diameter readings when probing wrong

[sirbean]

What is the process of proving everything is dialed on the probe and software. I centered my probe using tormach software and it says it is perfect. I cant seem to get a edge zero from my probe that matches my edge finder. When I use your software to find a center in a bore the diameter it reports is bigger than what it actually is. Thanks

[DaveCVI]

Hi,
First things that come to mind to check:
1) Did you enter the probe tip diameter into the tool table for the probe tool #?
If not, the system probably thinks the probe tool has a 0 diameter...

2) Have you been through section 8.10 of the user manual re Probe calibration and accuracy?


Dave

[sirbean]

Yes on 1 and no on 2. I will look it over quick. Thanks

[DaveCVI]

The calibration process will make a small difference in the readings - but not a large difference as it only takes out cosine errors.
When you say the measurement is off, by how much? what is the known diameter and what is the measured diameter?

Dave

[sirbean]

only thing I can seem to get me closer is reducing diameter of tip about .005" in the tool table. Tip is a ruby tip from tormach with there passive probe. Documentation plus measurement both put tip at .118". Im getting random probing errors also where it claims it cant find x or y plane.???

[DaveCVI]

Hi,

only thing I can seem to get me closer is reducing diameter of tip about .005" in the tool table. Tip is a ruby tip from tormach with there passive probe. Documentation plus measurement both put tip at .118".

It sounds as of you have NOT done the MSM probe calibration steps yet - correct?
What you describe sounds to me like it's probably cosine error.
The user manual describes what this is - the effective tip diameter to get precise measurements is not the same as the physical tip diameter. This error comes from the angle of the probe shaft at the time the probe triggers (which is also a function of the orientation of the probe in the spindle and the spindle rotation angle vs the probe direction).

Doing the MSM probe calibration process will find the effective tip diameter and update the tool table with the effective tip diameter. The process does require that you have a bore (a good true circular one) that you know the diameter of. The process probes the bore, finds the measured diam, compares it to the stated known diameter and calculates an effective tip diam that compensates for the cosine error of the probe.
The measurement differences from cosine angle error are typically a small number of thousands - and that's right in the area of the .005 that you've determined by trail and error.

Im getting random probing errors also where it claims it cant find x or y plane.???

That is probably a noise problem on the probe input line and/or a de-bounce issue with the probe electrical signal. See section 8.11 in the user manual for a step by step process to check out probe operation - including checking for noise on the input signal, what to do about it etc.

Dave

[sirbean]

Ok I set the calibration per the manual. I still get the error. I don't get it if I set my second slow probe routine above 10. Right now I'm setting at 20 for my fast pass and 10 for my slow. What to you recommend for these settings? Thanks again. The errors were bothering me enough to take my probe apart and polish everything. No errors with tormach's software but they don't do the slow pass.

[DaveCVI]

I know it is counter intuitive, but raise the probe feed rate. A common tendency when starting out probing is to go as slow as possible out of fear of breaking something. Alas, I've found multiple times that probing too slow will get one frustrated. What happens is that contact bounce becomes a problem at slow speeds; the contact bounce results in multiple pulses into the mach probe input.

FYI, I typically run my probing operations at 40-50IPM and IF I do the slower 2nd probe I run it at about 15-20ipm.
From experimentation I've learned that my probe gets flakey below about 12IPM; so I never probe slower than that. This threshold will vary from probe to probe and also depends a bit on the amount of electrical noise in the system, de-bounce settings etc.

Now then, notice the bold "if" above?
Here is a (semi)secret: You don't need the 2nd slow probe; it's not of much use really.

I've written about this on the old support forum, but I've done enough probing tests to verify (at least to me) that the 2nd probe does not measurably increase the accuracy of the measurement. At least I've never been able to get data to show that it is worth the time it takes.
You're welcome to do your own tests - set up a fixed block, probe to it at different rates and record the trigger point. Do this multiple times for each probe feed rate so you get data to plot re probe repeatability vs feed rate. Do this a bunch more times at different rates for a single probe and again for an array of speeds for the 2nd probe. My test never showed the 2nd probe op as increasing the accuracy.
If you get different results re the value of the 2nd probe, I'd like to hear about it. So far in 3 years no one has offered up contradictory data.

That raises an obvious question: So why is there double probing in MSM if the extra probe doesn't help?
Answer: Because some folks believe double probing has to be better. While I don't have any data to support that belief, I'm not in the business of arguing about this with folks. Customers wanted that feature and they told me that MSM would not be competitive without it and they would not buy MSM if it were not there. So MSM got the double probing feature built in early in MSM development and it's had the the 2nd probe feature ever since....

If you want single probing only, set the slow probe feed rate = 0. That disables the 2nd probe operation.
Dave

[sirbean]

Great answer and thank you for the good explanation . I tested the measurent of a diameter and it was only .0002 off. I did a test of my the edge of my vise for squareness over 5" and it said i was .0011 out. So I put the old interapid indicator in there and sure enough i was out .001". A lot better results than what I expected and am so happy I decided to work on dialing the probe in. I initially had the fast probe set at 10 and my slow at 5. That is defiantly what was causing my errors. I'm currently at 30 and 10. Now that we talked I will do those tests you asked about. Great service!!! I am really liking the automation of the probe and am starting to have confidence in it. The next step on this automation journey will be a touch off plate. Your manual is really alot better than what I had assumed and I will be reading what you have to say about that. With my machine I only have one accessory port so it will have to share inputs with the probe. I would appreciate any and all direction you are feeling I should head in. Any articles or videos are a plus. Any way thank you for making me start to trust my probe readings

[DaveCVI]

Yep, slower is not always better when probing!

Re interfacing both a probe and touch plate - see section 6 of the user manual; it covers the topic including a sample interface circuit you can use to do the interfacing.

For videos see the CVI web site (http://www.calypsoventures.com) under the MachStdMill menu item - lots of videos covering MSM )both mill and turn).

It's nice to hear you're finding the manual of use - it was a bit of work and it does cover 99.9% of the topics I get asked about.

Dave