Piezo20 probe and piezo kit now available
-
If you have some PTFE rod spare, I'd try it if you're having success with it. Just found some nice chromed steel 4mm rods lying around at work.
-
I wonder whether you're over engineering this? The movement of the hot end is tiny, less than 0.5mm and probably less than 0.1mm. I would suggest that a compliant mechanism would allow the necessary movement with minimum weight while maximising lateral rigidity. I've got a pic but nowhere to host it, what's the best way round that?
Moriquendi
-
Email sent Idris I will host it for you and give a link in this thread.
Doug
-
Moriquendi's clamp
-
This is a rough representation of the lower part (Blue and Red) of DjDemonDs effector and would bolt to the upper part at the four corners with standoffs.
Moriquendi
-
How would it clamp the E3D Heatsink firmly?
-
It would need to be split into two parts like DjDemonDs, I just knocked it together to demonstrate that you don't need dedicated bearings and shafts, you can achieve the same thing by arranging for flex in specific areas. I've been using a similar mechanism to support my bed with piezo sensors for months with no issues.
Moriquendi
-
I am quite certain that the design for this can be simplified and improved, and it will be, its been a very engaging collaboration so far.
But this is where I'm coming from: my printer was never better in terms of precision than when my hotend was clamped rigidly in a metal clamp, in a metal effector (with magnet arms and metal carriages on linear rails) and had literally zero play whatsoever. Now I know I am not going to get to that level of precision as I am now using a printed effector and carriages, and my hotend is attached to a compliant mechanism. But I do want to limit that compliance to 0.15mm of vertical movement, enough to trigger the piezo, and no lateral movement at all. If this would work with the piezo clamped firmly between two flat parts I would go down that route but it doesn't bend enough to have high sensitivity in this configuration.
I have noticed the nozzle move laterally as it slides over a slightly squashed down first layer, I don't like that at all.
All I'm talking about are 4 short rods pushed through some holes, I wouldn't consider that over engineering it, I'm just pushing to see if I can achieve the aims above. I'll leave the wobblier version available for anyone who wants to settle for that, it works.
I used to be very much a "that's good enough" pragmatist, but since building this large kossel and striving for a lot more precision I'm very much bitten by the bug. I'm only developing this way of using the sensor, as I was not happy with the precision I got from the IR sensor, which is to be fair fairly precise.
BTW the easiest way to image share for this forum is https://postimage.io just upload image and you get all types of links, and they host in perpetuity.
-
I use Photobucket but it is the same Idea
I'll drop into my Plastics place tomorrow for the rods I will get a length of PTFE and a Length of Acetal (Which is a form of Delrin) total cost for 500mm of each is £0.51.
DJ Drop my a Mail with your Address and ill get some of each of to you tomorrow if I get chance (If I can find a PO Open)
Doug
-
BTW the easiest way to image share for this forum is https://postimage.io just upload image and you get all types of links, and they host in perpetuity.
Much obliged.
Moriquendi
-
DJ no prbs mate you have done much to help least I can do to send you some bits of Plastic
-
Thank you. I;d love to get printing this new version but I've a whole heap of parts to print for a customer. Maybe Ill print it on the microdelta.
-
DJ Rods are in the post mate u shud get them tomorrow
-
Thanks mate, I did slightly rework that top sensor holder for 5.5mm rods which I scavenged from work but they are too big the piezo wont fit. The version I sent to you was Lykles original top plate for his effector but I modded it with an extra recess to improve piezo sensitivity. Ive hosted my version here: https://1drv.ms/u/s!Apv79JfGbPIwgu1TAuVGDSUBP5ecuA you need to add holes for the new rods.
-
I will wait till you have tried it out I think you will find 2 pieces in the package one PTFE good to 260 degrees and acetal which is good to 100 the acetal is the stiffer of the 2but I think if you can make the max exposed part of the rods to be about 1-2mm both will be plenty stiff enough. The PTFE also absorbs moisture much better at 0.03% acetal can by 0.3 IIRC.
Let me/us know how you get on
-
DJ
This looks interesting wonder if same idea could be utilised with Piezo's
http://www.robotdigg.com/product/885/Auto-leveling-force-sensitive-resistor-sensor
A fitting that will fit on the nozzle temporarily whilst the Cal/bed mesh is done then remove and the Z Hight set manually would mean not doing it for each print tho?
Hope the rods arrived ok
Doug
-
Hi Doug,
That looks interesting, not sure how it handles a hot bed, and certainly won't handle a hot nozzle, plus the sensing area is a big pad on the end however as an idea a clip on piezo module for probing would certainly eliminate the nozzle wobble issue and the need to change effectors etc…
I will try it, its not that difficult to do is it? The only thing I can't immediately visualise is how it attaches. That being said if we were going to develop a clip-on sensor is piezo the way forward or might FSR be better? Or maybe even a very precise microswitch or optosensor with spring loaded trigger flag a bit like Nebbians deployable touch probe http://www.thingiverse.com/thing:1517363
Haven't got the rods yet but the post in Bristol is notoriously slow.
Lol no pressure but David mentioned that he's developing a PCB/glass/epoxy effector and carriages and has a piezo version in mind but is waiting on the prototype… better get on with it then
Should get them tomorrow I am sure.
-
I've had great success in the past with a tactile switch clipped to the tip of the nozzle, simple, cheap and surprisingly accurate. The disadvantage is convenience and eventually you're guaranteed to forget to take it off before heating the nozzle.
Moriquendi
-
I have been following the discussion with great interest and have tried a 27 mm disc with my E3D Lite hotend on my coreXY machine.
I soon discovered that the inner wire soldered position on the disc would be in the way of the top of the E3D so the contact would not be flat.
I tried to move the wire further out to the edge of the active sensor but the mechanical bond was very weak and after a short time the wire came off, bringing with it some of the coating. I tried another position and the same thing happened. I now have a piezo disc with very little coating so I am unable to solder a wire to the piezo surface.While the wire was attached I was able to get good readings by tapping the hotend, without the bowden tube fitted. Pleased with this I drilled a hole in the centre of the disc to clear the bowden connector but this left only a small amount of the piezo in contact with the E3D because of this I could not obtain consistent readings, in fact they seemed quite random.
So:-
1. how have you all resolved the fact that the wired soldered joint is in contact with the E3D? I assume you have the active piezo surface in contact with the hotend as opposed to the brass side.
2. How are you finding such a small area of piezo in contact with the E3D is sufficient to obtain good readings?Any help and advice would be great.
Paul -
So in my design there is a cut out for the solder pad and wire in the rim of the lower part (which clamps around the e3d groove mount). Place the piezo with the side with solder/wires/active element downwards towards the hot end and then sandwich it between the upper and lower parts. I get very normal triggering despite cutting a 5mm hole in the middle, but unless you have a better approach use a spur point wood drill on a wood surface and drill few seconds at a time with plenty of breaks to let heat dissipate.
The trick has been to tune the recess above the piezo and the contact area of the hot end clamp to allow just enough flex in the piezo to get the signal above the noise, but not so much that the nozzle is wobbly. If you set it up with the piezo flat between two surfaces its not sensitive enough, if you have it so that it can bend dramatically, i.e. a very narrow rim to the lower part, you get amazing sensitivity but wobbly nozzle. Its optimising it between the two.
Assuming the rods Doug sent me arrive today, I will see if the latest revision eliminates the residual nozzle wobble enough to consider it the Beta version.