Decided to see if it is possible to use piezo sensor in the heatsink clamp to do z probing. The answer is yes, and its very sensitive. My delta autocalibration deviation has gone from 0.05 to 0.01, with all the benefits of using the nozzle as the probe but with a firmly mounted bed, and a firm nozzle too.
Last edited by DjDemonD (29 January 2017 00:31)
Very interesting. I'm looking at possible alternatives to DCs excellent IR probe which seems to be adversely affected by 3dLac that I've started using. I must investigate further..........
DJ am I right in saying that it uses just one piezo sensor ? and how do you resist the E3d from moving in the mount?
Maplins sell this one http://www.maplin.co.uk/p/3v-ceramic-pi … 2742-qy13p
Last edited by Dougal1957 (13 January 2017 12:46)
Yeah that maplin one would do the trick, the 27mm one, you need a small signal processor board to get a useful trigger signal from a piezo transducer, but Moriquendi on reprap has made some and still has a few available.
The hotend is solid as the upper and lower parts can be fairly tightly screwed together. The design is such that any force axially up through the hot end causes a tiny flex in the piezo and you get your trigger. Is it as rigid as the hotend fastened into a metal effector with a metal groove mount adaptor - no, but its fairly firm.
The force required to trigger it is tiny, it barely tickles the bed as it probes. Much less violent than FSR's with none of the wobbly bed that goes with it.
Last edited by DjDemonD (13 January 2017 16:08)
Did i read that right, your drilled the centre of the piezo sensor disk?
Yes I drilled out a 4mm hole in the centre of the Piezo transducer.
Piezo bimorph plate is quite tough, we used to cut circular discs out of them with a hole saw when I was building lasers. Don't let the piezo get hot when cutting or drilling it, because too much heat will destroy the piezoelectric properties.
Duet WiFi hardware designer and firmware engineerhttp://www.escher3d.comhttps://miscsolutions.wordpress.com
Yeah I drilled two, one with a regular twist drill on a piece of wood that one worked, the other with a diamond core drill, not so successful too much heat.
It is uncanny how precise the probing is though and how gently it does it.
I wonder if its possible to get plates made to a custom shape.
Duet Wifi Hardware Designerwww.duet3d.comwww.think3dprint3d.com
It would be interesting if a smaller unit with hole already drilled could be made. The 27mm unit is larger than ideal packaging-wise, but works by bending so it might be hard to get the same sensitivity from a smaller disc, although it could be made thinner.
Speaking of this, there is already "donut" hollow type of sensor.
https://www.kistler.com/gr/en/products/ … lbf_9135_b
While probably expensive they might be a good solution
what would be awesome would be the ability to get them with mounting holes to screw into the top of the hotend (3x3mm m3 for example) and another set of mounting holes to screw into an effector or carriage. does anyone know how they are made?
Found a 20mm Dia one at rs (they also have a 12 mm version)
http://uk.rs-online.com/web/p/piezo-buz … s/7214940/
Interesting to see how well the buzzer works as a sensor, do they manufacture piezo's with different properties ans actuators vs sensors?
These are the Murata ones from the same range that was in the original post so should be ok But I think I will get some of the 27mm ones and have a go with this.
DJ did you go back to the std E3D Cooler with the 30 mm fan?
I think a slightly smaller one would work but perhaps not a 12mm, once you've cut away 4mm from the centre you will have not much piezo-electric material left to generate a trigger, and it won't bend as much. I don't think there's any feasible version without cutting a hole in the Piezo.
I am envisioning a unit with two pieces, one piece with the hot end on, one for mounting to the carriage/effector which slide over one another with perhaps two small bolts to connect them,with the Piezo sandwiched between. For deltas I'd build this into the effector, for cartesians/corexy the best design would be to incorporate this into the x carriage.
If anyone wants to pick the ball up and see how far they can run with it I'm behind them all the way. It's a promising sensor on account of it being very sensitive and nozzle based plus the Piezo disc can be fairly tightly compressed reducing the compliance of the whole assembly getting rid of the wobbly nozzle issue you get with fsr's, microswitches etc..
Last edited by DjDemonD (15 January 2017 10:25)
Doug, I'm using a 30mm fan e3d cooler with adjustable dual 30mm part fans that I co-designed with Jinx from reprap. http://www.thingiverse.com/thing:1692742
My understanding is that they are essentially the same thing. Some produce more voltage, a sharp tap with a tool can get 90v out of one of these things according to Mike (leadinglights) but as I have mine setup on the scope it was producing around 1.5v with a light tap.
Have ordered a batch of discs and the board.
The issue with the 30mm fans is the noise they produce wonder if a redesign of the top of one of DC's 40mm Fan mounts may be in order and attach to the effector by all 6 screws in a radial ring arrangement (Just thinking out loud)
Sounds like a good idea. I find the sunon maglev 30mm fans are quiet, reliable and powerful you can get them from Farnell for around £13 each.
Last edited by DjDemonD (15 January 2017 12:19)
Here's another possible configuration:
1. Drill and tap the top end of an E3D heatsink with 3 small holes, probably M2.5 or possible M3. I hate groove mounts.
2. Take a piezo element with the active element about 20mm diameter. Perhaps this one http://www.digikey.co.uk/product-detail … ND/4358154. Drill a central hole for the Bowden tube and collet, and 3 smaller holes around it.
3. Attach the disk to the top of heatsink using 3 nylon screws, with a piece of insulation between the piezo and the heatsink. The purpose of the insulation is so that if the heater cartridge develops a short and makes the heatsink live, we don't get a short between the heater and the piezo.
4. Clamp the brass surround of the piezo disc between the effector (which has a hole in it for the top of the heatsink to pass through) and a printed ring.
For the electronics, I would take one of my IR sensors, remove the optical components, and feed the piezo into the analog input. Then program it to produce a nice clean pulse when it detects a shock. Piezos produce a good voltage from a very small mechanical shock if the load resistance is very high.
I'd have a go at this myself, but I don't have a drill press for drilling the top of the E3D heatsink.
That's a good approach and neatens the design. As long as the area of contact between the heatsink and the piezo is fairly small it will flex enough. Maybe a small washer with 4mm ID and 6-8mm OD between the heatsink and the piezo transducer. Good idea with the ir board it seems logical it would work. The board Moriquendi came up with is probably very similar to it but has two pots to adjust the sensitivity to suit various piezo discs. But as a finished product using a known and tested piezo this would be unnecessary. I'm fairly sure these boards can only be supplied in fairly small numbers at present.
My only concern would be the whole hotend is suspended from the piezo unless I'm not visualising what you mean correctly. This might reintroduce the wobbly nozzle issue. My rough design doesn't look nice but it does work and the nozzle is quite firm I think I'll work on making this smaller and neater for now.
Agreed re. groove mounts they are useless. The 90 deg tee I based mine on clamps the hot end hard and effectively turns a groove mount into a flat mount with m4 bolts.
Last edited by DjDemonD (15 January 2017 13:51)
That David sounds like a plan although don't quite get how you would fasten the Hotend to the Effector with this arrangement any chance of a quick Sketch.
I am sure I could drill and tap 3 holes in the top of a heatsink at M3 and maybe use a piece of Bare PCB G10 board as the insulator.
That David sounds like a plan although don't quite get how you would fasten the Hotend to the Effector with this arrangement any chance of a quick Sketch.I am sure I could drill and tap 3 holes in the top of a heatsink at M3 and maybe use a piece of Bare PCB G10 board as the insulator.Doug
See https://dl.dropboxusercontent.com/u/193 … can029.pdf (sorry about the poor drawing, I flunked Art at school).
A is the top of the hot end heatsinkB is the insulatorC is the piezo transducer. The thicker part is where the element is, the thinner part is where the brass substrate extends beyond the element all round.D is the effectorE is the printed ring. It will need a small tunnel in it to make room for the wires attached to the piezo element.F are screws. The two (actually 3) inner ones at least are Nylon to avoid shorting the piezo element out or connecting the piezo to the hot end metalwork.
If the 27mm O.D element I found proves too insensitive, we could use a 35mm one instead.
PS - RS also sells a 27mm OD piezo element, see http://uk.rs-online.com/web/p/piezo-buz … s/7243162/. They also sell M2.6x6 nylon screws.
OK I get that the bit that worries me is if 3 Nylon screws will be substantial enough to hold the hotend especially should it contact anything.
How about using Mica washers as in the type used to attach 7405 type things to heatsinks and use std Stainless screws would need to drill the Piezo disc a little larger tho and I wonder if there is enough space to get it all in.