Tool to Square Z-Axis

[richards]

This is a dial indicator tool that I use to square the Z-Axis on my Alpha.


3245

As you can almost see, the left quadrant of my machine is 0.001" out of square (however, the top quadrant is 0.004" out of square).

Using a dial indicator and a piece of plate glass makes squaring up the Z-Axis a five minute job. Since I'm not using shims, I usually settle on having the axis +/- 0.003" anywhere on the rotation, meaning that any two points could be as much as 0.006" out of square with each other. Perhaps if I tinkered a little with the machine and added some adjustment bolts, I might be able to get closer than that, but I think being within 1/150th of an inch isn't too bad, considering the low quality work that I do.

[gerald_d]

Mike, your "squareness" is way better than 1/150th of an inch.....unless your cutters are the same diameter as your trammel.

[richards]

Gerald, I like to use big cutters to get lots of kerf - 10-inches if possible - since most of my work consists of turning expensive wood into sawdust. The more serious question would probably be, if I'm really setting tram to about 0.0002 inch (when using a 0.375-inch cutter), why can't I hold 0.0002" tolerance on all of my parts? (For those who think I'm being serious, I assure you than my tongue is firmly planted in my cheek. Since adding a 3:1 belt-driven 'transmission' for my Alpha, I've been getting perfectly acceptable parts.)

[harryball]

I find that being 1/150th of an inch out over a 10" area to be unacceptable. I'm setting up a laser alignment system to eliminate all error whatsoever, perfect alignment. However, technology on your planet is not up to the challenge so I've contacted my home world...

or was that a dream I had last night?

Bats are very picky... so I build to very tight specs. :-)

Robert

[gerald_d]

Mike, I was playing a bit of devil's advocate with that comment above...showing that this type of "squareness" is diameter-dependent. But, surely squareness should be independent of diameter?

(Remember the argument in April last year? link (http://www.talkshopbot.com/forum/show.cgi?tpc=7&post=23066#POST23066))

The trammel method is indicating real squareness only if your spindle axis is perfectly parallel to the z-slide motion. With my Fimec spindle I discovered (when fitting a DIY fan) that the shaft & housing where not concentric with each other at the back end - ie. the shaft was at a slight angle to the housing. (won't be buying Fimec again)

Anyway, a true test of z-slide squareness would be to flip the dial gauge to a horizontal position and then plunge its tip along a reference square standing on the glass.

[richards]

Gerald,
That's not a problem. I'll just shim the square (or tilt the glass) as necessary until the dial indicator shows no deviation.

I feel your fustration with your spindle. Although I haven't checked my Colombo, I'd be surprised if everything were absolutely plumb and square. Sometimes it seems like the advertising department and the engineering department of tooling companies have severed all communication.

[gerald_d]

Don't you just put a folded napkin under the short leg of the table?

That non-concentricity of the Fimec wasn't the main reason for deciding to give them a miss in the future. Have since found another supplier who appears to be more co-operative.

[trakwebster]

This looks like a swell jig.

May I ask where you got the parts to assemble this? And what parts are they?

(Or even, if you feel generous with your time, a photo of the back side to see how it's assembled?)

[richards]

Arthur,
The jig is parts and pieces of various dial indicator magnetic bases that I've aquired. After checking several of the bases, I realised that I could take one part from one base and add it a part from another base until I had what I wanted. The only part that I had to make was the 1/2' shaft that is held by the spindle. I had to turn part of the shaft down on my little mini-lathe to 0.4775 to fit the clamp of the magnetic base. The dial indicator is simply held by a standard screw/clamp that is part of most magnetic bases. The horizontal arm is part of another base and the vertical shaft that fits the spindle's collet is the piece that I turned on my lathe. With a little careful shopping, you'll be able to assemble the same thing for about $75 to $100, depending on your choice of dial indicators and magnetic bases. Although it may not be appropriate on this forum, I imagine that I could supply the same device that is in the photograph for $80 plus shipping, if you sent me an email indicating your interest. (NOTE to the forum administrator: If this is not appropriate, please edit this post to delete that last sentence and this note sentence.)

Gerald,
A folded napkin would work just fine - thanks for the idea. Don't you find it just a little strange that machines capable of precision work have so few precision adjustments? If tramming in the Z-axis was important to the use of the machine, wouldn't there be adjustments to perfectly align one side of the Z-rails to be perpendicular to the Y-axis and to the X-axis? Then shouldn't there be an adjustment for the other Z-rails to be parallel to the master Z-rail? And finally, shouldn't it be expected that there should be adjustments on the spindle/router mount to make the shaft parallel to the Z-axis and perpendicular to the X and Y axes? Loosening four bolts and tapping with a dead blow mallet seems to be a very crude way of aligning a precision machine - although it does seem to work fairly well.

[gerald_d]

Ron Brown and one other mentioned setting their towers on adjusting screws (Ron even posted pics). As you say, the machine mostly works okay without having adjusters, so why go to the expense of designing/installing/producing them? Anyway, for cutting thinnish boards, a perfectly square z is hardly needed.

[terryd]

Mike, How do you read the dial indicator when its spinning? and how many rpm's do you set it for? Never be afraid to state the obvious....... ;-). Hehehehe!!! sorry just couldn't help myself..

Terry

[gerald_d]

With strobe light when it is spinning, and a mirror when it faces the other way. Set rpm's to strobe frequency. Isn't that obvious?

[richards]

So, Terry, you caught the motion blur in the photo. Good eyes. It's a good thing that I have a spindle with totally adjustable speeds.

Gerald, the shutter in my Nikon only syncs to 1/120th of a second, which is the limiting factor governing the shaft RPM and the bank of strobes. I haven't gotten around the problem caused by light fall-off in the mirrored sections (inverse square light fall-off law), so the photos of the mirrored sections are much too dark. I quess I could attach a servo to the aperature ring on the camera, but I think that the ID3004 servo that I have on the test bench might be a little awkward to attach (10-1/2 inch long, 10-15 lbs.).

Of course the only two things turned on in the shop is the vacuum, to pull the mdf platten/spoil board tightly against the support boards, and the lights. The spindle shaft is turned by hand with a few seconds pause at each position, with my hands off the table top, to let the needle totally settle.

If you want to see something interesting, just give any axis a light thump with your finger tip as you watch the dial indicator. You'll be surprised how much movement you'll see. It makes me wonder just how much jitter there is in each cut caused by the vibrations caused from a spinning spindle, cutter chopping through the wood, V-wheels chattering along the V-rails, steppers clogging through their steps, etc. Has anyone ever proved that this Shopbot machine actually works?

[harryball]

Shopbot machines are only experimental based on theory. No real practical application is known. Fortunately for me, I'm only a theory too so we get along great.

I did hear once that some guy tried to make a sign with one, can't imagine what he was thinking.

Robert

[mikejohn]

Mike
Does the inverse square law apply to circular motion?
Maybe your problem is more with the Gurney Mott theory of the formation of the latent image.

.........Mike

[richards]

Mike J.

You've brought up an excellent point - one worthy of more discussuion. Honestly, I hadn't even considered the Gurney-Mott theory, which is:

THE GURNEY-MOTT THEORY

When a photon of light of energy greater than a certain minimum value (that is, of wavelength less than a certain maximum) is absorbed in a silver bromide crystal, it releases an electron from a bromide (Br-) ion. The ion, having lost its excess negative charge, is changed to a bromine atom. The liberated electron is free to wander about the crystal (see the figure below). As it does, it may encounter a latent image site and be "trapped" there, giving the latent-image site a negative electrical charge. This first stage of latent-image formation--involving as it does transfer of electrical charges by means of moving electrons--is the electronic conduction stage.

What do you think, is it even applicable in digital photography, and, if so, is it more applicable than the inverse square fall-off law?

[harryball]

Of course you realize if you go to this level you won't be able to be in the room watching the bot during the alignment process. The Heisenberg Uncertainty Principle states the act of observing a shopbot during alignment will change the alignment of the system... or is that you can't know the alignment and the XYZ position at the same time... hmmm.... perhaps you could construct a Heisenberg compensator and install it first.

Robert

[richards]

Robert,

Now we're getting somewhere. The Heisenbery Uncertainty Principle is defined as:

In quantum physics, the Heisenberg uncertainty principle or the Heisenberg indeterminacy principle — the latter name given to it by Niels Bohr — states that one cannot measure values (with arbitrary precision) of certain conjugate quantities, which are pairs of observables of a single elementary particle. The most familiar of these pairs is the position and momentum.

I'm beginning to finally understand why no machine can ever be expected to work.

[3d_danny]

Here is a ready made spindle square.

http://www.penntoolco.com/catalog/products/products.cfm?categoryID=6153

not a bad deal for $139


3246

[mountie1808]

I'm really in the wrong ballpark now. How are you able to make such fine adjustments???? I've only been working on the bot since march but am taking it upon myself to get this thing to really make some money. I'm having to tweak things, even resetting the router to the top of the unistrust for added clearance. What I haven't been able to do effectively is perfectly level it. I still get a slight (a wee bit more than slight) downhill edge on the cut with a 1/2" endroute bit. The only method that I know if is to use a bullet level and tinker back and forth hoping to fall into place but have yet to get it to disappear. Are there fine adjustment screws somewhere to do exactly what I need to do and I'm just missing them??

Mike