How to Adjust and Calibrate an Inside Micrometer

An inside micrometer is primarily used to measure, as the name implies, inside diameters. They can also be used for a variety of other types of measurement, in particular with specialize accessories, like a height gauge adapter. We'll look at the construction and adjustment of several types of rod-type inside mic.

There are a lot of variations on what is, in essence, a stick. The photo shows a fairly common style: the interchangeable solid rod type. It has the key features of most inside micrometers, those being a micrometer screw inside to adjust the length between the precisely positioned measuring faces. The faces are spherical segments and should contact the inside surface being measured on an exact diameter to get an accurate measure. There are other styles, but this is the most common general purpose style, as it can cover a wide range (one set I use regularly covers 38mm to 2500mm) at reasonable cost.

Once properly calibrated, as long as there is no damage, there isn't significant wear, and the adjustment isn't messed with, the measurement can be directly read from the micrometer. There are sources that claim that you can not trust the reading, but if the tool is properly set and calibrated, the reading will be correct to the precision and uncertainty of the calibration. Wear can be an issue, as, by design, there is very small contact area when making a measurement.

First, definitions:

Calibration of the tool is where it is compared against a standard, error quantified, and uncertainty determined. There is often a permissible limit for the error, and if this limit is not met, the tool must be adjusted.

Adjustment is when a property of the tool is changed to reduce error

The uncertainty is primarily a result of the process used for calibration and the uncertainties of the references used to do the calibration, as well as the inherent uncertainty of the tool being calibrated.

The best way to calibrate an inside mic is with a calibration ring (Ring Gauge). These often cost more than the micrometer, and for a wide range set, you need to have a lot of them, and as the size gets much beyond 300mm, ring gauge are not generally available. Most small shops don't keep a range on hand, but either use alternative methods or send the tools out to a calibration specialist. For many, many applications, alternative methods are sufficient. For most hobbyists, alternate methods are the only realistic option.

So lets look at a few.

An inside micrometer

A reference suitable for the size to be calibrated

Lets start with one of the most common styles: the interchangeable solid rod type. These are made by a number of manufacturers, more, and less, reputable, with two top-line makes today being Mitutoyo and Starrett. The same general design has been made for well over 100 years by a variety of makers. For the purposes here, I will only show Imperial-unit tools (inches) for consistency, though most current types are available in SI units (mm) as well.

We'll start with the Starrett 124A, which covers the range from 2" to 8". The rods are in 1" increments, the head has a range of 0.500 inch, so there is a spacer provided for the upper part of each rod range. The 2-3" rod is used from 2.000 to 2.500" without the spacer, and 2.500 to 3.000 with it. As shown in the case, the mic is set up for 2.500 to 3.000", as the spacer is in place.

The larger version of this micrometer (the 124C) has three spacers, a 1" and two 2", and they may be used in combination to give each rod a 6" range with a 1" travel head.

Second-hand tools are often missing the spacer. They can be purchased from the manufacturers for name brand micrometers. If the tool will be used in the field, or shared in a shop by several people, having a spare spacer can save a lot of trouble.

The range is set by inserting the proper range rod, with spacer if needed, into the head, being sure that the alignment marks are match, then snugging the retaining screw. This is one reason calibration can be an issue. Since the rods interchange, each one must be properly adjusted and calibrated with the head it will be used with. DO NOT interchange parts between sets, as the heads may not be exactly the same.

Construction is simple, and most makes follow similar lines. The graduated thimble has a contact face at the end, and the micrometer screw inside. This threads into the body, which has the hole for the rods, the index line, and the major graduations (0.025" in this case), as well as the retaining screw and a short handle. There is a collar inside that can be used to set the friction: inside mic's generally are set for more thread friction than caliper (outside) micrometers, as they are set by feel by being swept or rocked against the part, and are often used in places where a lock is awkward or impossible to use.

Most mic's of this type adjust using a threaded contact point. These have wrench flats, and the micrometer set came with the appropriate wrench, but any appropriately sized small wrench, such as from an ignition service kit, works. The threads are tight. Some makes have locknuts, as well.

It can be difficult to hold the rod while making the adjustment, so if adjustment is needed, hold the rod by the knurled portion. The larger diameter and knurling require much less grip force than the smooth sections of the rod, and reduce the risk of damaging the rod (bending, scoring, etc) This particular Starrett set changes by about 0.018"/turn (56 threads/inch)

Now the question: How do we know if it needs adjustment?

The first method, often shown in textbooks, is to use an outside micrometer. This requires a calibrated outside micrometer with a range that overlaps the inside mic range being calibrated. Here, I used a 2-3" outside micrometer to calibrate the 2-3" rod.

First, check the outside mic using a standard. Then, there are two options. I set the inside mic to exactly 0 (the 2" point in the rod range) and measured it with the outside micrometer. The other option is set the outside mic, and adjust the inside mic. This need not be done at the minimum measure, but can be done at any point in the range.

In this case, the inside mic is about 0.0005 undersize. This is actually in spec, but I did adjust it to get it closer. If it had been significantly oversize, I would have double checked the seat of the rod to the head and rechecked against the outside mic before adjusting the rod.

Next, I rechecked with the 0.500" spacer, and it matched, indicating that the spacer is in spec.

At this point, I ran the inside mic over its range and checked several points to be sure the screw isn't unevenly worn, and it was ok over the full range. This should be done at intervals of about 0.105" so as to check different rotation points and detect certain types of thread non-unifority, and only needs to be done for one rod.

Now, do the other rods the same way, and the set is calibrated.

It is good practice to do calibrations using a reference standard that is at least four times, preferably ten times, better than the tool being calibrated. Better means both in accuracy (nominal actual value) and uncertainty (the bounds of how far the true value is from nominal).

The details are a specialty, but as a general guide, each transfer increases the uncertainty. In this case, two transfers occurred: the standard rod to the outside mic, and the outside mic to the inside mic.

A slight practical improvement comes by setting the outside mic to the standard and locking it. This is still a transfer of the measure, with an associated uncertainty, but it is likely to be less uncertainty than by measuring with the outside micrometer.How much less? As I said, it is a specialty in itself.

In this case, the outside mic used does not have a lock, so it wasn't an option.

If you want to learn more about measurement and uncertainty, see Bell: _The Beginners Guide to Uncertainty of Measurement_ (

https://web.archive.org/web/20210505184753/https://www.npl.co.uk/special-pages/guides/gpg11_uncertainty.pdf ) provided by NPL, the UK National Physical Laboratory.

For longer rods, there may not be an outside micrometer available. We can look at a couple other options that can be satisfactory for many applications.

First, many shops will have a conventional height gauge for use on a surface plate. They are a reasonably inexpensive way to get a range up to 18" (450mm) at good precision and accuracy, though they will have much greater uncertainty than an outside micrometer when used for direct measurement (this is due to Abbe's principal, a geometric consideration). They can be used as comparison against a standard gauge rod which may reduce uncertainty.

As shown, the inside micrometer is held vertically and the conventional height gauge scribe is swung over it. This is very sensitive to contact, and changes much less than 0.001" (0.025mm) are easy to detect. When the bottom of the scribe block is above the inside mic, it, obviously, clears it. When it very slightly less high, it will tend to grab the top of the inside mic by friction and cause the inside mic to rock.

Many things can be used to hold the inside mic vertical, such as a machinist square, a good vee block, and so on, but it can also be swept by hand, as will be shown later.

Using the same height gauge method, a fixed range Brown and Sharpe inside mic is checked.

In this case, I knew it needed to be adjusted, since the tip was relapped due to wear. This style allows for that, though many types get the contact replaced when worn.

In this case, I set the height gauge to 9.025" using a gauge block stack and rocked the inside mic under the scribe block. Why the extra 0.025"? So I have adjustment on the thimble both ways. The contact is roughly set by eye, then the mic is used to measure the height of the scribe block. The reading tells me how much to adjust the contact point, and which way. Two iterations is generally sufficient to achieve better than0.001".

The hard part with this style is setting the locknut without disturbing the contact setting. The locknut should be left barely snug at all times to take up thread play, and the final tightening is good snug, not lugnut tight. If needed a drop thread retainer (or organic solvent paint, or nail polish) can be used to prevent movement after it is set. This will also indicate if the setting has been tampered with later.

More expensive than a height gauge is a high quality, large range caliper. Calipers are generally less accurate, less repeatable, and less precise than micrometers (Really. Read the manufacturer literature, and read the standards) though they may be graduated, or report a digital value, with much greater precision than they are actually capable of. But if it is all that is available, or, better, if there are gauge rods available, then it may be the best option.

Why do I consider a height gauge preferable? It sits on the surface plate on its own. Calipers can be awkward to handle, and when both hands are needed for the inside mic being calibrated, one needs to get creative.

For the (fixed range) inside micrometer, shown here, a 24" master vernier is used. The caliper is supported on shims by the fixed jaw and the beam so that the middle of the measuring faces is the same height as the axis of the micrometer. This mic sits naturally with its axis parallel to the surface under it on the heat shield. Gravity does most of the alignment here, leaving hands free to manipulate the inside micrometer.

How good can one do this way? Well, that depends. In this case, the caliper was set using a 10.0000" standard, it has well lapped, parallel faces, and the inside mic was matched to it, so the inside mic is almost certainly within 0.0005". If the caliper scale was used as the reference, maybe +/-0.002". No, a digital caliper won't generally improve on that, though it may show more digits.

Tubular style inside micrometers (not to be confused with tubular frame outside micrometers pioneered by Tumico) have been around for a long time. Like the solid rod types, many manufacturers make them, and, in many cases, unlike the other styles, parts are sometimes compatible from one make to another. Like the solid rod types, parts should not be interchanged between sets.

The tubular style is usually "double ended". The extensions go to both sides. The base length, 1.500" for the Lufkin 680A set shown, uses two `caps'. One or both of these are replaced with extensions for larger ranges. For example, for 2-2.500", one cap is removed and replaced with the 1/2" rod. The wrench allows you to grip the caps and the rods, and if needed, the handle can also be used. The parts should be assembled snugly, but if you need pliers, they are WAY too tight. Note that the rod length designations give the difference between the base length with a cap and the length with the rod, so the 1.500" to 2.000" base length, plus the 1/2" rod, plus a 2" rod on the other end gives 4.000" to 4.500" range.

It is critical that the caps be the same length. Measure them carefully before beginning calibration. If they are not the same, pick the less worn one and do not use the other, and see the note below. You will lose the base range (1.500-2.000 in this case), but will be able to set the other rods. Even if the caps are the same, it is good practice to set the other rods using only one cap, then recheck with the other cap, and to fit the parts both ways (cap on thimble, and cap on body) to insure that there are no defects that will haunt you later.

The reference surfaces for the caps and rods are the ends, and the threads inside are relieved slightly at the end to provide a clean seat, but ANY nick, ding, burr, or blemish, on or near the end, can affect how the parts seat. Unlike the solid rod types where small burrs may drop into a groove in the seat, on the tubular types, the extensions will not seat correctly.

The tool is adjusted and calibrated at the base length first by setting the graduated edge of the thimble. The caps are solid, but the rods are individually adjustable, so this must be done first.

The technique is shown using a similar micrometer that, frankly, photographs better.

After the base length is calibrated using the caps, the rods are done. The rods adjust similarly to the solid rods of the Starrett 124A- the contact end is threaded into the tube and has flats for a wrench.

There is a confounding issue if the caps do not match, or a cap is missing: since the rods are used either with the cap or in pairs, everything can be fine with one rod and a cap, but two rods are out of calibration. Direct measure of the rod lengths against the cap length lets the base length to be properly set, though. The 1/2" rod should be exactly 1/2" longer then the cap. The 1" rod, exactly 1" longer, and so on. If the cap is 0.247", rather than 0.250", long, then the 1/2" rod should be set to 0.747". Once ONE rod is properly set relative to the cap, that rod and the cap can be used to set the base length, and the cap and head can be used to set the other rods.

The final calibration has more theoretical uncertainty than the solid rod type, as there are at least two adjustable parts, separately adjusted, in most measurements. In practice, it isn't likely to be an issue, but it is good to be aware that the errors and uncertainties stack.

Here, we see a Lufkin 681 head, with a base length of 4". It is being set using gauge blocks. The 4" (4.000015", actually) block has a wear block wrung to each end, overhanging enough for the 681 to measure the space. The OD of the 681 is against the 4" block's side so that here is one degree of freedom to swing the micrometer, making it easy to feel for the proper adjustment.

It is adjusted like the smaller tubular types.

The gauge block clamp is not needed, but makes it a little easier to handle, and a lot easier to photograph. If you have hollow blocks available, a tie rod does the same job. Accessory blocks that act like caliper jaws are available to fit the ends, as well.

Lets look at a few other designs.

First, a Millers Falls from early 20'th century. The 1218 is a 2 to 12" range, but does it with only seven rods. The secret? Two spacers: a 0.500" and a 1.000". They are used singly, giving each rod a 1.5" range.

All but the shortest rod are adjustable which means that, unlike the Starrett 124, the contact tip on the thimble is adjustable for the base length. All of the adjustments are secured with locknuts, rather than friction.

The rods are held in by collet chuck style. Be sure that the rods properly seat, and that there is no crud between the fingers of the collet.

Start with the shortest (2-3.500") rod and set the base length, then move through the other rods. Don't forget to check the spacers.

The Starrett 120- not the dial caliper; they reused the number- is really different. It is set by aligning reference lines on the rod to the reference line on the head. Starrett still uses the design in the 121 long-range mic, and several other manufacturers have used the same design over the years. Sawyer's version (the 142, circa 1904) was advertised as for "Close internal measurements by thousandths of an inch, where a definite distance in inches is not essential".

This is the first one we have seen here that MUST be calibrated at each use. The graduations give approximate setting, but only relative measure is accurate without calibration.

This is the style people will point to when pressed, should they choose to die on the "inside micrometers can't be trusted" hill.

The Sawyer 144 has threaded extensions that are stacked to make length. These are not uncommon on the used market, and similar concepts are still used.

If you have one, the base adjustment is done with the thimble end contact. The extensions are individually adjustable. The two cross holes are for tommy-bars to adjust the lengths.

There were several other Sawyer models made, as well, including one with a head that telescoped to allow measuring deep internal grooves easily. They used free rods like the Starrett 120, as well as extensions to increase range. Like the Starrett 120, they should be calibrated at each use.

The international Tool Co, San Leandro, CA, made an inside micrometer. I have no idea if they made anything else.

This is a pretty neat design, as the rods positively position for multiple ranges without the need for spacers. A tapered pin locates the rods. Quite a clever method, but it does seem to be subject to contamination.

Adjustment is by the contact tips, and similar to others seen here.