Calibration issues

[redvan22]

After trying to use the new edge finder function using a manual edge finder and getting weird results, I decided to recalibrate my mill (I thought new software - recalibrate). This became a daunting task. The new process of going to the axis settings and clicking the calibrate button kept giving me a message stating that the final number was too small to be used and to to try again. So I clicked the "try again" button and followed the instructions only to get the same message over and over again. So, I resorted to the old way of calibration but that didn't work either.

Now, for the Calibrate button experience; first I set the CPI to 10000 then clicked the Calibrate button which asked me for the size of my calibration block - 3" (using a 1-2-3 block). It then instructed me to locate the left edge and then click Continue, which I did and I had to wait for a progress bar to complete while DRO did some calculations. Next I was instructed to locate the right edge and then click Continue, which I did and I had to wait again for a progress bar to complete while DRO did some calculations and then I got the message stating that the number was too small to be used and to try again which I did several times but kept getting the same message each time.

So, I reverted to the old calibration method; I set the CPI to 30000 and then changed the readout to 4 positions then indicated the left and zeroed out the display. Moved to the right side and indicated it. Took the reading on the display and entered that as my CPI and when I returned to the display, it was way off.

Tried entering the standard of 5080 and then returned to the display but the display didn't change so I zeroed out the display (still on the right edge) and moved to the left edge and found it was way off also.

So, now I'm befuddled. DRO doesn't seem to work or I'm doing it wrong and the old way doesn't work or I'm doing it wrong and most importantly I can't do anything because it doesn't work. Tried looking for a new video on calibration or even edge finding but there are none yet for these topics for this new version.

I apologize but I'm quite frustrated at the moment.

Michael

 

[ycroosh]

After trying to use the new edge finder function using a manual edge finder and getting weird results, I decided to recalibrate my mill (I thought new software - recalibrate). This became a daunting task. The new process of going to the axis settings and clicking the calibrate button kept giving me a message stating that the final number was too small to be used and to to try again. So I clicked the "try again" button and followed the instructions only to get the same message over and over again. So, I resorted to the old way of calibration but that didn't work either.

Now, for the Calibrate button experience; first I set the CPI to 10000 then clicked the Calibrate button which asked me for the size of my calibration block - 3" (using a 1-2-3 block). It then instructed me to locate the left edge and then click Continue, which I did and I had to wait for a progress bar to complete while DRO did some calculations. Next I was instructed to locate the right edge and then click Continue, which I did and I had to wait again for a progress bar to complete while DRO did some calculations and then I got the message stating that the number was too small to be used and to try again which I did several times but kept getting the same message each time.

So, I reverted to the old calibration method; I set the CPI to 30000 and then changed the readout to 4 positions then indicated the left and zeroed out the display. Moved to the right side and indicated it. Took the reading on the display and entered that as my CPI and when I returned to the display, it was way off.

Tried entering the standard of 5080 and then returned to the display but the display didn't change so I zeroed out the display (still on the right edge) and moved to the left edge and found it was way off also.

So, now I'm befuddled. DRO doesn't seem to work or I'm doing it wrong and the old way doesn't work or I'm doing it wrong and most importantly I can't do anything because it doesn't work. Tried looking for a new video on calibration or even edge finding but there are none yet for these topics for this new version.

I apologize but I'm quite frustrated at the moment.

Michael

Michael,
Edge finder non withstanding [there is some weirdness with the current code, which I'm fixing], calibration smells like a hardware problem with the scale adapter. The app waits for 15 seconds to get 10 data samples. Even in "sleep" mode it should get 10 readings per axis.

What version of the adapter are you using?

Regards
Yuriy

 

[redvan22]

Haven't changed anything other than the app being updated. Still using the same kit I got from little machine shop for both my lathe and mill. Is there some way to determine the version?

Why doesn't the old calibration method work now either?

 

[ycroosh]

Haven't changed anything other than the app being updated. Still using the same kit I got from little machine shop for both my lathe and mill. Is there some way to determine the version?

Why doesn't the old calibration method work now either?

LMS kit is the same hardware as the old version of DIY kit, but runs a VERY old firmware version. It uses external HC-05 module, and those can fail or have bad UART connection. It's its sending junk down the line, TouchDRO won't work properly.

You can check what the adapter is sending using a BlueTooth terminal. There are some instructions here: https://www.touchdro.com/resources/troubleshooting/troubleshoting.html (in the TouchDRO Won't Stay Connected section).

I don't know why it was working with the old app, and stopped working with the new version. What you are describing is not a calibration problem. When you enter 5080 (which should be close enough), calibration is not involved.

Regards
Yuriy

 

[ChazzC]

A little off-topic, but related: I didn't re-calibrate after the app updated the first time nor after the recent update, but did a check using a 1-2-3 block and indicator and am within a couple of tenths on all axes (which could be the difference between the 1-2-3 block I originally used and the one I used now). I have the "good" iGaging SS scales and the current version of the Adapter, with no issues for two years.

 

[ycroosh]

A little off-topic, but related: I didn't re-calibrate after the app updated the first time nor after the recent update, but did a check using a 1-2-3 block and indicator and am within a couple of tenths on all axes (which could be the difference between the 1-2-3 block I originally used and the one I used now). I have the "good" iGaging SS scales and the current version of the Adapter, with no issues for two years.

You shouldn't need to recalibrate after and upgrade. The upgrade process takes care of any data migration (at least in theory ).

 

[ChazzC]

You shouldn't need to recalibrate after and upgrade. The upgrade process takes care of any data migration (at least in theory ).

That's what I figured (and expected), but checked just in case . . .

So far, I'm loving the new version of the app – thanks!!

 

[redvan22]

LMS kit is the same hardware as the old version of DIY kit, but runs a VERY old firmware version. It uses external HC-05 module, and those can fail or have bad UART connection. It's its sending junk down the line, TouchDRO won't work properly.

You can check what the adapter is sending using a BlueTooth terminal. There are some instructions here: https://www.touchdro.com/resources/troubleshooting/troubleshoting.html (in the TouchDRO Won't Stay Connected section).

I don't know why it was working with the old app, and stopped working with the new version. What you are describing is not a calibration problem. When you enter 5080 (which should be close enough), calibration is not involved.

Regards
Yuriy

So based on what you've said (along with what ChazzC said) what was once good should still be good. As for the edge finding, you said you're working on the"wierdness" issue.

 

[RJSakowski]

When I calibrate my DRO's, I always approach the reference edge from the same direction. This eliminates any issue with backlash and correcting for the diameter of the edge finder. I do this by setting up a stop block with a face perpendicular to the axis being calibrated and a guide bar parallel to the axis.Then I reference the stop block and zero the DRO. Then I move the table enough to insert the calibration block and find that edge.

I use a 1/2/3 block for the stop and a length of ground O1 bar for the guide bar. The bar is mounted first and swept to ensure that it is parallel to the axis and the 1/2/3 block is butted up to the bar. This gives me a reliable reference datum.

 

[ycroosh]

So, on Sunday I got an email from another person who had similar "distance too short" issue.
I got a bit paranoid, and tested the calibration wizard with every version of TouchDRO adapter I've ever made, and it works as expected.
With some "hacking", I can reproduce the "not enough samples" error on some other boards (and by hacking, I mean messing up the power supply so it drops below 2.5V).

Just got an email form him, and it turned out to be cross-wired scale. He connected A and Z lines, instead of A and B, so the reading was 4x shorter than it would normally be.

Regards
Yuriy