Going nuts

[WobblyHand]

I would think the magic number is near 5080. 5 um = 0.005 mm = 1.9685039370078743e-4" (give or take) 1/1.9685039370078743e-4 = 5080.
3" x 5080 counts/in = 15240 counts.

@RJSakowski 's method is very good and easy to set up. Clamp a 123 block to the table and make it true to the table. (Sweep both perpendicular faces to be sure.) Put in a 123 block butted to the clamped block and measure that face. Remove the 123 block and move the table to the clamped block and measure. Assuming a nominal 5080 counts/inch, you ought to get 3.0000" if your removable 123 block is on size. One is always approaching the block(s) the same way. I used a 0.0001" indicator to know when I was at the face (used the same reading for both faces) to ensure contact consistency.

 

[ltlvt]

When calibrating a DRO, you want to approach from the same side of the calibration block. I stack two blocks and zero on the first block. Then I remove the first block and move to the second block. It is also important to have the blocks square with the table to avoid cosine error.

So do I. That is the same method I use. Thanks

 

[RJSakowski]

I would think the magic number is near 5080. 5 um = 0.005 mm = 1.9685039370078743e-4" (give or take) 1/1.9685039370078743e-4 = 5080.
3" x 5080 counts/in = 15240 counts.

@RJSakowski 's method is very good and easy to set up. Clamp a 123 block to the table and make it true to the table. (Sweep both perpendicular faces to be sure.) Put in a 123 block butted to the clamped block and measure that face. Remove the 123 block and move the table to the clamped block and measure. Assuming a nominal 5080 counts/inch, you ought to get 3.0000" if your removable 123 block is on size. One is always approaching the block(s) the same way. I used a 0.0001" indicator to know when I was at the face (used the same reading for both faces) to ensure contact consistency.

I would add that for most accuracy, the 1/2/3 block dimension should be verified. Even name brand 1/2/3 blocks aren't exactly nominal. They are usually slightly oversized to allow for the customer lapping.

 

[WobblyHand]

So do I. That is the same method I use. Thanks

Sounds incredibly frustrating. Has to be something, we just have to find out what the heck it is.

Are these new scales? I had a set of scales get crud in them and they had accumulating error to them. I'd go 3" to the right, then 3" to the left and not end up at zero. Tried cleaning them and they got a little better. The only thing that actually worked was replacing the read head. Then things made sense again.

 

[ltlvt]

I finally got it PDC. Here are the steps I followed.

#1 Verified the scale was working properly by disconnecting it from Touch DRO and connecting it to the battery powered I-gage display. I then used the indicator to verify the left edge of the 1-2-3 block. Continuing with the I-Gage display I moved to the left side and touched off with the indicator. Thus, coming up with a measurement of 3.001 inches. I am confident the .001 was human error on my part.

Next, I re connected to Touch DRO and manually input the recommended 2650 CPI for I-Gaging. This yielded 3.091 inches on the display. NO BUENO !

Next I went back to the Auto calibration and followed the same steps that I was using before but being extra careful to ZERO both the Abs and Inc for the X axis. Following thru until I got to the apply button the CPI was 2561.667.

I clicked apply and then checked the travel and it yielded 3.007 inches. I am confident that the .007 was from Touch off on left and right sides. I am confident I can eventually get it within +or - .0005, The heat is terrible in my shop right now due to extreme heat wave, forecast 109 F tomorrow and 113 F Wednesday.

For now, I will end this thread and when weather and time permits, I will calibrate the Y axis. I've said it before and will say it again this Forum is the best I have seen for getting quick and helpful responses. Thanks again to all who replied. There would be no forum without the 2-way communication.

 

[ycroosh]

I finally got it PDC. Here are the steps I followed.

#1 Verified the scale was working properly by disconnecting it from Touch DRO and connecting it to the battery powered I-gage display. I then used the indicator to verify the left edge of the 1-2-3 block. Continuing with the I-Gage display I moved to the left side and touched off with the indicator. Thus, coming up with a measurement of 3.001 inches. I am confident the .001 was human error on my part.

Next, I re connected to Touch DRO and manually input the recommended 2650 CPI for I-Gaging. This yielded 3.091 inches on the display. NO BUENO !

Next I went back to the Auto calibration and followed the same steps that I was using before but being extra careful to ZERO both the Abs and Inc for the X axis. Following thru until I got to the apply button the CPI was 2561.667.

I clicked apply and then checked the travel and it yielded 3.007 inches. I am confident that the .007 was from Touch off on left and right sides. I am confident I can eventually get it within +or - .0005, The heat is terrible in my shop right now due to extreme heat wave, forecast 109 F tomorrow and 113 F Wednesday.

For now, I will end this thread and when weather and time permits, I will calibrate the Y axis. I've said it before and will say it again this Forum is the best I have seen for getting quick and helpful responses. Thanks again to all who replied. There would be no forum without the 2-way communication.

If you want to take a shortcut, just do the math and set the CPI directly:
The formula <actual CPI>/<current CPI> = <actual readout>/<expected readout>
If we rearrange this. we get <actual CPI> = (<actual readout>/<expected readout>) * <current CPI>
In your case, <actual CPI> = (3.007/3.000) * 2561.667 = 2567.644

Not sure which procedure you are following, but the "official" method is described here: https://www.touchdro.com/resources/dro-manual/calibration.html
There is an old video on YouTube:

(pardon the cheesy music. I was young and naive )
I am working on an updated video. It will be out in a few days, plus minus.


Finally, if you keep having these issues, check the scale. I've seen this happen when the scale is acting up. Right now, in fact, I had very similar issue with a Chinese magnetic scale. Each time I rerun the calibration, I get a different CPI. In this case the scale has very poor repeatability.

Hope this helps.
Yuriy