpg is the nest of the various piezo mentioned previously. I should also mention that the cables connecting the various piezos two are shielded audio cables and the shield connected to ground at the electronics. I also tested twisted pair and twisted 3 wire. Im leaving the piezo nest as is for any further testing that may be needed in the future. Of course, the Orion has printed circuits to connect to the electronics. Brilliant

Edited 1 time(s). Last edit at 04/19/2018 05:38PM by Chowa.

Just to answer your other questions 're the heat from the bed, in my rig you can see in the second image an aluminium covered foam blanket, dont know what kind but assume it is high temp, probably neoprene. Also assume the piezos in parallel rather than series for the Orion, Idris may shed light on that. From previous discussions parallel was used for multi sensors.
Stef

Edited 1 time(s). Last edit at 04/19/2018 05:58PM by Chowa.

There's no problem with signal available from the piezos, in fact we have to take precautions to prevent a hard trigger from causing a damaging voltage spike.

Temperature decreases the output of the piezo disks, practically what that means is that if you're using a heated chamber you have to tune at the same temperature you probe at.

Idris

---

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

Quote
Moriquendi
There's no problem with signal available from the piezos, in fact we have to take precautions to prevent a hard trigger from causing a damaging voltage spike.

Temperature decreases the output of the piezo disks, practically what that means is that if you're using a heated chamber you have to tune at the same temperature you probe at.

Idris

Heated chamber. Now there is a possible answer as to why I was having my issues with Orion which Idris should have had by now. At what temp is the Orion badly affected as I run my chamber at 80 deg when printing.

Quote

Temperature decreases the output of the piezo disks, practically what that means is that if you're using a heated chamber you have to tune at the same temperature you probe at.

Idris

Ever thought of adding temperature compensation in your amplifier circuit?

Quote
AlmostNeverAgain
...
1) Chinese are writing from right to left, as Leornardo did, to hide written words, but not for the same reasons !
....
2) Just one question: If a frame-based (bracket, hinge,...) solution would allow to trigger the endstop when PCB/Piezos compression is released, ..

1) wrong, Leornado wrote that way because he was left handed and it is "cleaner" when writing with ink.

2) right, this is the way to go, integrated in the carriage and it works. I did it, after a couple of quirks (like when the filament is pull when Z going down) has been Ok for a few months, just didn't use a piezo but could. A piezo is just one sensor among others.
Caveat, not a drop in for standard carriage, extruder, hot end. I made my own as at the same time I wanted something more compact, lighter and easy to mount/dismount and be my own !.
Suitable for a new line of machine though but is it worth it ? I do all this for fun.
So, again, you are on the right track. You have just to choose, use standard mounts, make it fit for the most common 3D printers or ignore what exists and go for something completely different that will be the optimum for you.
Indeed, as one can see in this discussion, to make it fit current set ups, follow up, is let say, annoying.

---

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

The effects start to show up at around 50c, I would expect to have serious problems at 80c. Heated chambers that go that high are pretty uncommon, I didn't think to ask. Adding temperature compensation to the board is possible but I think to ensure accurate compensation at every temperature would be quite complex (Read, expensive).

Idris

---

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

The heat chamber (and air humidity control, btw) is a very important device for many thermoplastics, and we can see that a lot of new, among them being quite high-end, filaments on the market.
This trend is clearly shown by high temp capable extruder, above 400°C, even if not having an heat chamber prevents from printing them.

Even the linear guide, like the MR12 have max temp around 60°C/80°C, so temperature affects many elements, beyond piezos. Fortunately, some manufacturers offer higher temp linear guide.

When the plastic comes out of the nozzle, it obviously interact with the ambient air, impacting the plastic properties, mechanical ones and surface aspect, too.
Some thermoplastics requires the heat chamber based on their properties, others thermoplastics are enhanced by using them.

For these reasons, finding solutions to keep the piezo sensor in an operational temperature range seems a "hot" item on todolist.
Insulation, heat transfer and indirect mechanical sensing, a combination of these 3, could possibly help.

MKSA, custom printhead carriage is something I am working on, allowing me to integrate the effector in the carriage structure.

Edited 2 time(s). Last edit at 04/20/2018 08:20AM by AlmostNeverAgain.

Temperature compensation and noise reduction goes hand_in_hand. We've discussed rectifier diodes before, which will reduce noise. then you can crank up the amplifier to a level ( read Schmitt-trigger ), that triggers reliable even at 80°C. The signal level is cut off anyways, so you don't risk killing the circuit by overvoltage.

Quote
o_lampe
Temperature compensation and noise reduction goes hand_in_hand.

Can you discuss your reasoning here? I'm not following...

Idris

---

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

Overvoltage? I understood that temp increase would have the ceramic voltage decreases, no?

It is not reliable to expose constantly a sensor frankly outside its domain, and the sensor lifetime could be affected, unless these sensor are thermally isolated, imho.
If you go high temp, in a constant manner, well above the ceramic piezo sensor specs, you probably need to find another flavor of sensor.

If you hit the piezos hard enough you can get a significant voltage spike, 80-90v is easily achievable, not something you want to have reaching your controller.

Leadinglights did significant testing on the the effect heat has on piezo disks (not all piezoelectric materials are the same, the PZT disks we're using are intended to be used as buzzers). I'll have to check the exact numbers but IIRC he found that the highest usable temperature for Murata disks was ~80c. However, once the disks cooled down again the performance was not impaired.

I think what o_Lampe is getting at is that if you have good enough noise cancellation you can set the circuit to be so sensitive that even with the decreased signal at 80c you're still getting reliable triggers without getting spurious triggers at lower temps.

Idris

---

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

Quote

I think what o_Lampe is getting at is that if you have good enough noise cancellation you can set the circuit to be so sensitive that even with the decreased signal at 80c you're still getting reliable triggers without getting spurious triggers at lower temps.

Exactly. That's called improving noise/signal ratio. One way is the reversed Piezo, the other cheap and easy method are the rectifier diodes.

One thing not being considered is that Idris already has clamping Shotky diodes on the input circuit. These clamp the piezo signal to +/- the forward voltage drop of the diodes. Putting in a bridge rectifier in advanced what is in there would be superfluous. In my under bed system I use one 20/27mm disk and the sensitivity is such that walking by on a wooden floor causes triggers. This is only an issue during initial probing at the start of a print job. I think that a more important consideration is the native noise frequency of the printer itself in selecting piezo size e.g. the Orion uses two 12 mm disks which have a resonant frequency of 9 kHz +/- 1kHz as per Murata spec sheet. The S/N is not an issue as there is more than enough signal to provide triggering. Just my opinion.

Stef

Edited 1 time(s). Last edit at 04/21/2018 11:36AM by Chowa.

The diodes on the input clamp the signal from the piezo to Vin + Vfd and GND - Vfd, they're silicon junction diodes so roughly Vin+0.7v and 0v-0.7v. What this does is route any voltage outside that range to the supply where the capacitance of the PSU (and elsewhere) makes it negligible.

Adding a rectifier to the input would cut the bottom volt or so off the signal (twice the Vf of the diodes in the bridge). It's something I need to look at with a scope, is the noise confined below that threshold? How hard do you have to hit the bed to exceed the threshold? Would it cause problems with softer bed materials? I need to investigate all of this.

Idris

---

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

Quote
Chowa
...and the sensitivity is such that walking by on a wooden floor causes triggers.

A guard dog that barks for any butterfly passing by, isn't a good guard dog.

Re: Diodes
I mentioned earlier, that a bridge rectifier has the side effect of triggering twice. First when the compression starts, second time when the pressure releases.
You are better off by using only the positive signal of the piezo with a single diode in series ( or two diodes ).

It is why the bridge configuration has been invented. DC42 uses it in its smart effector.

---

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