Precision Piezo Z-probe Now available.

Has anyone considered using toothed idler pullies with the same pitch as the drive Belts? It may reduce mechanical noise as the belt teeth would fit in better than using smooth pullies which could cause sub or hypersonic vibrations.

The scope pictures are great but the piezo signal is clamped by the 2 input diodes to ca 0.7+ - Vdc by D1 and D2 to the input of the differentiator IC1a.
The server rejected the message because it thinks I'm a Bot? Me thinks I'm human.

I can't be sure on the human or not thing

I have used toothed idler pulleys in the past some Chinese company was selling them as drive pulleys but they had bearings. Seems they're quite common [www.ebay.co.uk]

If possible I twist belts over and run them smooth side to the idler, which works on my kossel Xl but not on my corexy. In fact I might try this on my belt driven z axis on my corexy it might get the underbed sensors back in the game.

Edited 2 time(s). Last edit at 08/26/2017 01:36PM by DjDemonD.

---

Simon Khoury

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

That scope you admired goes for usd 390. Nice to have but a new Printer?

Cool it's not the money its the space might invest in some sort of PC based scope.

Edited 1 time(s). Last edit at 08/26/2017 01:43PM by DjDemonD.

---

Simon Khoury

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

I'm waiting for the heat sinks to arrive for the TMC2100 watterott drivers to see if microstepping calms things down.

They're a bit of a handful my experience was they have a very slender window of good tune too low or too high and they skip steps, big heatsinks are definitely recommended. Apparently the waterrot genuine ones work better than the Chinese copies.

I've had great results with lv8729s as long as you don't want to run them above 1.5amps.

Edited 1 time(s). Last edit at 08/26/2017 02:04PM by DjDemonD.

---

Simon Khoury

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

The BW uses drv8825 driversso will try the genuine watterott and tune them to specs.

BTW was it you who created the adjustable underbed Holders? If so is it on Thingiverse.com? Also rewired the cable from the pcb with shielded and from the piezo as well.

[www.thingiverse.com]

Yeah I've not tried DRV8825s but heard nothing good about them.

LV8729's are cool, smooth and quiet (not as quiet as TMCs but quiet enough) same chip as TB6600's I think, but limited to 1.5A max. Will do high stepping rates my micro delta runs at 1/32 mcirostepping on these drivers.

Do you think the shielded cable is helping?

---

Simon Khoury

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

I tried turning my one belt over that was running tooth side to the idler (which is a 608zz bearing anyway so not small diameter) and it made no difference, my z axis just isnt smooth enough for underbed sensors, but hotend sensor works just fine on that machine. I have thought about going to a multiple leadscrew/ball screw system with independent motors, and true auto levelling using RRF/duet but my belt driven z axis works so well, it just isn't worth it.

---

Simon Khoury

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

I have ball bearing lead screw on mine and still had loads of trouble getting it working but it is an acrilic frame and that seems to amplify any sound the mechanics make

Yes, the shielded cabloes appear to work well, don't have a scope to do anything . This is eating into my beer budget. I've replaced the stock motor /gear train with a Tevo Titan clone with a Nema17 stepper and saved over 500g of weight on the gantry. Need to adjust the driver as motor got very hot in a 1hr run. The B W uses one motor to turn 2 lead screws with a belt. Don't have specs as to what max current istate for the nema17. I'm sorely tempted to get that rigol ds1054 it's 50 Mhz, 4 channel, excellent specs,very Impressed!

The BW is all metal. I can tap the table and it picks that up. I'm using one 20mm piezo in the center of bed hot glued to the support carriage. Sensitivity is not the issue in my opinion. For my application using more than one sensor is not practical, tho I have drilled and tapped 4 holes for that possibility. My bed has a foam insulation covered with aluminum foil which is a V4 bed. So heat is not an issue neither.

Quote
Chowa
Has anyone considered using toothed idler pullies with the same pitch as the drive Belts? It may reduce mechanical noise as the belt teeth would fit in better than using smooth pullies which could cause sub or hypersonic vibrations.

The scope pictures are great but the piezo signal is clamped by the 2 input diodes to ca 0.7+ - Vdc by D1 and D2 to the input of the differentiator IC1a.
The server rejected the message because it thinks I'm a Bot? Me thinks I'm human.

The point was just to figure out the difference and how it works for myself and I just wanted to share the pictures here. The pike is also reached after 70-80ms so using the top would be a delay we don't want either. 1.7 gram was the lightest weight I had and 1 cm drop seemed like the lowest I could drop it from consitantly. Don't you think this will scale down linear to? Seems like the area of the disc makes a huge impact on sensitivity. Not that it's needed tho since It detects if people walk upstairs almost before I adjusted VR1. I set it down at 200kOhm now from the recommended 400k.

Quote
DjDemonD
Cool it's not the money its the space might invest in some sort of PC based scope.

That's a great choice from saving space I thought it was just a nice to have item when I got it but I've used it on so much. I'd never sell it again once I'm used to have it. PC based versions can be bought really cheap to so if you debug eletronics a lot I would recommend looking into one. Even if it's the cheapest it's still better than nothing.

Quote
Chowa
Yes, the shielded cabloes appear to work well, don't have a scope to do anything . This is eating into my beer budget. I've replaced the stock motor /gear train with a Tevo Titan clone with a Nema17 stepper and saved over 500g of weight on the gantry. Need to adjust the driver as motor got very hot in a 1hr run. The B W uses one motor to turn 2 lead screws with a belt. Don't have specs as to what max current istate for the nema17. I'm sorely tempted to get that rigol ds1054 it's 50 Mhz, 4 channel, excellent specs,very Impressed!

google RIGLOL also and you want it even more haha. In short it gives you the software unlock so the scope becomes 100Mhz. Still can't measure 100Mhz signal tho just like you can't measure a 50Mhz with the stock one but still you get twice of what you pay for.

Just have to post back here after probing lots of times and doing many prints. First layer height is now always the same and I'm not even using baby steps to correct anymore this sensor is just bang on each time. No way back to the IR sensor after trying this! I'm designing my other printer to run with piezo probes also now since this is so stable to use. Great product guys!

Thanks for the feedback, they take a little bit of tuning in, but very reliable once tuned, do you mind if I use your feedback on the website?

---

Simon Khoury

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

No problem!

I think I will do a review on it also as this really deserves some attention. There are a minor downside with the clean nozzle but when I probe at 130degrees and clean out the tip its not an issue

Any feedback or review would be gratefully received. One of the hardest things I've found about developing this into a saleable product is getting my head around other peoples experiences. I've been working on this system for over a year now and I'd like to think I know it pretty well. The problem this raises is that when someone uses it for the first time and has issues I have to be able to put myself in their place to understand how to tackle the problem, this has been harder than I imagined.

We're both keen to see this become more well known, unfortunately marketing is not something I have any experience of though we do have ideas in that area.

Idris

---

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

Okay so I finally mounted a titan aero direct to my x carriage via the motor bolts. But where am I going to put the piezo sensor?

Between the motor and carriage of course!

See it in action here [youtu.be]

The part is on thingiverse, it works but I have not done much testing yet, but please feel free to try it. I suspect it (or a variant) might be useful in variety of other locations also.

[www.thingiverse.com]

Edited 1 time(s). Last edit at 09/10/2017 10:23AM by DjDemonD.

---

Simon Khoury

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

Since I use PETG, I always have a drop of PETG on the bed at home position, or a dip if I remove it (the bed is covered with BuildTack) and sticking PETG at the nozzle ip. No problem with my modified Z switch mounted on the gantry even If I didn't preheat.
How do you cope with this unless you clean everything every time ? I don't as it doesn't affect the printer in any way.

Edited 1 time(s). Last edit at 09/10/2017 10:48AM by MKSA.

---

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

Yes I think there are two schools of thought here:

1) "I clean my nozzle everytime" - in which case no problem.

2) "I rarely clean my nozzle" - okay so you can either, rely on the advice we issue to probe at 130 deg C, this ensures the soft plastic on the nozzle gets squashed on the first few probing points, so consider doing a G30 probe at the side of the bed to achieve this, then completing the rest of your probing/calibration. Or probably find that it doesn't work quite as intended, the z-level will be variable.

It's very accurate, you can get easily 0.01mm precision and repeatability standard deviation of 0.002mm.
Plus no offsets, it works on practically every surface you want, you can swap surfaces and re-level very rapidly.

IMO that's worth clipping the nozzle with a wirecutter, or implementing a nozzle cleaning station at the side of the bed, to automate the process.

As for blobs on the bed, no I have a basically clean bed (I just run a blade across the surface) before each print, I find if PETG hits a blob especially on the first layer it ruins the first layer.

Edited 1 time(s). Last edit at 09/10/2017 11:14AM by DjDemonD.

---

Simon Khoury

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

Quote
DjDemonD
Okay so I finally mounted a titan aero direct to my x carriage via the motor bolts. But where am I going to put the piezo sensor?
[attachment 97949 20170909_095046.jpg]
Between the motor and carriage of course!

See it in action here [youtu.be]

The part is on thingiverse, it works but I have not done much testing yet, but please feel free to try it. I suspect it (or a variant) might be useful in variety of other locations also.

[www.thingiverse.com]

Nice and simple, I like it.

Idris

---

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

I like that idea very much. What drive current are you using for the Nema 17 pancake motor? I found that using the default DRV8825 drivers that the motor was operating extremely hot. I have not as yet run this extruder with the TMC2100 watterrot genuine but can't find the drive current for the Tevo branded OEM Will continue further work with the underbed mount sensor.

Steve

Hi Steve,

Thanks, I wondered if it would work as it seemed too simple, but it does. Might see if it can be miniaturised using very small piezos, that way they could be inserted in a variety of locations depending on the machine and setup.

My motor is a 1.8deg nema 17 20mm 13 Ncm from OMCstepperonline.

I run it at 700mA it gets warm but I'd guess around 40-50 deg C max its not too hot to touch or even uncomfortable to put a finger on. I suspect the probing results would be less consistent if I probed straight after a long print, but I'm not likely to.

I've never really had a machine with DRV8825s but I hear nothing good about them.
I've tried TMC2100's I believe the Watterot ones work reasonably well, the chinese boards I tried (whitePCB with blue heatsinks) were very unstable, couldn't really get a current level that worked properly. LV8729's actually work really well, but are limited to 1.5A which is enough for most nema 17's, they are quiet, reliable and cheap.

My current machine with the titan aero runs of tmc2660's on a duet board.

Keep us posted on what you're doing and shout for help if you need it.

---

Simon Khoury

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

Hi Simon, I received my Piezo20, I got it installed and wired, I am just trying to figure out how to use auto leveling with smoothieware with it. It lists an error in the terminal: "Probe failed to complete, check the initial probe height".

I can home with it, and the nozzle is right against the PEI but I cant do a G32 with it. I am also used to Marlin raising the Z up for safety and have no idea why I can't do that with smoothieware. When I home, the nozzle stays on the build plate even when X and Y move which isn't very safe.

Attached is my config file for Azteeg x5 mini.

If you can help I would greatly appreciate it!

Thank you

Edited 2 time(s). Last edit at 09/11/2017 08:15AM by klcjr89.

Hi kjcjr89,

The only thing I can obviously see in your smoothie config (and I must admit I only played with smoothie for a few months, I am not really an expert)

Is that I have defined

# optional Z probe
zprobe.enable true # set to true to enable a zprobe
zprobe.probe_pin 1.29!^ # pin probe is attached to if NC remove the !
zprobe.slow_feedrate 5 # mm/sec probe feed rate
zprobe.debounce_ms 1 # set if noisy
zprobe.fast_feedrate 50 # move feedrate mm/sec
zprobe.probe_height 5 # how much above bed to start probe

I have ! left in as normally the unit works Normally Open.

That might be all it is. I was using a re-arm so same pins I think as for Azteeg.
I found a low debounce seemed to help.

---

Simon Khoury

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

Quote
DjDemonD
Hi kjcjr89,

The only thing I can obviously see in your smoothie config (and I must admit I only played with smoothie for a few months, I am not really an expert)

Is that I have defined

# optional Z probe
zprobe.enable true # set to true to enable a zprobe
zprobe.probe_pin 1.29!^ # pin probe is attached to if NC remove the !
zprobe.slow_feedrate 5 # mm/sec probe feed rate
zprobe.debounce_ms 1 # set if noisy
zprobe.fast_feedrate 50 # move feedrate mm/sec
zprobe.probe_height 5 # how much above bed to start probe

I have ! left in as normally the unit works Normally Open.

That might be all it is. I was using a re-arm so same pins I think as for Azteeg.
I found a low debounce seemed to help.

I found that with M114 the Z probe endstop would return 1 (triggered) if I have it set to: 1.29!^ So I took the exclamation mark out. I am confused coming from Marlin how to get everything to work. At least I got a print working but this isn't ideal.

Edited 1 time(s). Last edit at 09/11/2017 09:38AM by klcjr89.

Please check the probe isn't mistuned. The blue trigger led should be off, when the probe is open. It should flash blue when the probe is triggered. If it is the wrong way around go to tuning in the manual and follow that process. It's a NO probe unless the tuning is off.

Smoothie takes a but of getting used to. There maybe be some more expert smoothieware users with precision Piezo products who might chime in.....

Edited 1 time(s). Last edit at 09/11/2017 09:50AM by DjDemonD.

---

Simon Khoury

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

Should be tuned correctly; when I tap the hotend it the led turns blue then goes off quickly. I have a cartesian machine so I had to turn VR2 clockwise slightly to prevent false triggering.

Okay then right now I can't see the issue. Maybe we can post on smoothieware support and see if anyone has an idea. Basically it works like a NO endstop so provided it's not triggering falsely it should be working.

---

Simon Khoury

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