Onefinity (Triquetra) 3 axis touch probe machined accuracy

Hi all, curious if everyone else’s measured X & Y values for there onefinity touch probes are fairly different? Ours are 2.1160” and 2.1250” respectively.

These should ideally be the same, otherwise probing from any other position than the bottom left requires swapping values in the probe window. This of course leaves room for errors to happen.

Wondering if anyone’s run into issues with this?

There seems to be much better options out there that claim very tight tolerances for similar or better price points.

Mine is different also. This is not a problem for me, I use the BB controller and it only probes to one corner. You could use your CNC to make it even then fix your offsets in the controller to match.

This was the first thing I thought of, but also its brand new and servs a very specific function. I really shouldn’t have to be modifying it out of the box. I know that in the vast majority of cases this isn’t an issue if you are just zeroing from the bottom left corner but in those instances that you aren’t this just leaves a lot of room for error.

Its also odd that there isn’t a tolerance spec listed anywhere on Onefinity’s listing or Triquetra’s website (the company that makes them). I reached out to them as well to see if this is just normal or an actual manufacturing issue. Will see if I get a response.

If this is just how they are it does make it a hard choice to recommend with other options out there at competitive price points listing tolerances of ±0.02mm (0.0007") that if accurate would certainly not have this problem.

Hey Kemperle-Industries,

the fact that the (Triquetra) 3-axis touch probe has wide tolerances is known.
You could use this one instead.

You could however try this:

I love my Elite, but the probe is the weakest link. Fortunately, it’s easily replaced. Here’s a video that shows the practical impact of the problem:

My ultimate recommendation is to just get a 3D touch probe like the one mentioned here:

I have a batch of them being shipped right now, and I’m taking pre-orders if you’d like to go that way. PRE-ORDER DIG3D.LP.M 3D Touch Probe – Ugly Dog Woodshop


Hey Aiph5u

Thanks for the links. Sad to hear its a known problem. I was looking at that probe and wish I had known about this before purchasing the Triquetra.

Ultimately I went with the 3D touch probe that @adamfenn28 mentioned but with a slightly different tip and cable length configuration. I actually came across it last night and thought this was a much better all around option.

Adam great video illustrating the problem, wish I had come across it sooner. I’m a bit surprised that Onefinity hasn’t found a better partner/supplier for what can be a rather important accessory to an otherwise great machine.

At some point when I have more time ill look into tightening up the tolerances on the Triquetra probe to see if I can eliminate the problem.

1 Like

I think all blocks of material used for probing are flawed by nature. Unless they are getting manufactured to Nasa standards and tolerances. I do like how you, Adam, touched on (pun intended) the use of a 3D probe in your “Zero Like a Pro” video. But I have decided on a more simple solution.

I bought an “el-cheapo” probe block from Amazon for all of $13 ( And with this, I no longer probe the X-Y axes using the outside edge of the probe block, but instead hold the block against the outside edge of the material! Then I configure the width and length of the probe in the Settings to be zero, and never worry about the X-Y dimensions of the probe.

I did have to hack the Buildbotics probing code, changing a few numbers, to accommodate this act of rebellion and force the spindle to travel extra distance on the X and Y axes. Without said hack, the process would stop prematurely thinking that it has traveled too far because the stock probing block is huge compared to mine (insert joke here). The default maximum amount of travel the probing routine is told to move for X-Y is only 20mm. I increased this to 40mm, along with reducing the amount of travel the spindle moves to “go around” the sides of the stock probe when switching from the X to Y axis.

Pros and Cons:
The only two downsides I have encountered for this insubordinate behavior include:

  1. I have to hold and move the probing block during the entirety of the process. This is really not that big of deal as I would think that most (if not all) of us will sit there and watch the probing process the whole time anyway. And, besides, I like to think of this evocative moment of time as more of a bonding experience with my machine and the probing process overall.

  2. I now have reduced how close to the physical limits of the machine’s 0/0 positions I can place my workpiece before being struck by the soft limit ban hammer of death pop-up warning. In other words, if I place my workpiece less than 40mm away from the machine’s X0 or Y0, the probing routine will see that my current position minus 40mm will result in a negative X or Y value and throw a fit. Whereas the stock probing block and code lets you get as closer from the edges of X0 or Y0 positions.

Aside from these downsides, I actually have the following upsides to this touch probe tomfoolery:

  1. Did I mention this probe is cheap? Not like free beer cheap, but the next upside compensates for that too.

  2. You can use ANY hunk of electro-capacitive metal, as long as it is relatively flat and doesn’t look like the surface of the Moon. It can be any size or thickness. It doesn’t matter how wide or long the probe block is because I don’t use the side of the block, but the actual outside edge of the workpiece.

  3. I don’t have to use the full, built-in, X-Y-Z probing routine either. At times, I have manually sent the separate G-Code probing commands via the MDI tab, then perform a manual move command for the radius of the bit, and then just click the “Zero” button on the web interface. This lets me zero an axis from the widest part of a workpiece that does not have a straight edge (I love you Harbor Freight chop saw), or if I simply just want to zero a single axis only.

1 Like

I think you’re probing the opposite side of the block than what we would with the Onefinity probe… that is, the side that is coplanar with the edge we are probing. You’re just holding it such that some portion sticks out past the corner so there’s something to probe against.

Is that right?

I believe your description is correct. Not sure where I can upload a video that is accessible, but here’s some stills showing, in order of appearance, the full Z, then X and finally Y probing routine and where I hold the probe for each: