Lathe DRO Issues yet to be solved; or, X Axis Head scratching...

[verbotenwhisky]

My issues started with the DRO which came on my lathe, it had some slight damage and was building error at 0.001" at a trip. Move the carriage off of a known point then return to that point and you would be at a plus 0.0012X repeat the process and it would be at 0.0027X and so on and it was occurring on both the X axis and the Y/Z axis. Talked to Mike at Precision Mathews (PM) and he was great help, very patient but finally we met in the middle and I now have a new DRO setup and I have installed it. I now have an Y/Z Axis which is rock solid but, the X Axis is increasing at 0.313" at a single trip and I cannot help but think I am missing a setting. I did set the scale resolutions X=1 and Y/Z=5 all linear and I am currently paying attention to my mounting of this scale to see if I possibly have an issue there.

Any thought on this would be appreciated, I am currently of the opinion that this is a setup issue, unfortunately the documentaion is a bit thin here.

 

[Inferno]

Sounds like something that got damaged was your X-axis scale

 

[darkzero]

When you installed the replacement scales, did you verify for parallelism between the scale & the read head using an indicator? Would differ depending on manufacturer & scale type (glass vs mag) but for glass scales in this size range should be no more than a 3-4 thou. If not adjusted properly you may get a cosine error but not nearly as much as you say if I'm reading correctly. Also need to set the gap between the read head & the scale, about 30 - 60 thou clearance. Glass scales usually ship with shipping protectors for the read head. Usually a strip or pieces of plastic inserted between the read head & scale. These should be removed before setting up alignment.

Move the carriage off of a known point then return to that point and you would be at a plus 0.0012X repeat the process and it would be at 0.0027X

Just to be clear, are you experiencing a repeatability error or accuracy error or both? According the quoted above sounds like repeatability. What are you using to keep track of the starting point? An indicator of some sort like a thousandths DTI? Don't rely on the handwheel dials for now.

the X Axis is increasing at 0.313" at a single trip

So using a dial indicator (most basic so I'm assuming you have one), set a zero reference, advance say 1.000" or 2.000", if you retract back to the zero reference point, you get a difference of .313" on the DRO instead of back to 0? If you advance 1.000" on the DTI (assuming no cosine error), does your DRO also say 1.000"?

 

[WobblyHand]

Was both the glass scale and read head replaced or just the read head? I had an issue with my X scale not returning to zero. The error gradually became larger and larger. Similar to you, my original skip counts were so bad that you could use a simple scale and see it was way off. I cleaned the glass, and that didn't help. The read head didn't look dirty at all, even under magnification. PM sent me a new read head and things seem to work great now.

The X scale was relatively easy to service and align - once I had removed all the cosmoline between the scale and the table! The cosmoline had glued the scale to the back of the table.

Use a pair of 123 blocks and a DTI to check the scale. It's a good reliable way of measuring the performance, since you can detect even minor differences. Clamp a block to the table. Adjust your DTI so some preload on the block face. Set the DRO to zero. Move table away. Insert second block flush to first. Measure distance to second block, using the identical load on the DTI. Read out DRO value. Remove 2nd block. Return table to first surface so that the DTI reading is identical. Result should be 0. The DRO measurement of the distance to the first block, can be verified using a micrometer on the second 123 block. Hope this helps. Maybe this is a variation of what you used...

 

[verbotenwhisky]

When you installed the replacement scales, did you verify for parallelism between the scale & the read head using an indicator? Would differ depending on manufacturer & scale type (glass vs mag) but for glass scales in this size range should be no more than a 3-4 thou. If not adjusted properly you may get a cosine error but not nearly as much as you say if I'm reading correctly. Also need to set the gap between the read head & the scale, about 30 - 60 thou clearance. Glass scales usually ship with shipping protectors for the read head. Usually a strip or pieces of plastic inserted between the read head & scale. These should be removed before setting up alignment.


Just to be clear, are you experiencing a repeatability error or accuracy error or both? According the quoted above sounds like repeatability. What are you using to keep track of the starting point? An indicator of some sort like a thousandths DTI? Don't rely on the handwheel dials for now.


So using a dial indicator (most basic so I'm assuming you have one), set a zero reference, advance say 1.000" or 2.000", if you retract back to the zero reference point, you get a difference of .313" on the DRO instead of back to 0? If you advance 1.000" on the DTI (assuming no cosine error), does your DRO also say 1.000"?

Yes, I carefully aligned the scale with the cross slide, used the DI to do so. I am currently making certain the reader is properly aligned with the scale, I had shimmed it out but may not have gotten it far enough.

It is an accuracy error at this point, it seems to add .313, darn near 3/8" which I find puzzling, I had a repeatability issue with the original DRO and it was doing it on both scales.

 

[verbotenwhisky]

Was both the glass scale and read head replaced or just the read head? I had an issue with my X scale not returning to zero. The error gradually became larger and larger. Similar to you, my original skip counts were so bad that you could use a simple scale and see it was way off. I cleaned the glass, and that didn't help. The read head didn't look dirty at all, even under magnification. PM sent me a new read head and things seem to work great now.

The X scale was relatively easy to service and align - once I had removed all the cosmoline between the scale and the table! The cosmoline had glued the scale to the back of the table.

Use a pair of 123 blocks and a DTI to check the scale. It's a good reliable way of measuring the performance, since you can detect even minor differences. Clamp a block to the table. Adjust your DTI so some preload on the block face. Set the DRO to zero. Move table away. Insert second block flush to first. Measure distance to second block, using the identical load on the DTI. Read out DRO value. Remove 2nd block. Return table to first surface so that the DTI reading is identical. Result should be 0. The DRO measurement of the distance to the first block, can be verified using a micrometer on the second 123 block. Hope this helps. Maybe this is a variation of what you used...

I actually replace the entire unit, the original had a building error which increased with each cycle on both scales PM worked with me to figure it out and the display had been damaged (Slight in appearance) so PM finally said lets replace it and I offered to meet them in the middle to get the DRO I wanted. I have setup the new unit and the Y/Z axis is perfect now but the X is being a problem child.

I have been using the stop on the carriage for the cross slide as my reference point, bump it, set the DRO to 0 then run it out to 3 and 4 inches and back to the stop, +.313xx at the moment. I wish this had better documentation because I suspect it is a simple setup problem but not certain....

 

[WobblyHand]

I actually replace the entire unit, the original had a building error which increased with each cycle on both scales PM worked with me to figure it out and the display had been damaged (Slight in appearance) so PM finally said lets replace it and I offered to meet them in the middle to get the DRO I wanted. I have setup the new unit and the Y/Z axis is perfect now but the X is being a problem child.

I have been using the stop on the carriage for the cross slide as my reference point, bump it, set the DRO to 0 then run it out to 3 and 4 inches and back to the stop, +.313xx at the moment. I wish this had better documentation because I suspect it is a simple setup problem but not certain....

Just for clarity, you replaced both the bad scale and the display? Or? And if you swap x for y, on the display, the problem follows the scale?
Likewise, x for z, at the display, and the problem follows the scale? Apologies if you did all that. Hard for me to tell from way over here.

Still could be read head alignment, but when I did a custom install on my lathe (glass 1um scales), it wasn't hard at all to get the read head alignment.

On my mill, I did a read head swap. (Non return to zero, like your problem.) To align it, I found that PM used a stack of washers, maybe some are missing? Even so, it wasn't difficult, or super critical, save for dropping the silly washers. Don't want the read head cocked, just reasonably straight.

 

[verbotenwhisky]

Just for clarity, you replaced both the bad scale and the display? Or? And if you swap x for y, on the display, the problem follows the scale?
Likewise, x for z, at the display, and the problem follows the scale? Apologies if you did all that. Hard for me to tell from way over here.

Still could be read head alignment, but when I did a custom install on my lathe (glass 1um scales), it wasn't hard at all to get the read head alignment.

On my mill, I did a read head swap. (Non return to zero, like your problem.) To align it, I found that PM used a stack of washers, maybe some are missing? Even so, it wasn't difficult, or super critical, save for dropping the silly washers. Don't want the read head cocked, just reasonably straight.

That is what I am looking at now. the bracket is .28 thick. the scale back plate is .382 Thick and there is a difference of .084 due to the cross slide ve the carriage and I shimmed it .034 so there may be the issue there I am hoping...

 

[verbotenwhisky]

Finally, Resolved the issue.... Turned out that the new DRO X axis had and issue with the reader, apparently it had been installed crooked. I talked to PM (Mike) and told him I had pulled the reader out just far enough to see that the ball was in its socket and thought it odd that the board of the reader seemed to be at an angle; he asked me to take a photo so I pulled the reader back out took a photo and as I stared at the photo I realized the little wheels on the back side appeared to be out of position. I pulled the reader completely out of the unit and carefully positioned it so the wheels were in the proper spot(s) as I slid it back together with ball in socket and installed it with properly shimmed bracket and now it is spot on.

PM (Mike), Thanks for your patients and help, it was appreciated.