How to square the Z

[harryball]

I don't know what you guys do, but this worked great. I modified my wrench and placed it on the shaft of the spindle. I then used zZero to set the spindle to 0. For me this meant when Z = .123" the wrench was not touching the zeroing plate, when it was Z = .122" it was touching. What do you know, my zeroing plate measured to .123".

I could then swing the wrench to the other side 180 degrees and set Z = .123" and see where I was at. If I touched the plate it was too low, if I didn't I could adjust and find where it did to know how much to adjust.

There are photos later that will help this make sense.

I set the spindle out on the table so I had enough clearance for the wrench to spin and still be on the table.

I set the wrench end off to the X=0 side of Y and zero'd. This now meant that at Z=.123 then zZero input was off, at .122 it was on.

I swung the wrench 180 degrees and set Z to .123. I had a large gap under the wrench. I had to set Z to .100 to touch. So it was too low.

I raised the low side and repeated the process, Zero Z, swing, check and so on.

In about 10 minutes I had both sides at .123". I did the same for the X axis, double checked everything and that was that.

My Y axis is less than .001 off square, my X axis is less than .003 off square... and that's out at the length of the wrench!

I used my second wrench to weight down the zero plate, no other function. Yes, I know I need to resurface my spoil board, that'll come later.

Here are some photos to help this make sense...

My wrench mod is a bolt added to the end eye.

3149
I was stunned, .123"

3150
All nice and zero'd

3151
What the PC shows, touching...

3152
not touching, like magic.

3153
now at the other side not so close

3154
What my dog was doing all this time

3155

Hope that makes sense.
Robert

[gerald_d]

These gents used dial gauges:
Ron Brown (http://www.talkshopbot.com/forum/show.cgi?tpc=7&post=6856#POST6856)
Mike Klafehn (http://www.talkshopbot.com/forum/show.cgi?tpc=7&post=32607#POST32607)
Dirk Hazeleger (http://www.talkshopbot.com/forum/show.cgi?tpc=7&post=23055#POST23055)

[scottcox]

Dog: Great Idea. You're an inspiration!

Robert: Your idea was OK too. ;-)

Does the wrench fit tight into the grooved flats of your spindle shaft? Mine has all sorts of vertical play (wide slot/thin wrench)so it's somewhat loose. How are you keeping it tight at your spindle while testing?

[harryball]

I will inform the dog of your approval... if he wakes up long enough. I think he's about 70 now in dog years.

I'm only keeping the wrench in place by gravity. I found by trial and error that it was not neccessary to secure it provided I didn't knock it around while working. I even removed it and hung it back and get the same readings. I just make sure it is fully engaged and hanging each time.

It's also important for the wrench to be free so when the spindle descends it doesn't push on the table surface. If it's touches the table the wrench just floats up.

When I place the zero plate on the spoil board I rub it in good, slide it around to make sure it's flat. Also, make sure your table is grounded well, you are completing a circuit and while it should be that way for zeroing Z anyway... you're going to be doing this over and over increasing the chance of a static charge sneaking in.

Robert

[trakwebster]

I was a little less scientific, but got pretty good results ...

I surfaced my spoiler board using the shopbot-supplied 1.25" flat end mill, and sure enough there were some ridges, showing me it was off in two directions.

I handled each direction separately, but using the same method.

For example, for the Y-direction, I taped a small very flat metal ruler onto the gantry so that it's end just abutted the Z-carriage, and on this z carriage I placed tape and marked where the top of this ruler touched.

From the hardware store I then got some strong stranded wire, cable-ties, a turnbuckle, and a couple of s-hooks. I connected an s-hook on one end of the wire, and on the other end attached the turnbuckle and the second s-hook.

I connected this 'winch' between the carriage and the z-assembly on the side that needed to be pulled in order to correct the z-plane error, and then (a) loosened the z-assembly mount bolts and then (b) tightened the turnbuckle until the marked tape showed that the z-assembly had moved slightly in the desired direction. Then I retightened the z-assembly mount bolts.

I did the same thing in the X-direction.

It still is not 'perfect,' as I can see the trails that a 'clear-rectangle' leaves. However, for my purposes so far, it seems to be just fine.

Your method of calculating 'correctness' is much better than my eyeball/guesswork method. However, I'd like to know ... how did you go about the process of moving the z-assembly mounting by a small amount to get what you wanted?

[harryball]

When working with the Z along Y I used a bar clamp on the low side to support the Z. Then to raise the Z on that side made small turns of the handle.

For the Z in the X direction, I have a small bar clamp I drilled out and converted to a push bar to place some ceiling lap boards. I used it to rest the Z forward (toward the X=96 end) and then slowly turned the handle pushing the top of Z back toward X=0 until it was where I wanted it and tightened the two top bolts.

Robert

[trakwebster]

Bar clamp set to push! I wish I'd thought of that! My bar clamps were only six feet away, but I had to go to the hardware store!

[harryball]

I never look at things at face value... I figure this entire world is just full of parts I don't know where they go yet.

I used a 6" white corningware bowl as a lake drain stopper, 2 years and still holding back a 2 acre lake :-)

Robert