Laser alignment tool

[davidallen]

just got a laser alignment tool from Northern Tool and Equip. that looks like a good addition to the ShopBot. it's their Torpedo Boy Laser Level w/ tripod for $24.99. it has a magnet mount on it and can be fixed to the Z-axis to provide an offset for the tool's x-y zero. it has a small spot size as well as a line generating lens. you can use the line to set zero as well as align stock.

da

[danbrewer@attbi.com]

That's a great idea, David. I bought one myself and am looking forward to trying it out. It sure beats running a v-bit down to .005" and sight-aligning.

One question: since it's attached to your router (assumption here) has it withstood the vibration test? Has the laser ever gone out of alignment (or just plain "gone out")?

[ron_cleaver]

How do you use it to set (0,0) and align stock?

[davidallen]

since it has a magnetic base, I mount it on one of the vertical channel pieces of the z carriage. this gives an offset line for either the x or y axis.

table zero is best set using your table stops. you can then use the laser (with the line attachment) to align stock and calculate or set the edge offset from the system zero. if you want, you can then use that offset to reset the table zero to the stock zero.

I usually remove it while the router is running

[danbrewer@attbi.com]

Ron

It depends on where you mounted your laser. Looking at the cutter head from above, you can split the area around it into four quadrants. Knowing which quadrant your laser is mounted in will determine the offset values you'll use when zeroing out the Bot. When standing in front of your Bot, the 3rd and 4th quadrants will be nearest to you.

Here's a top-level view of the cutter head and the laser mounted in the 3rd quadrant (this looks best when viewing this message in Courier font)


|
2nd quadrant __|__ 1st quadrant
/ \
----- | |------
\_____/
3rd quadrant * | 4th quadrant
^ |
|
Laser mounted here


My approach would be to move the cutter to 0,0 (wherever that may be; it really doesn't matter), chuck up a v-bit and then drop the head down to -0.005" to produce a small dimple. Turn on the laser and arrow the head over until the laser is pointing straight into the dimple.

Make a note of this X-Y location. These values will be used as your offsets.

Then set up a custom cut file (like CUSCUT5.SBC) to offset the ShopBot and Z2 the bot.

Depending on what quadrant you mounted your laser in, you'll need to adjust your offsets accordingly; one, both or neither offset may be a negative value. The rule of thumb for determining which values would be negative, which would be positive, is this:

Laser mounted in the....
1st quadrant: X and Y offsets are both positive
2nd quadrant: X is negative, Y is positive
3rd quadrant: X and y are both negative
4th quadrant: X is positive, Y is negative

For example, let's say your laser is mounted somewhere in the third quadrant, and after moving the laser into the dimple your ShopBot is at X-Y coordinates 6.212, 3.019. Write this in your custom cut file:

'Zero X-Y at Laser
'Move the bot to the NEGATIVE offset of the laser
'for 3rd quadrant mountings
J2 -6.212, -3.019
'Now zero out X-Y
Z2
'Now move to the new X-Y zero
JH
'Done!

Of course, your results will vary, depending on the quadrant your laser is mounted in and how far the laser is from your cutter.

Hope this helps.

[danbrewer@attbi.com]

Thanks for the quick reply, David. I saw a posting of yours in another forum that discussed creating a mount for the laser. I'll take that advice to heart as I REALLY don't want to chunk down another $25 for another laser!

The laser works beautifully. Again, thanks for the fantastic suggestion.

[david@fools.net]

David,
I saw at the description of the Torpedo Boy that its accuracy is rated at +or- .2 inches at 1.1 yards... is that acurate enough?
David Beede

[rgbrown@itexas.net]

Let's do a little simple math here. The specification for "the rest of the world" might read +/- 5mm @ 1 Meter.

My ShopBot has a "Z" of 6" (150 mm) but, to make the math simple I will say it is 200mm - near 8" from the end of our level to the top surface. Now, our accuracy would be +/- 1mm @ .2 Meter. This is the "outside" of the envelope. In reality it is probably much closer but, if you advertise, you must take everything in account or some user will come back and say "I used this level and it was off 1/8" in 100 yards. I want my money back!"

I have one of the older ShopBots with self installed Rack and Pinion up-grade and "home-made" gearboxes. It is more than visually accurate.

The tool is more accurate than most woodworkers. Dan is lowering his tool to "-.005" and making a small dimple. The dimple, if one is using a 90 V-Bit will be .01414". Overall cutting accuracy is advertised on the new models as +/- 0.015". Let me get my "Opto-comparator" and find those last six tenths.....

My point is I hear all this talk about accuracy and the most accurate tool I have seen used on a ShopBot is a dial-caliper or digital caliper. And, how many of you folks wondering if an "X-Y" location of .020 is accurate enough have checked their "Z" for true square with the table? Did you check the table for flat with a "straight-edge" ? A good 48" straight edge is about $150 or more these days. Did you sweep that table off before you placed a "square block" on the table? Did you scrape the table to assure you were on an "average" of the table. Did you do the necessary analysis for temperature compensation of your tools?

Or, as I suspect, are you folks trying to split hair with a sledge-hammer?

Ron Brown

Diplomacy is the art of saying 'Nice doggie' until you can find a rock. - Will Rogers

[gerald_d]

David, don't let Ron's maths confuse you!


The other posters were only talking about using the laser to make a dot. You have looked at the spec for the "level" part. i.e. when you have to level a playing field. That spec tells you how level the field would be afterwards. And quite frankly, 0.2 inches in 1.1 yards is disgustingly bad. But, if you want to use the laser part like the other posters have described, you are not going to use the level part. In fact, a keyring board pointer will probably do that job just as well.

If I really had to, I would buy an industrial quality laser unit, like this (http://www.worldstartech.com/uli.htm) or this (http://www.worldstartech.com/ucl.htm) and make a small adaptor to fit it directly in to the router collet. These laser units are only around 3/8" diam x 1"~1.5" long. I have used them on other applications and they are quite tough. They should withstand being permanently mounted on the z-axis.

But, quite frankly, I don't think that I ever would mount a laser on a ShopBot - I just don't see the need for it. If I really needed to find edges and spots that accurately, I would first call on a metalwork shop and find out more about the tools that they use in their milling machines for this purpose.

[papadaveinwy]

Ron, great narrative! I have always stayed away from theads like this because I always thought this is "splitting hairs" Lets all get real, if you need that kind of accuracy then your work must be for NASA or you are trying to develop a new atomic tunnel for the DOD. Or are you tring to use your bot for calculating quantum physics. David in Wyoming

[gerald_d]

Metalworkers use "wigglers" or "edgefinders" but they have to be used at quite a low speed. Link (http://its.foxvalleytech.com/MachShop3/basicmill/accessories.htm).

[rgbrown@itexas.net]

I'm wondering if he scraped and trued the "Uni-Strut" before he mounted the laser.

On a real note, if one did want to know where the center of bit is they could mount two line-casting lasers, one to the "X" cariage and one to the "Y".

And yes, metal workers do use "wigglers". They also use stops and dial-indicators to find lines and centers.

Imagination, trial and error have taught me more than instructors ever could. I still have "stops" at certain known points on my table. Dowels fit in holes and become locating pins.

There is a saying about more than one way to skin a cat. Trust me, one wants to make sure the cat is dead first. It is also helpful to have a reason and a sharp knife.

My table has a "grid" that acts as a place for vacuum gaskets and vacuum channels. the grid is a on a 2" x 2" pattern and the channels are 3/8" wide. I can usually "spot" a piece within 1/8". The "next time" I rebuild my table I will probably dye sections of the table various colors so I can tell where the bit is without looking at the screen.

[danbrewer@attbi.com]

Ron

Don't get me wrong here. I'm certainly aware of the limits of accuracy in the Bot, and my instructions were really not meant to draw anyone to the conclusion that the Torpedo laser will provide any more accuracy than the Bot inherently has (which is pretty good in and of itself!)

I just see it as a more convenient way of finding an X-Y zero-point than eyeballing down the X and Y axes with a V-bit chucked up, which I found incredibly tedious and time consuming, not to mention inaccurate because it's hard to sight down the Y-axis.

The Torpedo Boy comes with a magnet attached, and I've discovered that by just placing it snug againt the Z gantry that this provides a fairly reliable point of reference every time (it's plumb). No jigs to build for the laser, etc. I've had the extreme privledge to work under shop foremans who believe that "simpler is better", and I've used that philosophy here as well.

You're right, dropping a V-bit down to -0.005" does produce a VERY small dimple (I guess -0.010" would have been better advice!), but once I got the offsets for the laser relative to the bit, it works every time. It's not rocket science, and I wouldn't recommend using this technique if you require more accurate tolerances than most woodworkers need, but for a "quick and dirty" X-Y zeroing routine, this works for my money.

Again, my apoligies for any confusion or seemingly misleading information.

[rgbrown@itexas.net]

Dan,

Actually, my math was rusty, or I was not thinking right. The dimple diameter would be twice the depth. The "length along cut" would be square root of 2 times depth or diameter. And you weren't the one asking if a level was accurate enough.

If the goal is to find "Center Bit" you are on track. One might also think about using the "line generating lens" and put two plates on the "Y" carriage and have a "Cross" where the center of bit is.

The accuracy question is raised many times. I remember an argument with a supervisor about me using the 3-4-5 triangle for "square" and the supervisor insisted I use a 6-8-10 triangle. Some of the folks raise the questions and don't have the tools to measure the results.

There are certain truths in layout and engineering. Braided lines are more accurate than twisted line. A knife cut is more accurate than a pencil line. A fish quits growing if you take a picture.

It is hard to guess where the bit will hit....

[gerald_d]

We all know that the z-zero plate works for positioning the SB in the z-direction. Why don't we develop that principle to the x and y directions as well?

Instead of just a flat alu plate, let's take an alu "cup" (or "ashtray") with a viewing hole cut in the bottom. Put crosshairs across the hole (scribed clear plastic) so that the "cup" can be put on your workpiece accurately in the x, y and z directions. The cup is electrically connected the same way as the one-dimensional z-zero plate.

Now we bribe Bill Young to modify one of his edge-finding programs to probe the bottom (next to the window) and rim of the cup and automatically zero the SB in the x, y and z directions. Sounds too easy, doesn't it?

Why haven't I done it yet? Well, we normally don't need that type of accuracy in our particular line of work. Also, we use mostly single flute cutters that are not symmetrical. And lastly, we don't know if Bill will take a bribe.

[Dan Brewer]

Ron-

Haha. You could have REALLY torqued off your supe and insisted you use a 9-12-15 triangle, then walked of muttering about the postulate of similar triangles (with a few explectives thrown in just out of the audible range).

Some folks you just can't reach.

[rgbrown@itexas.net]

"Now we bribe Bill Young to modify one of his edge-finding programs to probe the bottom (next to the window) and rim of the cup and automatically zero the SB in the x, y and z directions. Sounds too easy, doesn't it?"

I believe I might have a copy of a file similar to that if you would care to see it......

Ron

[davidallen]

after 2 1/2 years, there's still this much interest? well, I guess relating some experiences are in order.

the accuracy of the laser is not an issue. it's not accuracy but precision that's the issue. it could be off by 10 degrees, but if it hits the same place every time, the offset to the bit will be the same. the laser will fit snuggly on my z-axis such that it can be reliably placed every time.

when I use the laser (which isn't often) I use the line attchment and align the laser to a cut in the table. that gets the line parallel to the axis of interest. I can then check that a piece is aligned correctly. there are easier and faster ways of checking stock alignment. my primary use for the laser is for aligning stock that has pre-existing patterns. either materials partially done by someone else or a job that had to be removed for intermediate processing then replaced. usually some forethought to establish keys and work template minimizes re-alignment concerns.

I generally 'zero' the table with contact switches bolted to the rails. the x-y-z zero program works fine for that purpose. when I need to move the table zero, I generally zero the table, apply an offset with the M2 or M3 command then reset the table zero.

occasionally, I'll throw some stock on the table and want to set the table zero to the corner of the stock. to do that, I picked up a machinist's square and mounted it on a block of wood so that it can be clamped to the table but still sit above the stock and aligned to the corner. the square has a wire on it, like the zero plate, I again use the supplied program to zero the x and y axis with the bit contacting the machinist square instead of the frame closing the contact switches.

sometimes I get anal and want to re-zero, either between bit changes or after the system's been turned off for a while. the square works, but I need to check that the outer diameter of the bit is making contact and not the valley between flutes. to eliminate that concern, I mounted contact switches on magnets that I can position along the rail to set a variable system zero. they're also useful if you want to set variable limits in order to protect a job in progress elsewhere on the work table. (remember, the switches and zero plate/square trigger the emergency stop input.)

so after all the high tech and gadgeteering, I'm back to the original SB design of using contact switches and thinking ahead.

da

[gerald_d]

David Allen, that was a great post! And it puts your recommendation of a simple laser perfectly into perspective.

[david@simplicityboats.com]

Thanks to David Allen and the rest of the gang...
I asked about the accuracy of the laser, not to cast aspersions on it... but for education, which I'm getting. In my book .2" or almost a 1/4 inch in 1.1 yards seems pretty bad, but I now know that is the level, rather than the laser. And if you are using it on the Z axis mounted magnetically to project a dot relatively close by, I can see the repeatability would be quite adequate. I'm new at all this, so am mostly asking out of imagination rather than experience, but what about just using the home command and prox. switches? Bill P. uses variations on the program with different offsets to place him repeatedly in differenct quadrants or on the mark on different jigs... are we talking about being sure the jig is in the right place? I like registration pins, and the wired square that contacts the bit idea as well. Has anyone wired simple aluminum x & Y stops to jigs with calculated offsets from the Zero home spot and used them like the Z plate? Seems that would circumvent all the eyeballing - they could even perhaps be firmly affixed to the edge of a jig, and pivot up and down on a bolt say, to keep clear of router bits. Perhpas jacks could be wired into the control box, or into each jig to allow for quick connect & disconnect? Am I being too anal here? I do make musical instruments, so I lean that way, but it seems the Shopbot specs are way better than my old eyes.
Thanks again for all the ideas.
David Beede

[ricks]

David,
I zero my bot very much the way you have suggested. I use a copper pc board with a wire soldered to it and split with a banana plug that goes to the control box. (The wire also gets plugged into stops at x&y zero points to zero the overall machine.) But I use it mostly to zero the xy&z of jigs on the table that I can place anywhere. I have a wooden fence at the x axis and use plywood spacers to make sure the jig is parallel to it. I lay the pc board against the jig and have the router spindle hit it. (I'm using the "Eliminator" chuck which is round.) I rewrote the zzero program to work in the x&y direction.
I also thought about use copper tape on the corner of jigs but I haven't gotten around to that yet.
I also make guitars and furniture so there are lots of different jigs.

Rick Samish

[beacon14]

I also do the same thing. My z-zero plate is a piece of 1/8" x 3/4" aluminum bar stock, and I can hold it against a fence, using a modified version of the z-zero routine to create an x-zero or y-zero routine. I edit the value in the program depending on the diameter of the bit , and allow for the 1/8" thickness of the aluminum. (I use mostly spiral bits, so there is always part of the bit at it's outer diameter to make contact.)
I use this process whenever I am placing parts against a fence.
For general re-zeroing of my x & y axis, I have simple homemade contact switches made of a carriage bolt, spring, and double-nut. When the x or y carriage contacts the carriage bolt head, it does the same thing as when using the z-zero plate. I have a routine that starts near the home (0,0) position, zero's both axes, then returns to home. I do that before running any kind of program or if I'm in any doubt as to positioning accuracy. My homemade switches (which cost about a buck apiece) have proven to be very reliable and accurate.