CMM Touch Probe Possible?

  • Forgive my ignorance and lack of experience in electronics and programming in advance,

    In the Precision Machining and CNC Automation Course, I recently graduated from, our Haas VF-2 CNC Mill, a $100,000+ machine, was equipped with a touch probe. To us, this was so the ignorant, incompetent, blind machine could learn where the vice was fixed, where the metal stock is/was which is useful for jobs, or parts, with more complex geometries that would otherwise prevent the part from being machines correctly and/or accurately.

    What this did was effectively turn a big expensive dumb metal cutter into an extremely precise and repeatable CMM, with the ability to learn about its environment.

    The probe tool add-on was several thousand dollars, 2, 4, 6…depends on who you ask and what model and how much they think they can get out of you in that instant.

    These are the types of probes we used:–32926–32927

    This is how we measured tool lengths and calibrated the probe:–9397

    It was radio or optical, I do not remember the part number. That is irrelevant though, the point is the precision and the repeatability at that precision.

    Apart from the way it is connected to the controller…is this essentially just an endstop/limit switch that can change what axis it is controlled to adjust (via a macro and simple g-code) by selecting where to touch off the part or surface?

    After looking at what appears to be the first interesting patent:

    It honestly looks like relatively simple trigonometry and a simple physical construction.

    Another cutaway I found:

    And a popular application:
    (For the record, this is a cylinder head for an internal combustion engine in an automobile. A common application is probing the cylinder head in order to bore, or port, the heads as a volumetric increase or performance benefit. This is more advanced, however, opens up a LOT of possibilites for reverse engineering or quality control inspection.)

    Are these specific brands of sensors/probes out of the price range, yes, obviously.

    The concept, however, is not…

    After looking at this:

    and this:

    Is it realistic to believe that a list of data points in an XYZ (and others that are involved) space could simply be stored as a lightweight text file or a file could be converted or imported into a CAD program to generate a 3D mesh and a mesh into a 3D solid model? Even an STL would be accepted.

    Even if this were limited to more precise leveling, or used as a tool to aid in the ease of use as CNC machines like routers and mills and lathes grow in popularity, it would still be beneficial.

    Is the physical/electrical connection of a sensor like this complicated and/or difficult, or does one imagine that it would be? It does not look very different to the sensors supported currently. It looks like a relatively simple end stop or limit switch, even if it is an expensive one.

    Support for a CMM might be an interesting path given interesting probe designs such as these:

    It just seems like CMM's are easily within the realm of this board. so long as memory does not become an issue should a large mesh be the desired result.

    I want to build one of these, curious if I am missing something or if it really is as simple, not easy, as it appears to me as of right now.

    Thank you for your time, I believe this could be valuable to people here and at the lab. I want to squeeze as much as possible out of this board.

  • I am familiar with Renishaw touch probe systems, they are based not far from me and dental impressions I make are touch scanned on a Renishaw machine at my lab.

    However whilst it might have some application you are aware of the bewildering array of probe technologies now being used in 3D printing, our piezo systems (repeatable to 0.007mm) and David's/T3P3's strain gauge effector? All offering a level of accuracy more than enough for 3D printing.


Looks like your connection to Duet3D was lost, please wait while we try to reconnect.