Inexpensive touch probe option for 2d on a router??

[swissi]

slodat,

I have not tested it yet but in theory you could just add a line like this to your M6 macro file:

Code: Select all

If #4120 == 99 THEN M68 ELSE M88

This will activate Output 8 when tool 99 is selected and close it for any other tool number.

One word of caution when you convert your probe to wireless. We know that a VFD can mess with wired connections so you can imagine what it can do interfering with a wireless connection. The other critical issue by adding sender/receiver electronics is the introduction of latency. I could not find any information on the xoomspeed website what type of communication they are using but it doesn't look like they have much experience in low latency wireless communication. It's very easy to introduce about 100 milliseconds of signal latency with a wireless communication. With the default CNC12 slow probing feed-rate of 76.2mm/min your probe will travel about 0.127mm within 100 milliseconds. So if you measuring a slot, the latency over travel will be on both sides of the slot, doubling your measuring error of the slot width to 0.254mm.

If the latency would be consistent, you could adjust the probing tip diameter to account for the over travel but the problem with most wireless technologies is that the latency can vary a lot.

I suggest you are running some comparison measurements wired and wireless to get a good understanding if the wireless technology used in the xoomspeed box is acceptable to you.

-swissi

[slodat]

swissi- thank you for the code snippet. I'm not averse to programming in general (CS degree), but I'm lost in how the cnc12 programming works. I'll do some reading. I'm thinking I may want/need some error checking associated with this. As far as the wireless latency goes, I asked the maker of the wireless conversion. I'll see what he says. Perhaps the wireless isn't such a good idea? I don't know. I do know the VFD won't be generating an output when probing, spindle will be stopped.

[ScotY]

I have been using the Xoomspeed wireless conversion kit for maybe 6 months or so and have had no issues with latency. I’ve compared the probe results to a Haimer and have seen no differences. Fast probe at 20 ipm...can’t recall the second slow probe speed at the moment.

The Xoomspeed base station shows the probe battery voltage as well as RSSI signal so you have some idea of the connection strength. I’m not very smart so can’t explain it all very well, but it has been working great for me with zero issues.

I should add that I am using the Xoomspeed unit on a Tormach machine running Pathpilot. Sorry. I had planned to run it as is and eventually convert over to an Acorn. Still have the Acorn from my old little mill but just haven’t gotten around to it. Only Acorn machine right now is the cheap mini lathe.

[swissi]

I suggest you run the same measurements hardwired and then wireless. I'm almost certain that if you are not using a very, very slow feed rate that takes the latency out of the equation that the wireless measurements will be larger for inside measurements like a slot and smaller for outside measurements like a block width. Also the repeatability delta between measurements will be smaller with hardwired connection compared to wireless.

slodat,
the information you need for macro programming is currently spread over several chapters in multiple documents. I know that Keith is working on a Macro programming manual that will be released soon. The most important information is in chapter 12 and 13 of the Mill Operators Manual that you can find in the cncm\manual folder of your CNC12 Install. I suggest you read the manual from start to finish all the way through.

If you want more checks in your macro, you could do this:

Code: Select all

If #4120 == 99 THEN M68 ELSE M88 ;activate OUTPUT8 when tool #99 is used
G4 P.5 ;dwell for half a second to allow CNC12 to activate the output and close the PROBEDETECT input
If #4120 == 99 && !50006 THEN M200 "#)WARNING!\n\nProbe in Spindle but PROBEDETECT Input not Active!\n\nPress CYCLE START to continue or RESET to cancel"

This example assumes that Output 8 on the Acorn is configured as OUTPUT8 and wired in the way that when activated will close Input 6 that is configured in CNC12 as PROBEDETECT. The 3rd line will check if the PROBEDETECT input is indeed closed when tool 99 is used and will prompt you with a warning message if not.

-swissi

[slodat]

Response from the maker of the XoomSpeed wireless conversion, posted with his permission at my request.

Re the radio. You can forget about radio interference. The radio module used is an RFM96 device. It uses LoRA spread spectrum techniques and is pretty much the gold standard for this type of application. If I wanted, I could arrange for it to have a range of a few miles.

You are correct that variation in latency, rather than latency itself, is the thing that really matters. Expect the variation to be well down in the sub microsecond range. If your machine had a repeatability of 0.0001" and you did the slow probe at 1"/min, this variation would be 10000x better than your machine.

What is important is that you should use a 2 stage probe. One at high(ish) speed - say 50 inch/min, then retract 30 to 40 thou before doing the slow probe at 1 inch/min. The probe has a 'half asleep' mode that it uses to save power when it's not being used. This mode will give noticeable latency, but as soon as the first touch occurs, the probe goes to 'fully awake' so the 2nd touch delivers the precision you need.

The time the probe stays in 'fully awake' mode is programmable and the timer that ends it and returns to 'half asleep' retriggers if there are any further touches.

Latency does affect the apparent probe tip diameter in exactly the same way as stylus flex. The best way to calibrate this is to get your hands on a gauge ring and probe its diameter at a variety of orientations. Then calculate the probe tip diameter that gives the best measurements of your gauge ring.

Ask someone with Mach 3 to show you the probe calibration routines in the 'ProbeIT' add-on.

Again, expect the errors caused by variation in latency to be at the very least 1000x better than the variation you get in stylus flex caused by the probe's 3 point suspension design as you probe in different directions.

Thank you all for contributing to the conversation. I really appreciate it.

[slodat]

I suggest you run the same measurements hardwired and then wireless. I'm almost certain that if you are not using a very, very slow feed rate that takes the latency out of the equation that the wireless measurements will be larger for inside measurements like a slot and smaller for outside measurements like a block width. Also the repeatability delta between measurements will be smaller with hardwired connection compared to wireless.

slodat,
the information you need for macro programming is currently spread over several chapters in multiple documents. I know that Keith is working on a Macro programming manual that will be released soon. The most important information is in chapter 12 and 13 of the Mill Operators Manual that you can find in the cncm\manual folder of your CNC12 Install. I suggest you read the manual from start to finish all the way through.

If you want more checks in your macro, you could do this:

Code: Select all

If #4120 == 99 THEN M68 ELSE M88 ;activate OUTPUT8 when tool #99 is used
G4 P.5 ;dwell for half a second to allow CNC12 to activate the output and close the PROBEDETECT input
If #4120 == 99 && !50006 THEN M200 "#)WARNING!\n\nProbe in Spindle but PROBEDETECT Input not Active!\n\nPress CYCLE START to continue or RESET to cancel"

This example assumes that Output 8 on the Acorn is configured as OUTPUT8 and wired in the way that when activated will close Input 6 that is configured in CNC12 as PROBEDETECT. The 3rd line will check if the PROBEDETECT input is indeed closed when tool 99 is used and will prompt you with a warning message if not.

-swissi

swissi,

I got a great deal on a used Mitutoyo 1" master ring setting gage. I will use it for my probe setup. I will do some wired and wireless measurements and post my results. Thank you for the code snippet! I'll do some more reading.

[FlySox]

I'm happy with my ITTP probe, but just bought myself a Christmas present to see if this Belarusian wireless probe (~$180 shipped) is any good:

https://vers.by/en/blog/user-s-guides/vers-wl

Will post my findings in the new year...

[slodat]

I got the wireless conversion installed in my ITTP last night. Will get it installed on my router soon.

[carbuthn]

FlySoc, thanks for the information on the Vers.by touch probe and tool setter, looking forward to what you think of them. Much more reasonable price for the hobbyist if they work well.
Chuck

[slodat]

Connected the XoomSpeed wireless receiver to my router today. Works really well. Even from across the room.

Small box on the spindle allows for a wired probe, should I want to use one. The real reason for the box is the ProbeDetect pushbutton for use with the wireless probe. Works great. Adding the additional probe connection was too easy to not do it. Not sure it will get used.









First order of business was getting concentricity failed in. I got it to a few tenths. As good as it got today.



Then run a bore probing cycle on a known bore size and see how things look.







I am really happy with the probe and the wireless conversion. I'm excited to add this to the control. Lots to learn!