Precision hole drilling

Except for the fact that his encoders have no shaft to couple and no bearings. His code wheel will automatically be centered on the motor shaft, he just needs to mount the reader head within tolerance around the code wheel.

The center hole he is drilling is just clearance for the motor shaft and definitely does not need to be precision bored, it just has to not rub on the motor shaft.

The mounting holes for the reader head, and the holes to mount the adapter plate to the motor are the important ones, and all of those are small holes that don't need to be bored either. With the tapered centering tool that he'll either buy from US Dig or make, he can readily center the reader head around the motor shaft and then secure the head and mounting plate.

You misunderstand me. this is not for tolerance, its for finding your gross measuring errors. I agree, the printer isn't exact, but it will quickly spot a measurement error.

Karl

Correct.

The reason I want to locate all holes precisely is to align everything.

Thus the hole for the motor shaft is very close (0.06mm) to the motor shaft diameter and I want all other holes to be located precisely, with minimal slop.

The first plate seems to have worked out well.

I am afraid that if I allow slop in mounting of the reader head, then if I accidentally knock the encoder, it will become misaligned. The mount that I already made and used, has no "play" in head mounting, so an accidental bump would not change any alignment.

I did some sanity checks in the spreadsheet.

Have you heard of toolmakers' buttons? I haven't seen new ones for sale but they aren't hard to make, drill the center of some ground drill rod and part off a few rings.

To use them drill your mounting pattern to scribed or worn-leadscrew accuracy for an undersized tap, attach the rings with screws and washers and knock them into position as measured with calipers or a micrometer. Center each one under the spindle with an indicator and bore out the threads, then drill/bore/ream to finished size.

This means you have to buy a small enough boring bar. The standard tap was #5-40, which is 1/8" OD, and my B&S buttons are 0.200 in diameter.

I put a ground rod in the drill chuck and measure from it to a plug in other holes.

I haven't tried this but if you have X and Y zero references like a vise stop and the back vise jaw or a protruding parallel maybe you could locate the spindle with an edge finder and the depth rod of calipers. This would be a good use for vernier calipers that lie flat face down.

jsw

Somewhere in my dusty library, I have a book that takes your procedure one step further. It described methods for making tooling fixtures for manufacturing timepieces. X and Y were located by stacking Jo-blocks between stops and the tooling fixture.

The old guys were pretty clever.

I printed some CD labels a while back that the foreground image was not printing. But the circle of the CD outline did - about .010 wide.

Four passes on a cheap Lexmark X2550 - Four paper reloads on the same page.

Absolutely beautiful registration.

Perfectly round with no widening of the line anywhere.

But I'm not mirroring anything.

Why would I want to???

It may be one of Dick Moore's early books, which I think covered the subject. When I first inherited my lathe, I spent the first three or four months learning old toolmaking techniques, especially the use of the faceplate and toolmaker's buttons (I have four faceplates and two sets of Starrett toolmaker's buttons.) I even learned how to crush diamond bort and make my own internal grinding tools. It's interesting to learn but I don't recommend doing it.

I still find it to be the most interesting part of old-time machining. Around 1900, they could make master watch and clock plates (the ultimate drill jigs) to +/- 50 millionths accuracy -- in plain-bearing lathes.

Moore's Jig Borer was the next step in the process of applying technology to this fundamental problem. (NC milling was the next.) At that time (1900 -

Ignoramus3537 fired this volley in news:A-ydnVcH-7D4VqHRnZ2dnUVZ snipped-for-privacy@giganews.com:

So did I. But I perceived that the "config files" were designed for certain specific types of hardware, and one might have to re-compile at least the hardware interfaces if you were doing something full-custom.

I don't know if you're designing your own hardware, or using off-the-shelf stuff.

LLoyd

That's more work. You have to clamp two bars to the mill table or lathe faceplate and set them square to each other, then locate the hole nearest the corner under the spindle axis somehow. Maybe you center drill it and hold it in place with the mill or tailstock spindle while you clamp on the bars. Subsequent holes will be spaced by the accuracy of the Jo blocks, minus any user error or clamping distortion.

The blocks in my set aren't really long enough to guarantee that the work piece won't tilt.

I've used the watchmakers' disk method to drill locating pin holes for the wedges for a pie chuck. The method worked well enough but it's very difficult and tedious, involving turning three identical disks to better than 0.001" diameter accuracy and identifying when they barely make contact with one on a locating dowel pin in a previous hole and the other on a pin in the drill chuck. Of course I had to drill the hole without the disk in place to preserve its ID. The chuck locating pins they had to fit onto are 0.125" diameter and 1-5/32 (1.15625) apart. The three spacing disks I made measure 0.5814", 0.5815", and

jsw

Encoders without a shaft?????? Blink blink.....blink.....

Gunner

One could not be a successful Leftwinger without realizing that, in contrast to the popular conception supported by newspapers and mothers of Leftwingers, a goodly number of Leftwingers are not only narrow-minded and dull, but also just stupid. Gunner Asch

With a mill, I would more likely have used a 2-flute center cutting end mill and plunged through somewhat undersized (e.g. 5/8") then used a boring head to take it out to the desired diameter.

But does the hole actually have to be that precise a fit? All it needs to do is clear the shaft, I thought. Bore it a little oversized, and turn something to be a sliding fit over the shaft and with the right OD to center the encoder properly. You were talking about getting a centering bushing from the encoder vendors, but why not

However -- if you want a drill which does not walk the way a normal chisel-point drill bit does -- get one which has a split point -- which does a much better job following a staring hole made with a center punch of appropriate size or a center drill.

Again -- go for split point drills -- less wandering.

Especially if you make your own centering bushing using the lathe.

How precisely did you take the measurements? Assuming a digital caliper and clean untapped holes all the same size, take the spurs on the back of the caliper, place them in a single hole, and zero the caliper. Then move one spur to the next hole, and expand so you are measuring between two points as far apart as possible (the calipers will naturally seek this position). This will directly read the center distance quite accurately. Yes, the zero setting of the diameter of a single hole will be a little undersized, because the spurs don't have a fully sharp edge, so the chord takes up a bit -- but you have the same thing happening when measuring between two holes so the errors cancel.

Slightly oversized, with a centering bushing (again, made on your lathe) allows you to get the encoder properly centered.

Good Luck, DoN.

In particular -- the linear dimensions across the short dimension of the paper tend to be more non-linear because they are drawn with a rotating mirror and the laser -- trying to convert angular motion to linear motion -- (with optics which attempt to correct the linearity

-- but are not perfect). The other axis is defined by the paper being advanced by stepping switches, and might actually be more accurate if the advance rollers do not slip. (Some plotters actually have grit on the advance rollers which form a track in the back of the paper, so it moves the same distance for all the passes back and forth to draw the whole image. Overall, a good plotter is more accurate (at least to start with) though the paper can shrink or swell with humidity changes, so use it as soon after plotting as possible.

Enjoy, DoN.

On Tue, 13 Jul 2010 16:35:24 -0700, Jim Wilkins wrote: ...

Well, if they were supposed to be identical, why'd you make them with so much slop?

Better if you start with a spotting drill -- or even a center drill -- which are not long enough to flex much. But if you use a split point, you will still have very good centering compared to a standard chisel-point jobber's drill. Using screw-machine length drills (shorter) will also provide less flex room.

I keep a set of split-point screw machine length drills for frequent use, and a 1/2" split point drill (jobber's length) in the stand by the lathe for starting holes for boring. I've also got a set of 1/16th to 1/2" jobbers length drills with split points.

To show how little walking there is, I used the 1/16" bit to hand held electric drill in place a cross hole in a 1/4" shaft for a cotter key (to keep the track roller for the garage door from walking out of its brackets), and the hole was close enough to straight diametrical so I would not be ashamed to show it to anyone.

Enjoy, DoN.

They have to be more accurate than the resolution of your gage, or they're confusing. I put mine on an electronic gage at Wasino and they were within

The dealer warned me not to grind or lap on the hardened-bed lathe.

jsw

A sharp-pointed wiggler can help to get you centered on the scribe marks. Then swap it out for the drill bit.

Or -- an optical centering microscope, if you have (or make) one.

Enjoy, DoN.

Why not lap? All of the old-time lapping methods contain the grit pretty well. And you can cover the ways with aluminum foiled or oiled newspaper if you're edgy about it.

If you're hand-lapping, it doesn't get in the way of anything.