[Geezer (OP)]

My mitre-saw draws a nice red "laser" line on the bit of wood I'm going to saw to show me (approximately) where the cut is going to be.
 
How does it make the beamy thing draw a nice straight line. It's too small for a spinning mirror, and I'm pretty sure there are no "frikkin sharks" involved.

[CliffordK]

Laser?
Don't you hold onto the handle and pull it back and forth...  many many times...  until the board falls in half?

[Sprool]

it uses a small lens to focus the beam. The big industrial ones have sharks attached.

[RD]

... It's too small for a spinning mirror

But doesn't the mitre saw have a spinning disc blade ? ...



attach a laser pointer to the hub and voilà it draws a laser guideline.

[CliffordK]

I don't think I have any laser saw guides.

I did find this after-market attachment for sale.


It appears to be a small disk, probably powered by the initial acceleration in spinning the saw blade.  That means that the laser pointer is only turned on when the saw is actually turned on and the blade is spinning.  It may also align to one side of the cut.

[Bored chemist]

They spread the beam out with one of these
http://en.wikipedia.org/wiki/Cylindrical_lens

[RD]

They spread the beam out with one of these
http://en.wikipedia.org/wiki/Cylindrical_lens

One of those would not produce a line wider than the incident beam.
It would not "spread" the light beam, only focus it into a line rather than a point,
(a line no wider than the incident beam).

[Bored chemist]

I guess plenty of people have laser pens or whatever.
If you pass the beam though a cylindrical glass or bottle of water you will see that it fans out into a line.


You can arrange for a line like that to coincide with the saw cut.

[Geezer (OP)]

Thanks everyone!

I think BC has it right. The "laser" draws the line even when the blade is not rotating.

For an additional twenty bonus points, if the beam is, say, 600 mm long (about two feet), and I try to position a straight-edge parallel to the beam, what sort of angular difference between the beam and the straight-edge could I hope to achieve?

I'm sure it's a lot better than one degree, but how much better?

(If you have not already guessed, there is an ulterior motive behind this question.)

[imatfaal]

disclaimer - I have never tried to line something up with a laser distorted through a cylindrical lens so I have no idea of the "sharpness" of the line that would be produced.  Also Geez hasn't mentioned what accuracy and smoothness there is in the thing being aligned.  But assuming high levels of sharpness and uniformity - and the error coming from human failings....

Personally I can align a straight edge to either side or the middle of a .3mm line - so I would guestimate that .1mm accuracy is obtainable by eye in a real world situation.  If we double that to be conservative and to take account of other factors (ie it not being aligning a straight edge with a line drawn on sheet of paper with a tech drawing pen) to a .2mm error, and assume that we have compounding errors at each end of the 600mm, we end up with a pair of right triangles with adjacent edge being 300mm and opposite edge being .2mm and the "error" angle in the middle. 







I would further widen the range and just give an answer of in the order of hundredths of a degree.  I don't think it would be difficult to make it into the thousandths (or lower) with a little extra effort and cross checking.

I have probably fallen into Geezer's heffalump trap by missing an obvious problem - but I am happy to play the fall guy

[Geezer (OP)]

Thanks Matt!
 
There is no trap here. I have a particular application in mind. 0.04 degrees is not bad, but it's probably not sufficient for what I have in mind.
 
The eye is very good at judging parallels, so perhaps we can do even better than you suggest. I'll probably need to try it to find out. I can provide more info on the application if anyone is interested, but it's extremely BORING.

[SeanB]

I have looked at my laser level, and the lens is a set of cylindrical lenses ( easy to see as the line has patterns from interference from the lenses) to make the line. As to the accuracy of the beam, it depends on the lenses, the power of the diode ( higher power means more spreading of the beam) and the ambient light. Over the distance looked at, the line would probably be dependent on you having the source normal to the surface projected on, and the accuracy would be quite good, as long as you are careful you will get as near perfect straight line. Of course the biggest issue is the blade, most have a kerf of between 3-5mm, and you will find that the edge wanders a little due to the wood being less than isotropic.

[Bored chemist]

Thanks Matt!
 
There is no trap here. I have a particular application in mind. 0.04 degrees is not bad, but it's probably not sufficient for what I have in mind.
 
The eye is very good at judging parallels, so perhaps we can do even better than you suggest. I'll probably need to try it to find out. I can provide more info on the application if anyone is interested, but it's extremely BORING.

OK, so check the name.
What's the application?

[Geezer (OP)]

Thanks Matt!
 
There is no trap here. I have a particular application in mind. 0.04 degrees is not bad, but it's probably not sufficient for what I have in mind.
 
The eye is very good at judging parallels, so perhaps we can do even better than you suggest. I'll probably need to try it to find out. I can provide more info on the application if anyone is interested, but it's extremely BORING.

OK, so check the name.
What's the application?

Don't say I didn't warn you!

I have a thing that's called a mill/drill. It's a vertical milling machine - looks a bit like a drill press on steroids. The motor/gearbox part is supported by a large vertical steel tube. It can be raised, lowered and rotated about the tube. During milling or drilling operations, the motor/gearbox bit is clamped securely to the tube.

The motor part accepts various drilling, milling and boring tools. The item that is being machined is secured to a table that can be moved in two dimensions (the x- and y-axes) by leadscrews during a milling operation.

For some operations it is necessary to unclamp the motor/gearbox bit, usually to change the tool. Then you want to return the motor/gearbox bit to exactly the same "z axis". But that's not so easy! There are mechanical methods, but they take a bit of time.

I was thinking I might attach a laser thingy to the side of the motor/gearbox bit and rotate the motor/gearbox bit until the beam was aligned with a large steel square on the table in the y-axis. To be of any use, I have to be able to to return the z-axis to within a couple of thousands of an inch of it's previous position.

[CliffordK]

I think it is more complicated than you are describing.

If you have some kind of a knee mill, then the table moves in the X, Y, and Z axes.  And, then the head may also rotate to give additional planes of cutting.

You wish to return the head to its last position, not necessarily a fixed position with respect to the table (which you move all the time). 

If you always keep the head absolutely vertical, then you can add some adjustable stops for the rotation part.

If you only need access to the tools, you can raise and lower your knee (or does your machine have a quill?)  That allows all other axes to remain fixed.  And, if you've changed tools, you likely have thrown your Z-axis out of whack anyway, so the new tool will have to be reset with respect to the Z axis. 

If you computerize your table and/or quill, then the computer should be able to return to the exact same position it was last in (again, realizing that each new tool can change the Z axis, and even other axes if the head is tilted.

I'm trying to think of how lasers might help.  If you had two moon shaped lasers, you could mount them to the side of your head, and essentially draw an "X" over the center of where the mill is cutting.  (in the X/Y directions).  The Z direction is still more complicated.  As you moved the head close to your work, a large cutting head would likely obscure the center point though.

A single dot laser could be aimed at a non-moving place in your shop.  At the back of your machine, wall, whatever.  But, that still is assuming you aren't intentionally tilting the head.

[Geezer (OP)]

It's not that complicated. All I want to do is rotate the head to some known angle on the tube. There is no index on the machine that allows you to do that.

On some mills, the vertical movement runs on V-ways, so the z-axis is constant, but on this machine the z-axis can be anywhere on an arc described by the rotation of the spindle about the vertical support tube.

[SeanB]

So you want to attach the laser level to some fixed support to generate a vertical line up the tube, and then use a steel tape ( or something that allows you to have a set of lines that you can read relative to the beam) to come back to the exact position on the tube. You want a post attached to the machine base out of the work area , and projecting a line onto the tube of the post, and then a section of the collar where you can read the position of the line on a scale, to get it back in alignment. Simple enough, and you probably will want to have an external regulated 5V supply for the laser ( an old USB charger will do the trick cheaply) so you do not run through batteries too often. You probably will have a old steel tape measure that you can epoxy to the collar to provide the reference pattern at a position where you can get it all the way around the circumference in view of the laser.

[CliffordK]

If it is an angle of rotation,
You can do it the other way around too.
Attach a round laser to the head.  Perhaps design a slit for it to shine through to give you a short, narrow line.

And, a protractor lower down on the back of the mill...  to judge the angles of rotation.  That way, you get a somewhat larger scale than putting it on the head.

[SeanB]

That would work too, though it may be adversely affected by vibration. there you would not need a line, just a dot projected onto a scale, which will give better accuracy if the scale is across the room on the wall.

[Geezer (OP)]

Thanks Cliff and Sean.
 
The idea is to purloin the laser line drawing unit from my mitre saw and attach it to the head of the milling machine so it projects a line on to the milling table. The line would be parallel with an imaginary radial line connecting the axes of the spindle and the support tube.
 
To align the (imaginary) radial line perpendicular with the X-axis of the table I would rotate head on the tube until the laser line was just touching (really "equally touching") the edge of a steel square making a perpendicular with the x-axis of the table (the y-axis). The square could be jiggled left and right slightly as the head was being rotated to fine tune the adjustment.
 
I have a hunch that this method will make it quite obvious that the beam and the edge of the square are parallel within some small limit, although I think that's going to depend on the quality of the lens in the laser.
 
The other tricky bit will be to ensure that the plane of the laser is parallel with the axis of the tube, but that should be a one time setting.