Home made 3d Digitizing Probe

[Krinilssen (Kristian Nilssen)]

Howdy! I am designing a 3D Digitizing Probe to be used with my hopefully soon to be built lowrider2. I did some looking around for a cheep 3D digitizing probe but had no luck finding a good cheep option. After seeing some people’s attempts to build their own DIY Digitizing Probe I thought to myself, why not give it a shot while you are waiting. I looked at a few of those attempts and tried to make something of the sort of best of both worlds and this is what I quickly came up with! What I still need to finish is the top circuit board and a port for a wired connection seen started in the top left assembled markup and possibly some LEDs to give visual feedback while in action. Let me know what you think about the design and give some feedback what changes I should make! If you have any pointers, recommendations, or lessons learned if you have tried making one of these please let me know as well!!!

Little info:

Minus the rods that sit on ball bearings, circuit board, probe tip, wiring, screws, and springs, everything else is going to be 3D printed.

How this works is that a current is fed into one end of the circuit containing the ball bearings. The three metal rods coming out of the center hub sits on top of these duel ball bearings and completes the circuit. When the probe comes into contact with an object the rods or rod will lift and thus not be in contact with the ball bearings it sits on, breaking the circuit.

I plan on using three screws from the bottom that push on three small plates underneath the three sets of dual ball bearings so I can fine tune the accuracy of the probe.

Three springs which surround the three structural screws that keep the device together are in constant contact with the top of the ball bearings structure keeping everything in place and pushing the object down towards the leveling screws I just mentioned. This should cheaply allow for some decent fine tuning (hopefully, that’s the plan at least ).

I haven’t drawn it yet but I’m thinking of placing on the top circuit board a circuit that inverts the signal so that instead of constantly feeding 5V to what ever is driving this, that it only sends a signal when contact is made.

I have read that after quite some time, having power run through the circuit and doing tons of contacts, that as the metal probes lift off the ball bearings, it may arc slightly causing an oxidizing layer that can effect the performance of this. Ive also read that when this happens, all you need to do is brush off that oxide layer and it will work just like new again? Not too sure about this but if anyone could give me some info on this I would greatly appreciate that!

I haven’t decided on how im planning on mounting this to the cnc but more than likely I will incorporate some type of easy tool changer so I can just place this inside the bit and not have to find a way to mount and find the offset and everything else involved with that.

Any ways, hope someone found this interesting or has some great feedback on this design for a 3D Digitizing Probe that I’m planning of incorporating into my cnc!

[timonjkl (Tim W)]

That would be super hope you can get the hard part done so we can all use it try and make the board so we can cnc it too double bonus

[jamiek (Jamie)]

Your design should work. I wouldn’t worry about the contacts creating arcs. That is only really an issue if you have high current, and if you are using the normal endstop method where a pin has a weak pullup and is shorted to ground, then its not an issue. (Unless you’re in an explosive atmosphere, then you might worry.)

I dont know why there aren’t cheap chinese “hobbyist” probes available. It seems it should be possible to throw one together for a couple bucks in volume and sell for maybe $20. But the cheapest one I found for sale was way too expensive, which is why I made one myself.

[Krinilssen (Kristian Nilssen)]

That’s a relief, one less thing to worry about! Let’s hope that I never will have to use it in an explosive atmosphere then haha!

Thank you for giving this a look and giving me some advice so quickly! I’m pretty new to CNC and the V1 forums but I have seen so many awesome people who really care and are super interested in creating a great community around such an amazing product!

[Krinilssen (Kristian Nilssen)]

Sorry it has been a while. Somethings have come up and the times we are in right now haha. I was sitting down to finish my files after printing and assembling two prototypes and was about to work on the pcb. I am new to Cnc ing and was curious, what is the standard for 3D digitizing probes, is it for them to be normally open (constant power and a drop to 0 for a “Touch”) or normally close (where it is 0 and current for a “touch”)? Thank you and I’m almost ready to release all the files!

[jamiek (Jamie)]

It is simple enough to configure the firmware to accommodate NO or NC. NC is recommended for endstops for safety but I don’t think that applies here.

If the probe and object are both conductive then it’s simple to send a current between them, which would be a normally open configuration. For the probe in your drawing I think NC is simpler since you can just connect through all three contacts in series. I built a probe of this style in the NC configuration and it worked pretty well.

[Krinilssen (Kristian Nilssen)]

Jamie, I looked up your probe and its awesome! I really like it good work!

I have done some tinkering and I think I have come up with an alpha/prototype/version 1! I am only a little worried about the circuit design. I will post my circuit design and the calculations I used to get the values I am using in the next post with all the files and info for the version one/alpha/prototype. The only thing stopping me is that Im not too sure on how to go from here. The things that are stopping me are:

1. What is the connection? I have not built my LowRider yet so im not too sure what to follow. I see on the new board in the shop, the SKR PRO V1.2 it looks as though the connection for the probe would be to one of the endstops position and would be using 3.3V? most of the documentation I have found show that endstop connection as 3.3v. My design accounts for 5v so im a bit confused there. Is this the source for the pull up? If so can I just source power from 5v rail on the board and would that work for other controller boards as well?

2. What does this compare to with the other boards in the shop? I want to make this probe so it can work with most cnc’s and all someone has to do is just download, print, and assemble the parts.

3. The dimensions, should I make them smaller or would this be suitable, I have read some people have problems with 3d touch probes due to a small z axis or the probe not allowing the machine to reach the full horizontal axis lengths. This is just the size I was able to make it for a first version/prototype/alpha and I do think with some work I can slim in down in both its height and its width.

[Krinilssen (Kristian Nilssen)]

Prototype/v1/Alpha

This is a long time coming, it took a while but here we are! I have already started working on v2/beta by reducing the height of the probe to about 52mm but I do think I can go a bit shorter if people think that it should be. I have ordered some custom pcbs for cheep to try out my old circuit design which you can see sitting in the picture of the probe I attached (Pictures 4-7). This is not what I am going to be using going forward, and the new circuit I have designed takes into account the current over the probe to negate oxidation on the contacts between the balls and the pins but still have enough current to make the connection reliable. After some research I ended up with about 5ma (Let me know if this is not right). I have added two LEDS, one that is green that displays power and is constant on (Pictures 2-3). The other is a red LED and it is triggered by contact with an object visually signaling a touch (Picture 3). The LEDS I used are Adafruit sequins and have a 100 ohm resistor baked in to them so that had to be taken into account when designing the circuit. If anyone finds and mistakes or has any improvements they think would benefit the circuit design please don’t feel afraid to chime in . All 3d printed parts are printed in Prusament PETG on a mk3s at 0.2 layer height and 20-40% infil. Since this is a alpha/prototype/v1 i’m not going to include the STL’s but if you want them please send me a message or comment that you want them and ill send them over. I have a 50mm ruby tipped stylus that is arriving any day now that I am excited to attach and work with!

Circuit Design

(Picture 1)

Testing Circuit on probe

(Picture 2: Probe is closed and not triggered)

(Picture 3: Probe is open and is triggered)

First Prototype

(Picture 4)

Models of prototype

These are models of the prototype that I am using to create my STLs and to design around. These do have some v2 improvements added to them (I apologize). The changes in these models from the prototype are the reduction in height, change in the ball carriers and pin carriers, and changes in the bottom plate. In picture 7 you can see three empty holes on the underside. These holes are for small m2 or m3 screws (haven’t decided yet). The screws are to make fine adjustments to the ball carrier plate so that one can truly center the probe if for some reason it is not already centered or the probe tip you are using is off.

(Picture 5)

(Picture 6)

(Picture 7)

[Krinilssen (Kristian Nilssen)]

Went ahead and designed a PCB for the updated circuit. I hope someone who knows about circuit design and/or electrical engineering could chime in and analyse it! Would love to get a small batch of these ordered soon from a quick cheep custom PCB shop so I can test them out. Im still debating on wether soldering the balls together via wire or making a PCB with copper pads that the balls sit on would be the better route. Currently they are soldered together but I went and designed a PCB for that as well. Would this just be one more complication or would it be better and easier than soldering the balls?

-Thanks!

Pictures:

(Picture 1: Main PCB with circuit)

(Picture 2: Lower PCB with copper pads)

[jeffeb3]

Looks fine to me. I think you can definitely make is simpler and smaller and drop some components. But if you want to do it this way, this looks like a good way to do it.

[jamiek (Jamie)]

Adding the LEDs definitely adds complexity compared to a ‘dumb’ switch, but I can see it being a nice feature. Although if it makes contact only briefly it might be hard to see.

From the schematic it looks like you are driving with 5V power with a 3.3 V signal. When the switch is closed and the transistor is off, your LED 1 is going to put 5V (minus some for the diode) onto the signal line, which the controller might not tolerate. I think you might need two transistors to keep the LED and the signal separated.

The other thing which is probably a non-issue is when a 3.5mm jack is plugged or unplugged, I’m not sure whether it makes brief partial contact with the wrong pins. It might be worth checking that it won’t blow up anything having the polarity reversed for a moment or that it won’t short +5 onto to the 3.3 V line.

[Krinilssen (Kristian Nilssen)]

Thanks @jeffeb3! I am currently working on making it a bit simpler and smaller but I am also working on a version that is just the bare minimum without any cool features that should be a lot small and a ton simpler.

@jamiek, thanks for the insight! I was going off of a lot of other circuits people have designed (Like a few you linked to in your thingaverse post as inspiration for your own probe) and silly me just didn’t put one and two together. The unfortunate thing is that the forward voltage on the green LED is 3.4v so I don’t think a 3.3v supply would work for powering that. As a work around do you think that by simply putting a diode between LED 1 and the 3.3v probe signal would do the trick? That would still allow the prove signal to be pulled low when activated and when the probe is not activated the diode would stop the 5v power from flowing through LED 1 and through the signal line to the controller?
When looking into a connector I just chose a 3.5mm audio jack because I had a few extra laying around and i’ve read of a lot of projects using them. Do you have a recommendation for a different easy to get ahold of and easy to find cables for connection system that would work better?

Thanks for the feedback guys I really appreciate it!

[Krinilssen (Kristian Nilssen)]

I was thinking of something like this to incorporate the diode to fix the problem that Jamie pointed out

I also looked into mini xlr plugs for the connection. They seem to be easy to get and they would fix the possible problem of a wrongful contact connection with a 3.5mm jack.

[jamiek (Jamie)]

I think it should work, although the diode drop on a typical silicon diode is about 0.7 volts which is not much below the maximum off voltage of 3.3v circuitry (0.8 volts). It will probably work okay, but another way to do it is to short to ground and switch the LED with a combination of PNP and NPN, like this:

When the switch opens, the NPN Q1 turns on, which then turns the PNP Q2 on, which lights the LED. Note that putting the resistor R1 on the emitter instead of the base is deliberate, so most of the current through R1 will come from Q2 and the 5V supply, which should put a very low current load on the pull-up resistor of the signal pin. If the resistor were on the base and the emitter were connected directly to GND, you would need to balance the consumption of Q1 against what the pull-up can supply while keeping the logic level high.

This would invert the logic relative to your circuit, i.e. open switch is logic high instead of open switch being logic low, so the firmware would need to be expecting that.

[xorrox (Adrian Calota)]

Hi would you be so kind and send me the files you have. I want to build such a thing and I love your design. Adrian@calota.org

[timonjkl (Tim W)]

Any progress on the design?

[Tokoloshe (Philipp)]

I am going to revive this thread, because I found some cool ideas:

https://martinriches.de/cncmore.html#scan

(First found here: DIY 3D CNC touch probe.)

I think I am trying to build the last one, but connected to my spindle ground.