[cordvision]

It all started with alignment pins for which I drilled holes in two plates not aligning up properly. I posted about it here: http://www.cnczone.com/forums/machin...-dont-jam.html

The drill might have drifted a little bit out of axis but I think I actually have a different issue. I looked at "square" cavities that I milled recently, and I believe my axis (x/y) might not be square to each other.
I also checked for lost motion and all the values seem to be within spec. What is the best way to accurately measure if the axis are actually square to each other? There must be a better way than just milling a right angle into a block of stock and then using a machinist square (that is how I have done it, but I'm looking for an exact measurement).

[pickled]

Are you having a parallelism issue with the spindle too? At any rate I use this tool quite a bit for troubleshooting the type of issue that you are referring to. Just mount it and use it along with an indicator and/or a Haimer.
MAGNETIC CYLINDER SQUARES by Suburban Tool, Inc.
It really helps when setting up to perform bore operations. You can indicate vertically or mount it to the table horizontally and indicate along its length. Look around at the usual tool disposal sites and I am sure that you can find one. I keep mine stored away with my master set of gauge pins.

[pickled]

Here in case you need to see it in action. Rucker does a good job explaining it as he tries to bring his index mill back into tram. Cylinder square use starts at approximately 9 minutes and 30 seconds into the video.
https://youtu.be/U07X_0oo74o

[keen]

Hi - the easiest way to check it is to use (or borrow) a precision angle plate or square eg 6 x6 ' and clock one axis true eg X - then clock the other. If you have doubts about the angle plate or squares accuracy, flip it over and check it the other way - if it is square you will get the same result - if it is different then split the difference to get the same result. Keen

[keen]

6 x 6 " that is ....bigger if you have one. 4" x 6" will work but it takes more care to get good results.

[philbur]

See PCNC1100 milling machine certificate of inspection - procedure G1.

1100 Personal CNC Mill | Tormach Has Affordable CNC Milling Machines

Phil

[handlewanker]

Are you having a parallelism issue with the spindle too? At any rate I use this tool quite a bit for troubleshooting the type of issue that you are referring to. Just mount it and use it along with an indicator and/or a Haimer.
MAGNETIC CYLINDER SQUARES by Suburban Tool, Inc.
It really helps when setting up to perform bore operations. You can indicate vertically or mount it to the table horizontally and indicate along its length. Look around at the usual tool disposal sites and I am sure that you can find one. I keep mine stored away with my master set of gauge pins.

Wow, at that price we'll have a couple of those....LOL, a few spares too.

If you have access to a lathe and know how to turn, just buy a piece of 75mm bright mild steel bar, (the cold rolled stuff with a shiny finish), approx. 300mm long.

Stick it in the lathe, drill centres and run it between centres, checking that the tailstock is giving you a total indicator reading along the bar of a thou or two....top and bottom of the bar......some lathes are so far out you'd never believe.

Then face the end at the tail stock to clean up and as far in as the centre allows.

Lastly, run the bar in the fixed steady rest and undercut the centre of the end face to leave a land at the outside of approx.10mm wide.

The cold rolled or bright steel bar material is plenty good enough to make a precision square with without going to extremes of having the OD ground parallel and round.....it's already within a thou or so from the supplier, and over a 300mm length it won't be bent to any degree, but if you really know your socks, take a light cut over the OD and machine the OD between centres parallel.

Just facing the end of a bar that is as good as you'll get from the supplier will be more accurate than your out of accurate mill will ever be, and it cost just the material cost to do.

BTW, having a parallel bar with the end faced allows you to check all your squares against the side as it doesn't have a flange at the base which you cant butt up against.

If 75mm solid bar is too expensive, try getting some 10mm thick wall steel tubing and machine the OD round and parallel.

BTW.....it's impossible to face the bar end face and have it not be square.
Ian.

[tinmancarving]

For some reason I was interpreting his post as asking about squareness of the Y axis to the X axis.

I do like the video and magnetic square though. When I was a machinist all we had was a dial indicator. It was always fun to watch the new employees tram in a head on a bridgeport. Talking days for some.

Anyone have some tips on that Y axis vs the X axis squareness?

Rob

[handlewanker]

Hi, yes I know what he wanted, but as he doesn't have any reference things to indicate off of, making a square that is 100% accurate means he can calibrate his squares and angle plates and anything else that needs to be known to be accurate......... one good reason why you need a flat surface that is known to be flat, like a surface plate to check the accuracy of your finished work.

Once you have a reference square to work from you can mill up a square piece of plate and check it's squareness against the square........then the X, Y and Z axes can all be checked against one another.....they will be as accurate and square to one another as the square plate you milled up.
Ian.

[underthetire]

Usually a precision square is aligned on one axis, set as reference, then other axis measures the deviation. A laser system can also be used, but squareness optics are expensive and it a PIA to set up. I use a ceramic 6 sided square that's been lapped under a micron total. Less moving the square around to check all 3 axis. Most builders use the square method since its quick to set up.

[handlewanker]

Hi, that would be Ok if you can afford to have such instruments, and also have a definite further use as well apart from checking the X, Y and Z axes once in a while to ensure they remain accurate.

Making one as I described is well within the reach of the average machinist, if you have access to a lathe etc, and it's quite inexpensive........I don't think micron accuracy comes into the squareness of mill axes......nice to be able to achieve them but even at the manufacturing stage it doesn't occur.....tolerance etc, and we don't know what machine he's worried about..
Ian.

[underthetire]

Ya, that square would be overkill. I'd just check it with a good 123 block.

[pickled]

For the folks on a budget that don't have a lathe nor the patience for EBay and auction sites. Here is a page full of alternatives for checking squareness. Epoxy a n52 rare earth magnet to the bottom making sure that it is slightly recessed and you will have a pretty sweet gauge.
Online Catalog
And the angle plates
http://www.shars.com/company/online-catalog?p=177

[handlewanker]

Ya, that square would be overkill. I'd just check it with a good 123 block.

Hi, yes you could check it with a 123 block.....short distance etc, but it's made to a tolerance, so can not be classed as a reference......good enough for average work though.
Ian.

[pickled]

123 blocks are essential to have on hand. You can get a lot done with a basic surface plate, an indicator, a matched set of blocks and a machinist square. I have to do a lot of contact and non-contact metrology for dimensional troubleshooting here so I have a lot of the higher end stuff sitting around to aid me with any troubleshooting. I have seen too many people go down the long winding road of lost productivity while fighting measurement system variation (i.e. phantom errors vs. real machine problems). Heck I remember thinking that I had a squareness problem myself when my reamed holes didn't allow a pin fit. Come to find out it was a tiny chip of stainless that I didn't see stuck to the bottom front of my vise! You could sweep the rear jaw all day and it looked parallel, but the entire workpiece was off in the YZ plane by 0.2 degrees. For me it was a lesson in the need for improved lighting and more elbow grease while cleaning. I now indicate the ways of the vise as well as the jaws for z height deltas any time I move or remount them. Fixtures are treated the same way here too. Hopefully the OP can get things reconciled because having parts that you pour tons of work into and not function as designed really stings!

[philbur]

Hi, yes you could check it with a 123 block.....short distance etc, but it's made to a tolerance, so can not be classed as a reference......good enough for average work though.
Ian.

Isn't a reference made to a tolerance?

Phil

[keen]

The square or angle plate does not need to be perfectly square if you set it up one way - set on table , dial indicate measure eg the X to Y, then flip it over and re measure - eg square stock on table X right side to table X left side - if it is different then there is an error in the square or angle plate. But no matter, just half it to get the X Y squareness measurement.

By flipping it becomes a squareness generating process. keen

[handlewanker]

Hi, I expect the debate can rage into realms where the very smallest increment known to Man is still not good enough for average work........you have to start somewhere if you aren't into having some form of reference equipment to check your requirements........just get it.

I'm not being hyper critical when I say a lack of experience is not rectified by having high class equipment, but the learning curve to get there can be long.

The solutions posed here will no doubt suffice for what you need.......so go for it.
Ian.

[philbur]

There is more than one way to skin a cat, especially when that cat needs skinning very seldom (or maybe only once) in the life of your mill. Purchasing a reference that is an order of magnitude higher in both precision and in price in comparison to your no-name angle plate is not always the way to go when a bit of fiddling and care with your angle plate can produce the same result.

The big problem comes when you find your x/y is not perpendicular to the degree you had hoped, sometimes it can be best not to know.

Phil

[keen]

I think we have some doubters ? - try sketching it - or cut a paper square obviously out of square, sit it on a straight edge to replicate indicating it on the X - then flip it over. ...see how the Y is out in opposite directions.

The process takes a little longer but how often do you need to do it? Perhaps that is what Phil was referring to - "especially when that cat needs skinning very seldom (or maybe only once) in the life of your mill"

The best precision relies on mechanical generation not an attempt at absolute accuracy. Keen