Accurate threaded rod help needed

I'd go with this one. You've got the accuracy, easy intuitive measurement, and resolution down to the tenths if you use one with the .0001 scale on it. A micrometer or threaded rod would require you to turn that after moving the stage, where the caliper could be mechanically linked to the stage as it moves, so you are measuring it directly without the risk of having an inconsistant error.

Even if you don't couple it to that which you're measuring, and choose to move it manually, you won't be adding any new stresses to the system, and you'll come up to that which you're measuring from a consistant direction.

Digital calipers could be another option along the same lines here.

You cannot make a leadscrew using only a die that has the accuracy you need. How much money to you have to spend on this part of the project?

Jim

Ok, so I use a dial caliper. But how do I know how accurate it really is? Granted I can get something with resolution to .0001" or .0005" but I need to know how accurate it is. Does it lose or gain .0001" over an inch? Or over 3"? That makes a difference. I need to know that when I set the stage to 2.76" from 0" that it really is 2.76" from 0 and not 2.78 or 2.75. When I'm measuring a mirror I have to be accurate in the x measurement

I know I'm being picky but I don't know that I can trust the caliper to measure accurately. How do I test the caliper?

Thanks,

Ed

Sure he can. Start with the die threaded rod, then single point die stock set at 90, then 120, then 180, then 270 degrees. Then spend a week lapping the result with a long lap (ie the nut is as long as the screw). Turn end to end often when lapping. Then measure the result, and correct (by more lapping, but with a shorter lap and selective pressure).

Not easy but definitely possible. Rowland (scientist behind the first ruling engine capable of producing really good difraction gratings) undoubtedly started with a single point turned leadscrew but the difference between a turned screw and one produced with a diestock isn't very relevant when the target is something accurate to 1 or 2 microinch per foot. Long, long, long, way to go either way. OP isn't looking for anything that good, of course, so effort drops from months to days, I would guess.

Regards,

Adam Smith, Midland ON

Find someone with a set of Jo Blocks, or (less good) compare measurement from it vs. the measurement of an object from a known good micrometer.

Right. The need for accuracy is another reason _not_ to go home-made, at least not within the ability of my machines' accuracy.

By measuring something you know the size of, or having the caliper in question tested by a metrology lab.

Dave Hinz

Can you give us a bit more info about this requirment? You don't want to go spending all kinds of time and money on a system you can't use (or is too capable, thus resource-thursty).

Consider the following:

Lets say your gage is good to +/-0.001" Also, your part is allowed to deviate from nominal by +/-0.001"

Now, lets say your gage typically reads too small by 0.0009" (within the gage's tolerance). Further, the part you're measuring is too big, 0.0019" over nominal.

Now your gage reads +0.001" making you believe your part is within tolerance, which it is not.

The rule of thumb for gaging is to have your gage 2-5 times more accurate than the tolerance of the feature being gaged. Sometimes up to 10 times (or more).

Now, your process may not require +/-0.001" accuracy (which calls for a gage that can read within between 0.0005" and 0.0001") but you do need to understand the requirements of your process before you start to order parts.

Regards,

Robin

Let me check my Stuff. Ive got a 55 gallon drum filled with microscopes, stages and illuminators. Mostly Polaroid type scopes but Im sure there are some stages and Stuff with precision screws in them.

Ever see a barrel full of Leitz and Leica optics?

Ill check this weekend

Gunner

Read my post. I said "with only the die."

Unless he has some method of measuring as you say, he cannot do it with just the die.

Jim

Ed,

Having fabricated over a dozen telescope mirrors up to 16" F/4.5, I wonder what test you are implementing that needs that kind of accuracy, over that distance?

But, to your question, I once tested a 1/4-20 threaded rod from the local hardware store. It showed a period error of +/- 0.0015 every 0.05" of travel (one complete revolution of the threaded rod).

Over 0.4" of travel I measured a cumulative error of only

0.005", which is most likely an angular error in my test setup.

Here's links to the data, if it helps any:

Excel file with raw data)

image of test setup)

Now, back to optical testing, the only test I can think of that needs the type of precision you prescribed, would be an Offner Null setup, which I know of no amateurs, nor commercial opticians, using for common telescope mirrors.

Here's a short list of common tests used by opticians to asses optical mirrors, starting with the simplest. #1 Ronchi test A qualitative assessment of the optic, usually with no measurements need at all. Only time measurements might be used with Ronchi is if your comparing what you see verses what a computer generated simulation suggests what you 'should' see.

#2 Foucault Test Both a qualitative and quantative assessment of an optic. While there is a need for mechanical precision, that precision on even the largest optics, is generally limited to less than 1 inch of total travel. For instance, a 16" diameter F/4.5 parabolic mirror, tested with a fixed light source Foucault platform, only needs accuracy over a 0.4" range of travel. Staying with the 16" F/4.5 example, and assuming a measurement error standard deviation of 0.01" per measurement at 8 positions (zones) on the face of the mirror, would still result in measuring a good mirror from a bad mirror. IE, if the optic was perfect to start with, and each of the 8 zonal measurements had a measurement error standard deviation of 0.01" would result in showing a perfect mirror as having a surface error of 12 nanometers (or 1/22 rms wavefront error)

#3 Fringe traced Interferometry Myself, and many others, have built and use interferometers for testing optical mirrors. A Bath interferometer is the simplest to fabricate, and can be fabricated from common and inexpensive surplus optics.

Once you have a working interferometer, a common tape measure will suffice for any measurement needs while testing an optic. With the recent development of several free software packages for reducing the images produced from the interferometer, optical interferometry is now a viable option for amateurs.

If an amateur accurately implements any two of the three tests above, assessment of an optical mirror is nearly trivial. The difficult part is implementing and interpreting the tests, combined with the art of actually producing the finished optical surface.

Take Care, James Lerch

(My telescope construction, Testing, and Coating site)

Press on: nothing in the world can take the place of perseverance. Talent will not; nothing is more common than unsuccessful men with talent. Genius will not; unrewarded genius is almost a proverb. Education will not; the world is full of educated derelicts. Persistence and determination alone are omnipotent. Calvin Coolidge

A couple of ideas. One is get a Harbor Freight digital caliper. A friend got one and managed to get someone at a Navy Calibration lab check it. It was within .001 over the full six inch range.

Second you could get a dial indicator with a one inch travel and use a one two three block to extend the range to you three inch needed range.

Those would be inexpensive and give you a way to read your position. If your need is to set the x stage to a given position, I would think about something more like a depth micrometer used as an adjustable stop. Or if speed is not a big need, then a inexpensive set of gage blocks, used to set a stop.

Dan

Ok. I thought the point you were making was that he'd need to single point turn the thread to get something accurate enough to use. If what you were saying was that being able to measure the screw is fundamental, I certainly can't disagree with that.

Adam

Jeebus! 8-O We need to put a sign up at your place, Gunner:

"Gunner's Stuff - AKA Apex Surplus North" ;-P

I can walk into that place and lose two or three hours real easy. (Hey, it was daylight when I went in...)

-->--

Univesal thread grinding company makes very accurate lead screws, cost $$$

Jack

[ ... ]

Hmm ... there are dial indicators which reach to at least 5" travel and measure to 0.001".

Does what you are doing on the stage generate small chips? If so, they are likely to get into the rack gear that runs the dial, and cause it to skip several teeth. They can be cleaned out, but that would be a royal pain.

That is simple -- the lead of a die-cut thread is *not* that accurate. Better to cut it on a lathe, and better to go for a much finder thread pitch -- say something like 40 TPI (what inch micrometers typically have). But you will still need something to verify that accuracy once it is made and set up.

Lathe to start with. Fine pitch. Note that brass will wear rather fast. Better to use bronze against steel. (And, there are tricky ways to mount the nut so it will turn as it moves along a slide to correct small errors in the lead of the screw. I've seen such things in Gaertner "traveling microscopes" -- which may even have the travel which you need. (A quick eBay search has the only thing which comes up under "traveling microscope" is far from what I was talking about. The one which I was talking about was designed to fit on an optical bench (like a lathe bed), with a vernier reading elevation, and the cross-slide reading microscope lateral position to 0.001 mm (a *lot* more accurate than you expect to need).

Find a micrometer spindle which has 2" travel (or even 1" travel) and make provisions to put a 1.000" or 2.000" spacer block between the stage and the micrometer spindle to extend the range, with a spring holding the stage against the micrometer spindle. (This is assuming low forces acting on the stage.

Or -- get a digital caliper and mount it in there. (Even try one of the digital readouts which will mount on the front of a mill to measure spindle travel. It is like a digital caliper with no jaws, and will probably be easier to mount to measure the stage travel.

Lots of possibilities.

Good Luck, DoN.

Hmm ... probably not the best choice. It is low friction, and very repeatable, but it is a coarser pitch than a good micrometer leadscrew, so where you get 0.025" per rotation of the typical micrometer, you would get perhaps 0.100" to 0.200" per rotation for the ballscrew, thus requiring a more precise calibration of the thimble. And it is easy to drive a ball screw backwards, so if whatever is being done applies any force to the workpiece, it may shift the readings, unless yo have a way to lock the ballscrew against rotation. (Normally, a CNC machine tool, which does use the ballscrews, has either a servo motor or a stepper motor holding the ballscrew against turning under load.

Enjoy, DoN.

Ive been thinning it down a bit. I can actually see dirt now.

When I go to Apex I go early..and bring both breakfast, lunch and a compass

Gunner

"Pax Americana is a philosophy. Hardly an empire. Making sure other people play nice and dont kill each other (and us) off in job lots is hardly empire building, particularly when you give them self determination under "play nice" rules.

Think of it as having your older brother knock the shit out of you for torturing the cat." Gunner