[How do I?] DRO troubleshooting

Your welcome. Glad you figured it out!

It’s worth checking your indicators for repeatability, but they are probably good enough at measuring distance. They use gears internally, not screw threads (unlike most table and quill travel dials) so they won’t have the metric/inch issue.

To check repeatability, put something with a known dimension on top of your mill vice. The shank of an end mill works well if you don’t have gage pins, 1-2-3 blocks, etc. Roll the shank under the indicator until you find the point of maximum deflection, then raise or lower the quill to zero the indicator and lock it at that point. Then repeatedly roll the shaft under the indicator point, ensuring it returns to zero each time.

If you do check the calibration for distance measurement you’ll need a small set of gage pins or gage blocks. Zero at the top of the reference and read the value when removed. Then do the opposite, zeroing on the table or vise top.

Beware of cosine error (if the shaft of the DI isn’t perfectly vertical). Also, always roll the pin from the same direction — it’s usually best to come from behind, especially with an angled DTI lever arm.

Lastly, if your dials consistently read 0.015748 less than expected per inch of travel, then they are mislabeled metric screws. (25.4 x 0.984252 = 25)
 
Last edited:
I might have solved the problem. ALL the screws were loose! LOL! I've reinstalled them with Loctite.

I think this explains the readings being off at first then getting somewhat more consistent. The screws were loose enough to have a small amount of play that was taken up by travel.

I'm waiting for the Locite to cure, then I will
Glad you found the issue.

Too late now, but for future installations I’d suggest Socket Head Screws (either cap or mushroom head) and lock washers with threaded holes or K-Nuts for through bolting:
IMG_8561.jpeg
IMG_8562.jpeg
The socket heads make it easier to securely tighten (and easier to untighten) and lock washers/nuts hold securely while allowing easy loosening for adjustment. Also, use hardware that matches (metric or inch/imperial/English) the original hardware on your machine; makes it easier for the next owner (or if you forget in 10 years what you did ;) ).

Good luck moving forward and have fun!
 
Also, use hardware that matches (metric or inch/imperial/English) the original hardware on your machine; makes it easier for the next owner (or if you forget in 10 years what you did ;) ).
This^^^^

It's a pain in the neck with mixed fasteners, been victim to that. Especially with fasteners that are really close like M5 and 10-32! They are almost the same, but NOT quite.
 
This^^^^

It's a pain in the neck with mixed fasteners, been victim to that. Especially with fasteners that are really close like M5 and 10-32! They are almost the same, but NOT quite.
Exactly - this is why many of my designs have mixed dimensions. For example, DRO installation on my Mini-Mill – hardware, metric; stock, inch; hole/feature locations, metric; overall dimensions, depends:
20241124 X-Axis DRO Scale Mount.png
 
use hardware that matches (metric or inch/imperial/English)
I am so guilty of not doing this. I use whatever I have on hand (usually SAE), thinking “I’ll replace these eventually”. Then promptly forget to order the correct fasteners (or worse: receive them, not use them immediately, then forget why I ordered them). God help whoever inherits the stuff in my shop.
 
I was able to do some testing today. I got the DRO and the dial indicator to read within 0.001" of each other. There are a few anomalies, which I suspect is caused by the z-axis dial gear sticking/having play. I heard and felt it move when I was measuring, but I was looking at the DRO and DI and didn't see it, but I heard it. Here's the table for your viewing pleasure:
mill DRO 2.JPG

I checked the mill z-axis dial as well. I used the DI as the reference. It appears the dial is off by about 1.75%. This is close to 1.6% which is similar to the metric/inch error. The margin of error is not really consistent, though. The last zero is from taking the quill all the way up from the last reading. Again, I suspect this is from some sort of gear slop on the dial.
mill dial reading 3.PNG

Lastly, here is the setup I used. The indicator stand is on the vise out of the frame. I aligned the back of the DI with the side of the vise to get it as close to upright as possible to eliminate any angular errors.
DSC01668.JPG DSC01669.JPG

I think the issue seems to be resolved. I'll have to machine something and see how ot goes.
 
You really need some fixed dimension gages instead of relying on your indicators for quantitative measurements, but it definitely appears that you have a metric dial labeled for “inches” (inch-ishes).

If you have 100 divisions on your dial, it appears it’s actually moving 2.50mm/rev (0.098425”/rev) instead of 0.100”/rev. (Mine was the same, fwiw.)

Doesn’t matter since you have a functional DRO now. :chunky:

Also, the final error from raising the quill is almost certainly due to backlash.
 
Last edited:
Re: backlash, I bet your +/- 0.001” readings were due to backlash, too.

Don’t zero the dial and indicator with the quill at its full top position. Use the hand wheel to travel downwards a bit to take out backlash, then zero your dials and the DRO. Then record all your measurements ONLY moving downward.

Note that there is backlash in the quill/head/column gearing as well as within the indicator.

Assuming your scales are mounted rigidly and parallel to the travel, the DRO should read much more accurately/repeatably than +/- 0.001” (like 10X better).
 
Yuriy has an in depth calibration procedure (this uses the TouchDRO wizard, but the setup and procedure is applicable to manual calibration as well):

 
Last edited:
Re: backlash, I bet your +/- 0.001” readings were due to backlash, too.

Don’t zero the dial and indicator with the quill at its full top position. Use the hand wheel to travel downwards a bit to take out backlash, then zero your dials and the DRO. Then record all your measurements ONLY moving downward.

Note that there is backlash in the quill/head/column gearing as well as within the indicator.

Assuming your scales are mounted rigidly and parallel to the travel, the DRO should read much more accurately/repeatably than +/- 0.001” (like 10X better).
@ARC-170 has iGaging EZ-View scales:

IMG_8577.jpeg

Even the newer Absolute Plus scales are not accurate to 0.0001”:

IMG_8576.jpeg
 
Back
Top