[narrobb]

Been using this for few months, my take on it is at the end of the video. Just had some time to make a video about it to share.

https://youtu.be/_ZXgpXioe4E

[mactec54]

Been using this for few months, my take on it is at the end of the video. Just had some time to make a video about it to share.

https://youtu.be/_ZXgpXioe4E

Great Job and a good video, I would add a quality micro-switch and a sealed top is doable by almost any one that needs to do build a tool setter like this, it's easy to damage a Cutter with a had stop switch or a touch plate that a lot use, that is why a micro-switch or a hall sensor for similar switching would work better yours design is probably the cheapest way to make a tool setter like this

[mountaindew]

Very nice little tool and video!

[SCzEngrgGroup]

An actual switch mechanism with moving parts is not really necessary. Take a small piece of copper-clad PCB material, and mount it to a piece of metal or plastic, supported only by its edges. Solder a wire near one edge, and connect to a controller input with a pullup resistor. The input will be pulled low when the tool contacts the top of the PCB. The PCB is flexible enough to absorb any small over-travel. This setup will give excellent repeatibility, as there are NO moving parts whatsoever. Should you manage to damage it, simply put in a new piece of PCB material. Total cost under $5, with enough PCB material to last a decade.

Regards,
Ray L.

[narrobb]

An actual switch mechanism with moving parts is not really necessary. Take a small piece of copper-clad PCB material, and mount it to a piece of metal or plastic, supported only by its edges. Solder a wire near one edge, and connect to a controller input with a pullup resistor. The input will be pulled low when the tool contacts the top of the PCB. The PCB is flexible enough to absorb any small over-travel. This setup will give excellent repeatibility, as there are NO moving parts whatsoever. Should you manage to damage it, simply put in a new piece of PCB material. Total cost under $5, with enough PCB material to last a decade.

Regards,
Ray L.

To trigger there is no moving parts in this either. The spring is just a 0.5" travel protection for non-conductive materials but it will touch the inside screw after 0.3" travel, and this trips/stops Z travel.

So even if non-conductive material is used, once it goes down to the screw it will stop. This can be made to hit the screw after 0.1" or even less compression if needed.

For standard bits which are all conductive, the contact closure here in instantaneous and its repeatable to 0.0001. In fact, after 100's of usage, I can see a difference of 0.0001 only after restarting the machine.
And yes, it can't get any cheaper than this. :cheers:

[smokediver576]

Nice concept!!!! Are you looking to make an improved version to sell?

[narrobb]

Nice concept!!!! Are you looking to make an improved version to sell?

Thanks. But not sure if there is a big enough market for it.

Anyone with any CNC should be able to make this from scrap. Those struggling with such a small Copper plunger and no access to Lathe, can use Aluminum. And it does not even have to be round shaped, a rectangle, flat, bigger size piece with a 0.2" or so pocket at the bottom to secure the spring can be used. As the trigger is instance and the plunger should not even move, not even 1 thou, so the type/kind/size/shape of plunger should not be an issue. 10 IPM is plenty safe on end mills for first touch, then PathPilot i think makes it 0.2 IPM for the 2nd touch.

I wanted to highlight two main points really in the video. First point is the concept of joining Pin 4 & Pin 5 to make an Active Setter or even Active Touch Probe without any additional electronic components. The main design point is to separate the Pins until the bit touches the top surface. The end mill or bit once inserted into the spindle is now connected to the bed (conductive to the CNC bed). From there, any shape or design can be used.

The 2nd point is that Setter Height is not important, so there is no point of paying $$$ for a Setter that is within 0.0001 of 50mm or what ever published height. As if the Height is not accurate and you put 50mm while its 49.5mm then your tools are off by 0.5 mm (about 19 thou !!) plus any other errors. Putting Tool Setter Height to 0, makes accurate Setter height is no-factor at all.

Here is another idea for those who may have one of these cheap Chinese manual LED tool setter, remove the battery & LED, and just connect the spring to PIN 5, and base (from inside) to PIN 4. These tool setter rela on being exact in height, this is where it will be hard to be repeatable; not all of them are that accurate.

[Snecx]

Is there any undesirable consequence to use the machine to complete the circuit of pin 4/5?

[narrobb]

Is there any undesirable consequence to use the machine to complete the circuit of pin 4/5?

Not on my 770 S3. And I don't see why it would be different for other models. Pin 4 is DC ground. Pin 5 is the wanna be ground and it's designed to connect to DC ground to trigger the board.

Generally when both AC & DC buses are used on the same device; unless there could be a potential ground loop for micro-controllers, it won't hurt connecting DC ground to AC earth. This is also active probe means it only connects the lines just for those milliseconds required and no connection rest of the time. If you check my other video "Make a $25 Full 3D Probe" that uses a Passive Probe design which connects DC ground to the spindle body all the time.

If you want true isolation; a 3V button battery can be inside the unit easily. And this connects the button battery ground to spindle body and once that detected then the battery powered MOSFET or SSR can be used to join Pin 4 & Pin 5. And I believe this is the only proper way for full isolation. The other expensive Tool Setters on the market that do not have a battery and uses +5V/GND pins from the DIN connector to power the isolator chip won't be really isolated as they are still using the Spindle Body / DC GND as a reference. I don't think isolation is needed. It works like a charm as it is.
.

[jwatte]

To figure out whether DC ground is isolated from the machine body, bring out a multimeter and measure resistance between the two.

*If they are isolated*, then that's for safety; in the case some hot wire comes loose and touches the shell, it will provide a return path for the current, and if you have working GFCI, it will blow, else after a short while the fuse will blow.
Depending on the resistance in the path between the possible short, and the DC ground, this MAY cause a DC offset voltage that could be big enough to hurt the DC electronics in the system. There's really no way of knowing without careful study of the internals of the system.
In this case, as long as you don't have a stray high-voltage wire loose (or blown capacitor bending out, or fraying wire, or somesuch,) then it "should" be fine, and once the protection blows, it "should" be fine again. However, you will also want to measure the voltage potential between pin 4 and the chassis when the machine is on and idle, because sometimes electrical designs aren't as "natural" as you'd expect, for whatever reason. If there's a significant potential between pin 4 and chassis ground in idle-on state, it would not necessarily be safe to tie the two together at all.

*If they are electrically connected* with the multimeter, then there's zero additional risk in using it for this signal path.

[narrobb]

To figure out whether DC ground is isolated from the machine body, bring out a multimeter and measure resistance between the two.

*If they are isolated*, then that's for safety; in the case some hot wire comes loose and touches the shell, it will provide a return path for the current, and if you have working GFCI, it will blow, else after a short while the fuse will blow.
Depending on the resistance in the path between the possible short, and the DC ground, this MAY cause a DC offset voltage that could be big enough to hurt the DC electronics in the system. There's really no way of knowing without careful study of the internals of the system.
In this case, as long as you don't have a stray high-voltage wire loose (or blown capacitor bending out, or fraying wire, or somesuch,) then it "should" be fine, and once the protection blows, it "should" be fine again. However, you will also want to measure the voltage potential between pin 4 and the chassis when the machine is on and idle, because sometimes electrical designs aren't as "natural" as you'd expect, for whatever reason. If there's a significant potential between pin 4 and chassis ground in idle-on state, it would not necessarily be safe to tie the two together at all.

*If they are electrically connected* with the multimeter, then there's zero additional risk in using it for this signal path.

Back when I did the Passive Probe circuit I've checked the connection; I recall that the resistance was infinite, and the potential was less than 50mv.

[jwatte]

So, actual isolation, but at least not built dumb :-)

If you're willing to take the risk on some possible fraying wire or exploding component blowing out your DC power side (which honestly is not big) then this sounds like a good way to go.

[narrobb]

So, actual isolation, but at least not built dumb :-)

If you're willing to take the risk on some possible fraying wire or exploding component blowing out your DC power side (which honestly is not big) then this sounds like a good way to go.

Dont think there is any risk. Or at least extremely low. For any major issue, the AC side protection would have to be offline also.

Been using the custom Passive probe for over a year now. All good.

On the other hand, it’s is much better using it as Active type similar to this. Now I plan to convert the Poor’s man 3D probe to Active type this way no need to keep changing the Tormach settings

[hersh]

Thank you, looking forward to the weekend project.