PDA

View Full Version : Checking DRo accuracy



loose nut
01-20-2012, 07:36 PM
I want to see how accurate my lathe DRO is in the Z axis.

First I would hold a gauge block against the spindle face, D1-4, and then move a dial indicator up against (about .100" compression on the indicator plunger) it and set the arm to zero. Then remove the block and bring the indicator up to the face until the indicator shows the same reading as it was with the gauge block in place.

The problem is I keep getting different readings on the DRO while doing this, about .0002 to .001" /inch of travel but always different. I beleive this is due the the way I have to hold the gauge block while trying to set the whole thing up. Or not.:( .It is hard to get a good stable set up when there is such a small amount that can cause an error.

Any ideas on a better way to check the accuracy of this DRO.

tdkkart
01-21-2012, 12:48 AM
WAY too many variables in your setup to claim to be checking accuracy. You've got a DRO that measures to .0002, but you've got a dial indicator that is probably graduated to .001, how are you supposed to compare the two??

loose nut
01-21-2012, 10:33 AM
That's my point. How would I check it.

On my mill I place a gauge block in the vise and check the length of it against the dro using a 3D Taster which is super accurate and I get good results but I can't do that on a lathe. I can't mount the taster on the lathe and I have no way to hold the gauge blocks properly.

There must be a better way to do this.

JoeLee
01-21-2012, 12:06 PM
You have the right idea but from what it sounds like you have to refine you process and remove some of the variables. The way checked my readouts was I mounted a dial on the cross slide...... moved it into the face of the chuck....... zeroed the DRO and set the dial to 0.
I then took a 4" gage block held it firmly against the face of the chuck where I touched off with the dial point...... moved the carriage to the end of the gage block and came into it with the dial point. When the dial read 0 the DRO did also. I used a dial that read in .0005 increments. It worked fine for me.
I also repeated it a couple times and had the same results. Also repeated the same process with a 1" and a 2" gage block and the results were accurate.


JL.................

dian
01-21-2012, 01:12 PM
i think you are checking your dial against the dro. you need this:


http://i973.photobucket.com/albums/ae218/romandian/komode010.jpg

loose nut
01-21-2012, 01:17 PM
You have the right idea but from what it sounds like you have to refine you process and remove some of the variables. The way checked my readouts was I mounted a dial on the cross slide...... moved it into the face of the chuck....... zeroed the DRO and set the dial to 0.
I then took a 4" gage block held it firmly against the face of the chuck where I touched off with the dial point...... moved the carriage to the end of the gage block and came into it with the dial point. When the dial read 0 the DRO did also. I used a dial that read in .0005 increments. It worked fine for me.
I also repeated it a couple times and had the same results. Also repeated the same process with a 1" and a 2" gage block and the results were accurate.


JL.................


That's basically what I did but holding the gauge square and still is a bit of a problem for me, shaky paws. Also I don't have a Dial gauge accurate enough that I can put on the lathe.

JoeLee
01-21-2012, 03:28 PM
OK, understandable........... it is a little tricky to hold a 4" gage block square and firmly against the chuck with one hand. The dial gage I used is a Starrett, not a digital one.

The other thing you could do if your not comfortable with the fore mentioned
way is to put a piece of stock in the chuck...... face the end, zero your read out, turn it back 2" with about a 1/8" DOC and then measure from the faced end to the shoulder. That should give you a pretty good indication.



Those are really the only two ways I can think of doing it as you need something of a known size to check against.


JL................

loose nut
01-21-2012, 05:40 PM
Today I went out to the shop and tried again. This time I stuck the gauge block to the spindle face with a magnet, no more shaky hands.

I also dug through my collection of dial indicators and found one I had picked up used at Names many years ago. Its graduated too 0.00005 (half a tenth) but it hasn't been calibrated in who knows how long, so it's accuracy is questionable but will have to do for this.

Now I have a more stable setup but I was still getting some variation in my readings. For the most part they where within a tenth on the gauge which is within the minimum divisions of the DRO (.0002) if I was checking the same point over and over.

If I set the start point and then moved the saddle to another point and compared that to the gauge block I was still getting different readings. Most of the time it would be right on the money but sometimes it would vary a couple of thou. I don't know if I am putting an error into this set up or there is something wrong with the DRO.

The strange thing is that if I go back to my starting point (were I set the indicator against the gauge block originally) it is always dead on.

macona
01-21-2012, 06:09 PM
As long as your DRO scale is mounted parallel to the axis and the head is as well your dro will work.

If there is something messed up with the scales with DRO will probably detect it and give you a fault. Simplest way it to check is to mound an DTI to the carriage, zero off on the face of the chuck and move the carriage towards the tailstock and return to see if it returns to the same spot.

Maglin
01-21-2012, 07:41 PM
Today I went out to the shop and tried again. This time I stuck the gauge block to the spindle face with a magnet, no more shaky hands.

I also dug through my collection of dial indicators and found one I had picked up used at Names many years ago. Its graduated too 0.00005 (half a tenth) but it hasn't been calibrated in who knows how long, so it's accuracy is questionable but will have to do for this.

Now I have a more stable setup but I was still getting some variation in my readings. For the most part they where within a tenth on the gauge which is within the minimum divisions of the DRO (.0002) if I was checking the same point over and over.

If I set the start point and then moved the saddle to another point and compared that to the gauge block I was still getting different readings. Most of the time it would be right on the money but sometimes it would vary a couple of thou. I don't know if I am putting an error into this set up or there is something wrong with the DRO.

The strange thing is that if I go back to my starting point (were I set the indicator against the gauge block originally) it is always dead on.

Do you mean a couple tenth's? A couple thou .001+ would be a very large error for a DRO and might be a broken or dirty scale. A couple tenth's .0001+ could be dust or even the heat expansion on the gauge from your hands. Most older cheaper DRO's have a resolution of .0002" and if you are in between a scale graduation could explain the +/-.0002". Also check your reader heads mounting bolts on the carriage to see if they might be loose. That can cause a few tenth's error if the head is not firmly fixed on the carriage.

I think the best check is with a quality 1" .0001" indicator with it reading directly off the face of the chuck. I wouldn't use gauge blocks to many variables to cause .0001"+ errors. Oil Film could even be getting involved.

Rich Carlstedt
01-21-2012, 08:07 PM
If you are checking the Z axis, Why are you using the spindle ?
Accuracy of the scale has nothing to do with the spindle, or chuck.
They only confuse the issue due to possible bearing slop.

Using an indicator is fine for "repeatability" such as
Indicate the face of the chuck , back off and return to "0" and see if the scale repeats. adding a jo-block puts force on the chuck ...don't do that

For scale "verification", clamp a block/angleplate to the ways.
Put a known standard ( like 6", 12") firmly against the block, and indicate it, then remove the standard and indicate the block by moving the carriage towards the block.

Rich

loose nut
01-22-2012, 10:19 AM
Good idea, that might explain the errors I'm getting.