Precisely measuring a right angle

[bilinghm]

I want to make a precise measurement of the angle between the saddle way (Y axis) and the column on an X2 mini mill. Any thoughts on how to do this? I suppose a carpenter's square would get you close. Is there a good method to get the column really square front to back?

[Geof]

You don't use a square you use a dial guage on an arm in the spindle so you sweep out a circle on the table. When the gauge shows the same reading all the way around the circle your spindle is perpendicular to the table in all planes.

[bilinghm]

Ideally, I want to square the column independently from the head. The spindle bore may not be square to the column. Tramming the spindle without knowing if the column is square to the table could introduce error as the head moves up and down.

[Geof]

Originally Posted by bilinghm Ideally, I want to square the column independently from the head. The spindle bore may not be square to the column. Tramming the spindle without knowing if the column is square to the table could introduce error as the head moves up and down.

Ah.

I think this works:

Tram the spindle. Now you know the spindle is true to the table.

Now clamp a piece of stock to the table and bore a hole in it, and don't move anything.

Now mount you dial gauge so the probe is about the same distance out from the end of the spindle as your boring tool was. Dial the bore and it should come out perfect, within the tolerances of your dial.

Now raise the head and mount the dial on a long extension, and dial the bore again.

If the reading show something out of line then you know your column is not square to the table.

You also know your spindle is not parallel to your column ways, and if that is the case I am not sure how you fix it.

[One of Many]

If the main issue is spindle tram, then yes, as Geof points out, you could shim the base of the column in order to get the spindle tram front to back.

However, if you tilt the column this may put it out of square to the y-axis ways. The point here is the surface the head mounts to can also contribute to the spindle misalignment.

With a decent quality machinists square, clamp the beam to the column ways and place an indicator on the table. Touch the bottom of the square's blade with the indicator and run the y-axis in and out to see how parallel the blade is in relation to the column squarness before and after you shim it. At least with this information, you will know if your compensation messed with any other alignment.n A thou or two on these light duty machines may be close enough considering the spring they seem to be prone to.

Now I see you have already addressed this!

Secondarily, the quill may also come out of the head at an angle.

DC

[lerman]

If the spindle is square to the motion of the table in both X and Y, you can correct for the Z axis motion's not being square in software.

It would be pretty reasonable to add a module to do this to EMC.

Ken

[toastydeath]

You unfortunately need a reference standard to do this, but this is how they align high precision machines.

You stick a dial indicator on the Z slide, and the angle standard on the table. Z goes up, and the errors recorded. The standard is flipped end for end, and it's done again. Again, recording the errors. The difference between the two is the error in the Z axis.

[under-dog]

To do mine I:

1)swept with a gage to make sure the column(Zslide)was square all around to the table.

2) put a piece of very straight rod(I use an ejector pin which is very true)into the spindle collet and the using an indicator against the rod ran the Z up and down. This should tell you if the head is square to the Z slide. Approach it fron the x and Y plane to make sure it is square both ways.



I think it can be a back and forth type of thing. If one of these is out it can throw both off.

[Andre' B]

In days gone by you would hire a strong young guy, who did not have a real clear idea of what he was getting into but thought he wanted to learn the art of scraping.

Give him some tools and three rough plates an a few words on technique, and the three surface method. If he could read give him a book like this.
http://books.google.com/books?id=npYNAAAAYAAJ&pg=PA233
If he can't read set him up with some tutoring time with a good looking girl about his own age.

When he is done and you have three nice flat plates you give him three rough blocks and tell him to scrape one surface on each block flat. Let him decide to use the three surface method or one of the reference surfaces he just finished.

When he is done with that tell him to scrape a second surface on each block using the three surface method but when checking them against each other the surfaces he just finished are to be resting on one of the flat plates from step one.

If he does not walk out or run off with the girl, you will have three surface plates, and three 90° reference angles and one good worker who is not easily distracted.

[Mike Nash]

I recently did mine. You do want to get the column trammed separately first on the mini. It's also a pain, because you are going to have to scrape or shim and either way you are going to screw up the side to side tram while trying to get the front to back right. So you get to do both over and over and ...

I put an indicator on the column and ran it up and down the edge of the blade of a (supposedly accurate [but not really]) machinist's square held blade up, parallel to whichever direction I was checking. I used heavy duty aluminum foil because it was thicker and mine needed quite a bit. I'm pretty sure you are reducing rigidity when shimming as opposed to scraping where it would be potentially increased. One thing to remember; put the shims at about the 10 and 2 o'clock positions. This keeps the compression over the ways. I tried in the center (12 o'clock) and only succeeded in temporarily warping the cast iron. Build up both sides equally. And tighten the big nut TIGHT and recheck. But make sure the side to side is good before you get it much more than snug because it isn't going to want to move otherwise, even with a dead blow hammer. The shim positions are assuming yours leans toward the table. Both of mine did.

Have fun! Bwaa ha ha ha ha!

[toastydeath]

Any test of the position of any axis needs a reversal in it, or you aren't testing the axis, but the axis error plus any error in your artifact. Even a scrapped 90 degree block won't do the job properly.

There's actually a couple ASME standards on how machine tool axes are to be set up, and naturally I can't remember what they are. B-eighty something?

[philbur]

Put an angle plate on the table, either at right angles or parallel to the column ways depending on which lean direction your are trying to check, and attach the DTI to the head. Run the DTI up and down the angle plate using the Z axis and note the change in reading over a measured travel. Reverse the angle plate by 180 degrees and repeat the DTI reading. Do some simple math on the results and you have your answer.

Tramming the spindle is irrelevant to the trueness of the column.

Phil

Originally Posted by bilinghm Ideally, I want to square the column independently from the head. The spindle bore may not be square to the column. Tramming the spindle without knowing if the column is square to the table could introduce error as the head moves up and down.