Piezoelectric disks for Z contact detect and bed levelling

Thanks I may be dim and obviously it works but I can see how the hot end pushing up does anything to the bracket. The reason I ask is that I would to use a modify this thing to use a piezo sensor but can't quite wrap my head around how to do it. I though I would get away with the groove mount part but as it's direct drive, it doesn't fit the setup [www.thingiverse.com]

I see, this is a bit more of a challenge, I have been playing with a new idea, not published yet but you might want to try it. Its a plate which mounts between the motor and the bracket, you would have to mount your motor infront of the bracket sas shown here:


Which I have done by removing 2 (dont take all four out the motor will fall apart and is a PITA to reassemble) of the motor bolts, and then fixing the motor through the bracket with this plate sandwiched between them:
titanmotormountpiezo0.1.stl

The piezo goes in there and just gets squeezed when the nozzle pushes up.
Its very much a concept at this point but I've got one mounted which works okay.

---

Simon Khoury

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

Ah I doubt I could mount my mounter in front of the bracket as it would limit the printable area of my bed. Could it not sandwich between the motor bracket and extruder?

No, not as it is, but feel free to redesign it, the only reason I didn't got that route was that it needs more cut away for the motor shaft, and I wasn't sure it would work as well. But feel free to try it. Were giving away a free piezo kit to anyone who comes up with a new implementation. PM me if you're interested.

---

Simon Khoury

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

Ultimately i'm moving to a flex3drive. Is there already an option for that? If not, looks like i'll be designing one

No specific option in the sense that the piezo system is usually independent of the flex3drive. I have a flex3drive and just mounted it above the piezo sensor (this was on a delta so the flex3drive was above the effector and the piezo sensor below), running a short guide tube between them (for 1.75mm). The Piezo20 screwmount just needs 3 holes, one for filament and one either side 25mm spacing for the sensor module.

Edited 1 time(s). Last edit at 10/07/2017 08:25AM by DjDemonD.

---

Simon Khoury

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

Great. Care to share a picture. Once I get my flex3drive I can figure out the best method now I understand the theory. Incidentally, I ordered my piezo kit Thursday, how long does it take to ship (UK)

I think I shipped it yesterday so 3 days or so.

---

Simon Khoury

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

I'll find an image of my setup with the flex3drive.

---

Simon Khoury

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

Here the setup with flex3drive.


The version I have is the one designed to be integrated into a kossel mini effector, as such the lower part of the flex3drive housing is the effector itself in the original design, so I made an adaptor which has the flex3drive reliefs in it on its upper side (the white part) and has two mounting holes with heat pressed nuts to attach it to the effector. Different flex3drive versions probably don't need this and just have a flat base.

The effector here is a waterjet cut aluminium part, with a piezo disc in a mount like this [www.thingiverse.com] (the green parts shown there on thingiverse). The filament just enters a PTFE tube beginning inside the flex3drive and then goes down through the piezo sensor unit into the hotend.

Edited 1 time(s). Last edit at 10/07/2017 05:49PM by DjDemonD.

---

Simon Khoury

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

Since the first trials with piezo discs at the beginning of this thread I have continued to be surprised by them. Back then I set up a trial to see what output I could get from a simple pressure change across a disc, expecting to see maybe a fraction of a volt but I got 8 volts at a very low speed and fairly light pressure. The biggest disappointment was when I found that no-name piezoelectric discs had a very severe loss of output with increasing temperature, so much so that they were unusable above maybe 50°C or 60°C. Trials with genuine Murata discs were inconclusive as although they gave a higher output this seemed to be very variable; but this may be because I got discs without leads and damaged them by soldering leads on.

The situation changed completely when I started testing discs to find what sort of life they would get. Using Murata discs supplied with leads already soldered on, I get about 5% to 10% loss of output with simulating 5 years of very heavy use - 500 grams of nozzle contact relaxed every 5.4 seconds for 100,000 cycles. Heat trials were a bit of a surprise as I got MORE output with cycling at higher temperature and also the output at 80°C was quite usable and got better with time. These parts did have some reduction below the original output when returned to ambient temperature.

The biggest surprise came though when I tried a simple long soak at 80°C for 37 hours with simulated nozzle contact events only to measure the output: The output rose from 2.85V to 4.0V over this period.

The next thing is to see if I get the same with no-name discs.

Mike

Its great work Mike. Could I request that you include a drilled 20mm Murata 7BB disc in your analysis. Its just that's what a lot of our devices rely upon to function?

---

Simon Khoury

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

Quote
DjDemonD
Its great work Mike. Could I request that you include a drilled 20mm Murata 7BB disc in your analysis. Its just that's what a lot of our devices rely upon to function?

Will do although it may be a 27mm disc as I have one previously unused drilled one handy.

Mike

Quick question my piezo arrived today thanks, my existing z endstop only has 2 wires but the kits has 3. What do I need to do?

Edited 1 time(s). Last edit at 10/09/2017 02:34PM by totalitarian.

Okay so you would use our kits GND (black) and Sig (green) wires for your existing endstop connection (what controller board are you using?). But you need to power the piezo PCB with 3.3-5v from somewhere else.

---

Simon Khoury

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

It's a ramps board, I have 3 pins on the z-min (board end) so I can make up a new cable.

[www.aliexpress.com]

Okay this is a ramps board
[s.aliexpress.com]
But the one you showed me is based on it. The endstop connectors are 3 pin +5v (red), GND (black) signal (green). Did we not send you a cable or did you order pcb only?

---

Simon Khoury

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

Thanks, you did but it needed to be twice at long for my CR-10. It original had a melzi board with 2 pin end stops. No worries, almost got it going. Just need to modify my mount now

Cool

---

Simon Khoury

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

Can I PM you, I have the mount installed but get no red led on the pcb (v1.22) also unsure how the piezo leads should connect to it...

PM me, I make the boards and should be able to help you out.

Idris

---

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

OK i'm having a hell of time with this firmware (marlin) When I do a G28 it homes x & y and then homes z. It then seems to do a double z home but at a slower rate. This is slower rate causes a crash. Any idea how to get around this?

Hi yes thats an easy fix.

In configuration.h look for

// Speed for the first approach when double-probing (with PROBE_DOUBLE_TOUCH)
#define Z_PROBE_SPEED_FAST HOMING_FEEDRATE_Z

// Speed for the "accurate" probe of each point
#define Z_PROBE_SPEED_SLOW (Z_PROBE_SPEED_FAST / 2)

This sets the fast rate to be your defined homing_feedrate_z which should be around 5mm/s with piezo probes but you can go up to 7 or down to 4.

The next one is just that divided by 2.

So what I've done is just specify this numerically like this

// Speed for the first approach when double-probing (with PROBE_DOUBLE_TOUCH)
#define Z_PROBE_SPEED_FAST 5

// Speed for the "accurate" probe of each point
#define Z_PROBE_SPEED_SLOW 5

---

Simon Khoury

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

Thanks, for some reason that didn't work as expected.

Noticed this in the configuration_adv.h file

//homing hits the endstop, then retracts by this distance, before it tries to slowly bump again:
#define HOMING_BUMP_DIVISOR {2, 2, 4}  // Re-Bump Speed Divisor (Divides the Homing Feedrate)

changed it to

//homing hits the endstop, then retracts by this distance, before it tries to slowly bump again:
#define HOMING_BUMP_DIVISOR {2, 2, 1}  // Re-Bump Speed Divisor (Divides the Homing Feedrate)

A few posts back DjDemonD asked if I could compare the effect of temperature and the number of pressure cycles (nozzle contact cycles) for a drilled piezo disc. I have had one running now for 40 hours at 80C and the results look pretty good.



Top trace (dark blue) is a 27mm Murata Piezo at ambient temperature (20°C ± 2°C)with 500 grams of pressure applied directly to the surface. The pressure is reduced to 50 grams once every 5.4 seconds for 100,000 cycles (6 days 6 hours). The peak voltage was recorded at the point when the 500 grams load was re-applied.

Red trace is the same but at 80°C.

In order to see if the voltage increase was caused by the mechanical stress or the temperature another Piezo was tested with minimal cycling. The green trace is shows the output at 80°C under 500 grams of pressure which was relaxed only long enough to get a reading with cycling then switched off.

The violet trace was the drilled piezo with the load applied at the center. The load was 500 grams but was only relaxed to 250 grams (hence the small signal) for mechanical stability. The temperature again was 80°C and the test was run as a soak, that is, the pressure was only cycled to get readings and then left under load. I will keep this going for a week to see what happens but I don't expect much further change.

Mike

Is it just me, or are the diagrams so small as to be illegible?
V1.1_piezo_signal_conditioning_board.png

I'm very interested in using piezos - I have no doubt that this is the immediate future of cheap endstops - and am trying to figure out if I can/should make a board, or buy a precision peizo (which after shipping seems *rather* expensive), or does a cheaper chinese version exist?
Similarly is there a similar device that can be safely substituted....like a instrumentation or other variation of amplifier... already available?

re: the last option: I see there are plenty of 'piezo sensor' boards available, but they do not have trimpots or anywhere near the same electronic complexity as whats been discussed in these pages, so presumably they are'nt suitable for printer application purely because of the lack of adjustability...even before all the other electronic failsafes.
Can anyone enlighten me on the various pros/and cons of such a board for this application?
[www.aliexpress.com]

Also, to the electronics noob that I am, I see a functional similarity with instrumentation amps... binary "sound sensors" etc. Is this a valid assessment? If an amp with sufficiently high voltage tolerance were paired with a piezo disk that wasn't going to kill it (apparently 40V in the case below), would it work?
[www.aliexpress.com]

Quote
BETLOG
Is it just me, or are the diagrams so small as to be illegible?
V1.1_piezo_signal_conditioning_board.png ...............

I think it may be your browser or computer although the colours on mine don't make it clear.

Quote

.......
Similarly is there a similar device that can be safely substituted....like a instrumentation or other variation of amplifier... already available?

There are simpler alternatives, I have even seen piezo discs connected straight into an Arduino port without any other circuitry. The circuit developed by Moriquendi cleans up the signal and allows adjustment to most combinations of 1 or more piezo discs. One cheap alternative that has been tried is this one:- [www.ebay.co.uk] It is just a comparator with neither gain nor filtering but it works as long as there is not too much electrical or mechanical noise.

Quote

re: the last option: I see there are plenty of 'piezo sensor' boards available, but they do not have trimpots or anywhere near the same electronic complexity as whats been discussed in these pages, so presumably they are'nt suitable for printer application purely because of the lack of adjustability...even before all the other electronic failsafes.
Can anyone enlighten me on the various pros/and cons of such a board for this application?
[www.aliexpress.com]

I have not looked at these boards so can offer no information.

Quote

Also, to the electronics noob that I am, I see a functional similarity with instrumentation amps... binary "sound sensors" etc. Is this a valid assessment? If an amp with sufficiently high voltage tolerance were paired with a piezo disk that wasn't going to kill it (apparently 40V in the case below), would it work?
[www.aliexpress.com]

Piezo discs used in this manner are not really sound sensors but pressure sensors that only register changes. The difference is small but significant if you are trying to get the best sensitivity and noise rejection. Watch this and the other piezo thread for the next couple of months as there is work being done on making what is already the best in class probes and amplifiers even better.

Mike

Quote
BETLOG
Is it just me, or are the diagrams so small as to be illegible?

Here you go...

Quote
BETLOG
or buy a precision peizo (which after shipping seems *rather* expensive), or does a cheaper chinese version exist?

I make these boards so I have a financial interest in encouraging people to buy them, with that said...

When you buy a board from me, or more generally when you buy from an open source developer you're supporting innovation within the community, and you're getting more than just a PCB. DjDemonD and I have provided indepth support to every customer who needed it. This product has grown on these forums from an idea to a stable and popular solution to a difficult problem within reprap. With regard to the price, you're buying a piece of custom designed, hand made electronics, I don't believe the price is unreasonable. Paying for shipping is always a painful experience, especially internationally.

People who clone open source developments, and by extension people who buy cloned products stifle innovation. If it's not possible for someone to make a modest return on their investment of time and expertise then you limit development to people who can afford to give their time away for free, we're already seeing evidence of this, to the detriment of reprap.

I'll get off my soap box now...

Idris

---

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

We've no unhappy customers so far.

Here's a larger version of the schematic
[docs.wixstatic.com]

---

Simon Khoury

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

Quote
leadinglights
[www.ebay.co.uk] It is just a comparator with neither gain nor filtering but it works as long as there is not too much electrical or mechanical noise.

Thanks, thats what I was asking. So then it appears there are a number of alternatives available, with varying degrees of suitability and cost.
[www.aliexpress.com]
[www.aliexpress.com]

Quote
leadinglights
Piezo discs used in this manner are not really sound sensors but pressure sensors that only register changes. The difference is small but significant if you are trying to get the best sensitivity and noise rejection. Watch this and the other piezo thread for the next couple of months as there is work being done on making what is already the best in class probes and amplifiers even better.

Right. But if i understrand correctly the simplest form of the switch just tolerates a potentially large input differential, and at an adjustable threshold outputs a binary 0/1 signal.
So, ignoring the usual audio connotation, therefore isn't this what an amp is for?
To my relatively uneducated eye they seem to be very close to functionally the same.

Quote
Moriquendi
[attachment 98963 PiezoZ-probeSMD6.png]

Nice.

Quote
Moriquendi
I make these boards so I have a financial interest in encouraging people to buy them, with that said...

Yeah, I wasnt sure if nobody had already asked the same question due to it being a sensitive topic. But at the same time, this appears to be a circuit that probably already exists, so I had to ask.

Quote
DjDemonD
We've no unhappy customers so far.

Here's a larger version of the schematic
[docs.wixstatic.com]

Thanks.

If it were a $3 chinese module I'd just buy one of each and figure out which was more useful to me by experimentation...but 30 pounds is AU$50, so requires a few days research first
And I can design a lot more compliance into the mount in this case, so a less refined product may be more efficient for my application.

Edited 2 time(s). Last edit at 10/18/2017 09:51PM by BETLOG.