Best way to test dro x axis

[kevin.decelles]

I'm trying to verify the dro is just installed in my 6x26 mill.

What is a good technique for measuring this?

The dials are one reading , the dro is another..... a dial indicator is another ......

I've ignored the dials as I feel this is the least accurate method. I've been using a dial indicator but I believe the misalignment I'm seeing is a result of it being 100% square with the indicator

Is there a full proof way to set up an indicator(s)/ 1-2-3 block or something that is foolproof for base lining the x?




Sent from my iPhone using Tapatalk

 

[Bofobo]

a mic standard?

 

[kevin.decelles]

Sure, but placed/measures from where?




Sent from my iPhone using Tapatalk

 

[Dabbler]

Gauge blocks and a tenths indicator. If you have a very good, verified 1-2-3 block it will do, along with a good test indicator.

 

[kevin.decelles]

Yes, but again, what procedure? Ie: place block here, place mic there, etc




Sent from my iPhone using Tapatalk

 

[Johnwa]

Kevin
Measure the length of your 123 block. Use the indicator mounted in a collect or an edge detector to accurately find the ends of the block.
John

 

[PeterT]

Here is how I did mine. I have a Chinesium 1-2-3 block with signed factory specifications (which I really, really trust HaHa) that says its accurate to 0.0001" length. Using my best micrometer of supposedly same accuracy & pre-referenced to the included length standard, I check the block dimension. Depending on where I measure it I get 2.9998 - 3.0002. So recognizing this tolerance, the block becomes my 'measuring stick'. I doesn't particularly matter what the actual number is, whats important is trusting the reliability of the number.

- set the block up in the mill table & secure it. It must be perfectly aligned to the table action. You verify by running DTI down the length across the ends
- clamp something accurate to say the left side of the block so a surface stands proud, maybe a gage block or something
- mount your DTI in spindle, face it towrads the left side, lock the spindle
- approach the protruding surface from right to left to minimize backlash until the DTI deflects a bit & set the DTI dial position to zero. Zero your DRO & zero your dials while you are at it but we wont trust it yet
- now back up & do the same thing to contact the right end of the 3" block. Stop when the DTI needle reads zero. Now compare this block displacement measurement to DRO? If your block was 3.0000" and your DRO reads 3.0020" then DRO is reading high by .002" per 3" or .00067 inch/inch.
- BUT, this is where the original tolerance of the block validation comes back in. My block could could be +/- .00020" which is +/- 30% of the reading error. So its somewhere in that range. And we also have possible parallax & repeat error, so do the test a few times in different block positions.

You can see if you used the 6" side of a 2-4-6 block of the same +/- grind tolerance it would be better yet because you are performing the same test over a longer distance. You could also stack several 123 blocks to achieve a longer distance yet, but that introduces new tolerance issues. My Chinesium 123 are +/- 0.0005" so 3 shorts could result in 8.9985" length or 3 longs would be 9.0015". Or it could be some in-between combination like (short + long + short) somewhat cancelling one another. But how would you validate this reference length unless you had an instrument that was accurate to .0001" on a 9" range? Also we are adding more mating surfaces into the equation.

So my own opinion is, if you have decent enough 123 blocks & trustworthy DTI & micrometer to validate the blocks you can probably get 'pretty good' using this method. Or at least hi-lite any glaring DRO errors. Failing that, if you can borrow a known accurate standard for DRO validation, better yet. BTW my Taiwan mill RF-45 leadscrew screw dials turned out to be surprisingly accurate vs DRO in the end. I seem to recall .002" over 10" with backlash factored. The thing is it can also vary on different segments of the leadscrew depending on how it was thread cut & wear etc.

I think DRO's can suffer variations but for different reasons. You are counting on encoder head resolution & scale etchings or windings or balls or whatever system they employ. Some DRO's allow you to input a fudge factor map to 'correct' it to some physical standard. But I'd say do everything you can to get the system mechanically straight & use accurate standards to validate. In reality very few hobby machines are going to cut anything to .001" anyway. Good luck.

 

[kevin.decelles]

Thank you -- this is what I needed


Sent from my iPhone using Tapatalk

 

[kevin.decelles]

Thank you -- this is what I needed


Sent from my iPhone using Tapatalk

 

[Dabbler]

You can borrow my gauge blocks. they are accurate to 20 millionths.

 

[Tom Kitta]

I bet the dro is correct and the dials are off a bit.

Its easier to test reputability over whole table length - It should be at least 0.001 +-

Bigger problem on mine is that clamping always moves it by few tenths - same as heavy operations.

 

[kylemp]

You're supposed to put a good surface as a reference, zero a test indicator to it, move away and then put a known size gauge in, come back until you hit zero on the known and look at what reading you're getting.