The Mostly Printed Probe is a side project to a resistance based probe interface board I was developing as this/last year’s Christmas project. Sadly my 21 year-old Renishaw ebay ‘bargain’ gave up the ghost and I needed a quick substitute. But the results were so impressive I’ve developed it into a full open source project.
Yes, there are dozens of probe designs out there but this one takes a quite different approach. By using pins instead of balls, a fully functioning probe can be made without any soldering and even the spring is made from filament… Accuracy is obtained through strategic use of tapers, bearing faces and concentric adjustment.
Hmm, seems I can’t show more than one picture… so here’s an exploded view:
If you search for ‘Mostly Printed Probe’, google might reward you with a link to a video short of me testing it in my mill and being slightly surprised by its repeatability - link to gihub is in the description.
All the models are available in that not-to-be-linked repository together with a print guide (the assembly guide is a work in progress).
That is a unique electrical circuit you have designed - every kinetic probe I’ve seen relies on the ‘three NC switches in series’ approach - probably because it also helps locate in X/Y as well as Z.
If you have a 3D printer, my design only requires 9off 3mm diameter x 12mm long dowel pins and a stylus (which can be made from an M3 bolt) and takes just a couple of hours to print.
Thank you. The dowels are set at 15° from vertical so 30° between them. That detail (and pretty much the whole project!) was inspired by this post on the LinuxCNC forum by Andy Pugh:
Looking at the diagrams from manufacturers of the ‘rod and ball’ type setting, the angle of the tangent between cylinder and spherical circumference appears to be similar.
Additionally, the resistance characteristics (ie the rise in resistance as the contacts open) is remarkably similar to my old Renishaw.
A few years ago I made one that used three metal radial rods and six pan-head screws (two for each rod). It seemed to slide easily sideways but it was much stiffer in Z, so I think the steep “V” would probably make it more balanced across X/Y/Z. I guess the length of the stylus would also play a part in that. I was just curious.
I was very impressed with the 3x3 rod design - primarily because I felt I could design it into an accurate 3D printable insulator (my original design was not wholly 3D printed).
The angles seem to be optimal for both location and electrical contact - both require very little spring pressure. The hardened steel (or tungsten carbide) pins probably help mitigate any risk of sticking.
Stylus length has a big impact on accuracy or ‘responsiveness’ - ie the longer the stylus, the further it will have to move (and perhaps bend a little) before the contacts start to open.
One of the very first things I was taught when I started my training in an architectural practice in the 1960’s was the use of the word “OFF”. Like most things, over the intervening time the strict use of the word changed, and perhaps my mentor wasn’t correct in the first place.
Bearing in mind that even at that time we were in the last days of drawing with nibbed ink pens on linen, and at the tail end of a time when water colours were used on key drawings to denote materials rather than written notes.
So here we go - if you were ordering something “bespoke” or made particularly, you’d have “6 copies run off” or a run of 6 copies. If you were ordering 6 screws however - you’d order “6 of those!”.
In engineering this was abbreviated to “6 off” to indicate that manufacturing was required, or “6 of” to indicate you could just buy it off the shelf.
I guess is that it was by the late seventies this had morphed into “6 off” to describe anything.
This may or may not be the real origin of the expression of course - but it’s been etched into my mind just waiting for a chance to pop out!
Thanks Peter - that kinda aligns with my experience as I would have acquired the habit in the early 80’s.
IIRC the missing space was to prevent a work order being changed when picking up supplies - ie eg ‘9 off’ could easily be modified to ‘95off’ whereas ‘9off’ removed the opportunity to get (quite a few) extras from the store/supplier.
I spent a little time working out how to get LibreOffice to do polar charts (they’re called ‘net’ charts in LO) and with help from the LinuxCNC forum ended up with this chart:
I am fascinated by the repeatability of a 3D printed probe - the concentric markings are 0.01mm apart. There are lots of factors at play (ruby ball size, backlash compensation, my competence) but still fascinating.
