How to read a micrometer

[colin hamiltonParticipant @colinhamilton16803]

I'm just back from my local car boot where I picked up a cracking little micrometer for only £2. It seems very well made it all lines up on zero and the dials are clear. Only trouble is it's a bit different to anything I've seen before and is marked as 1 to 10000 and I'm struggling to work out how to read it. The left hand side and the number in the window seems easy enough (in the photo 0.18) but what next?

[colin hamiltonParticipant @colinhamilton16803]

[Anonymous]

I assume the micrometer is imperial and can read to a tenth of a thou. I'd read it as follows. The linear scale on the barrel gives tenths of an inch.The digit in the rectangular window gives hundreths of an inch. Assuming that the scale immediately to the right of the window doesn't move the two scales on thimble together look like a vernier and will give thousandths and ten thousandths of an inch.

Andrew

[Mike HurleyParticipant @mikehurley60381]

My thinking also. Andrew beat me to it!

It does say 0 – 1" just above the Steinmeyer label

Regards

Edited By Mike Hurley on 21/11/2021 10:17:12

[peak4Participant @peak4]

I've never seen one in real life as they don't appear very common.
The company still exists making ballscrews amongst other things, but see P181/2 of this different link
https://docplayer.net/41628763-An-incomplete-compendium-of-micrometers-and-their-makers-peter-marks-photos-copyright-their-respective-owners.html

Bill

[SillyOldDufferModerator @sillyoldduffer]

Posted by colin hamilton on 21/11/2021 09:51:35:

…

The left hand side and the number in the window seems easy enough (in the photo 0.18) but what next?

…

It's a vernier scale, have a look at this link to see an example. They're read by scanning the graduations to find an alignment and doing an addition. The example is metric, which makes for easy reading, but decade inches are just as good. Inch fraction verniers exist too; nasty! Verniers have to be practised and it's easy to misread them. Take care.

Sorry to rain on your parade, but whether or not you have a good buy is unknown. To measure to thou and tenths, you have to prove the instrument reads correctly across its full range. It won't if the internal thread is worn or the anvils are damaged, or the clutch is bad.

Accuracy doesn't matter if the micrometer is only used for rough amateur measuring, but don't assume it really reads tenths! Paying to have the micrometer recalibrated is one option, otherwise check it with Gauge Blocks.

Just an observation: it's not difficult to work to about 1 thou / 0.02mm in a home workshop with the sort of tools Model Engineers own. And that level of accuracy is more than good enough for most purposes. In contrast, getting genuine tenths accuracy is hard work.

Being a suspicious old Hector, I suspect it's rarely achieved at home, and certainly not with un-calibrated second-hand micrometers. As getting good results out of a new instrument requires a fair amount of practice, I wonder how many amateurs really get consistent high accuracy results by measuring? It's not how I work: I measure to get 'close enough', 0.02mm, and then fit as necessary. Fitting is done by using one part as a gauge and comparing: I don't know or care what the exact size is. I'm not making jigs, gauges or turbine blades!

Does anyone really work to tenths, and if so what for? How is that level of accuracy achieved across an assembly? Have the measurements been confirmed by someone else?

Dave

[larry phelan 1Participant @larryphelan1]

Anything being sold in a boot sale at that price has to be suspect !

[Jeff DaymanParticipant @jeffdayman43397]

Steinmeyer are still in business and there is a UK rep – link below

They may be able to help with a manual or operating instructions.

https://www.steinmeyer.com/en/steinmeyer/contact/international-sales-representatives/?tx_nwaddresses_addresses%5Bcountry%5D=74&cHash=a11af10c43c9ca4c2ecb0a8a1aa3578f

[Chris GunnParticipant @chrisgunn36534]

Bill, thanks for the link to the Peter Marks info on mikes, a fascinating read, where did the morning go?

Chris Gunn

[Nicholas FarrParticipant @nicholasfarr14254]

Hi, well it looks to me that the 8 at the extreme right is in line with the 7 to the left of it, Therefore I think it is reading 0.187.6" But of course I might be totally wrong.

Regards Nick.

[Brian MorehenParticipant @brianmorehen85290]

Hi Colin use some twist drils if imperial as mike looks a quick check will give you the right answers looks like a good buy

Good Luck Bee.B

[Anonymous]

Posted by Brian Morehen on 21/11/2021 12:59:17:

……use some twist drils………

Need to be cautious using twist drills to check measurements. Quality drills often have a shank that is a few tenths to a thou or more undersize. ideal would be to use a gauge block, or at least silver steel/gauge plate.

Andrew

[Chris CrewParticipant @chriscrew66644]

" I wonder how many amateurs really get consistent high accuracy results by measuring? It's not how I work: I measure to get 'close enough', 0.02mm, and then fit as necessary. Fitting is done by using one part as a gauge and comparing: I don't know or care what the exact size is. I'm not making jigs, gauges or turbine blades!

Does anyone really work to tenths, and if so what for? How is that level of accuracy achieved across an assembly? Have the measurements been confirmed by someone else?"

I wholeheartedly agree with these statements from Dave. It is exactly how I work, although for me it's usually in imperial units. I have read posts that appear indicate to that some contributors work to NPL standards and whilst some people may be capable of achieving this level of precision, or think they are achieving this high standard, I doubt if many actually do in a home workshop. That is not to criticise their work which is probably superb and far above anything I could possibly achieve myself, but it does somewhat call into question the 'holier than thou' attitude as regards accuracy and precision.

I have a 50 year old J&S 540 calibrated to 0.0001" and a 60 year old J&S 1310 calibrated to 0.0002" and whilst I set myself little exercises in working to these theoretical limits, given the wear in the machines, I doubt if anything I turn out is better than 0.001", if that. And in any event I only have 'workshop grade' micrometers and calipers to check the dimensions so I also have to resort to a bit of traditional 'fitting'. My workshop philosophy has always been if it looks right, fits right and works right, then it is right!

[Howard LewisParticipant @howardlewis46836]

As Chris says.Beware of delusions of accuracy!

Even if an instrument reads to 1/10000 of an inch, that is only really accurate under calibration conditions, which usually implies 20'C.

This why Standards Rooms and Calibration Rooms are temperature and humidity controlled, and items are only measured after they have been in those conditions for at least 24 hours.. If they are checking gauge blocks, or measuring instruments for accuracy (To determine if the item should be downgraded from "Standard" to "Inspection" or even "Workshop"

They really are looking for 100ths or 1,000ths of a thou!

Look at the sort of 6" rule that we use. Most are marked "At 20'C"

Depart from that temperature and your "Inch" may be larger or smaller, in absolute terms.

The instrument may read "accurately" when checked against gauge blocks (slips ) if both are at the same temperature, but that does not mean that the measurement IS that, relative to a national; standard..

Just that the instrument is indicating the same as the gauge block used to check it, at THAT temperature. (Whether both are fresh out of a night in the freezer, or a pot of boiling water! )

For most of what we do the temperature does not need to be so accurately controlled, just that the piston is 0.499" to fit into the 0.500" bore. to function as intended..

As an example, a vehicle engine piston may be picked up, and the gudgeon pin is immoveable in the piston. Hold the piston in your hand for a few minutes, and the pin will fall out!

A sense of proportion and fitness for purpose are useful allies!

Howard

[Oily RagParticipant @oilyrag]

So Chris….

…How do you go on with fitting a ball bearing? make the housing to the nearest thou (i.e on the loose side ) and then use Loctite? or if it is a little tight 'press' on regardless (pun intended )?

There are ways to achieve engineering fits which are simple if thought through enough. Firstly the tool needs to be sharp and set to centre height (I use an old saw blade and the tangent principle ) and one way to get the infeed accuracy is to set the topslide over to give 1 thou infeed on the topslide to equal 1/10 thou infeed on radius (i.e Sine angle for 1/10 ).This is where a Zeus book is invaluable.

Colin,

As for your micrometer I am reading it from the photograph as 1 + 8 + 1 + 9 (not being able to see the 9 on the left hand vernier scale I am assuming it lines up! )  = 0.1819"

Again, for accurate measurements I only use a micrometer (a digital caliper is a 'near enuff' instrument ) and I always calibrate the mike against slip gauges set to the size I'm measuring to.

Martin

Edited By Oily Rag on 21/11/2021 22:05:13

[Chris CrewParticipant @chriscrew66644]

Martin,

Fitting bearings has always been a bit hit and miss for me, in fact I have that very job coming up again shortly for the Radford Thread Milling attachment which requires four ball races and a thrust bearing fitting for the worm-shaft. I may have been lucky on a couple of previous occasions when I have measured the diameter of the bore of the bearing recess using a telescopic gauge and micrometer. I made the recess the same diameter as the bearing outer ring, leaving it on the small side, and pressed the bearing in using the vice although I was very careful and conscious that I may crack the casting, so I needed to feel the bearing seating properly without too much pressure but with sufficient resistance to prevent it rotating in the recess. I have used the same 'feel' for the fit of the bearings on a shaft but I admit it is a little 'hit and miss' and is where the cylindrical grinder comes in for sizing the shafts a little more accurately than in the lathe. I am not a trained metal-worker, although I did receive some professional training on lathe work during a later abandoned apprenticeship in the 1960's, so I have to do the best I can which I suspect is the case for the majority of amateur back-shed workers. Of course, I will only get one chance to fit the four bearings as I cannot obtain another casting, so if it goes a bit pear-shaped I will have to resort to the Loctite. Fingers crossed!

 

Edited By Chris Crew on 21/11/2021 22:55:44

[Martin KyteParticipant @martinkyte99762]

You cannot work to a better accuracy than you can measure. Generally you cannot even do that so you really need a better measurement than the level of prescision to which you want to work. Generally we make things to fit items we already have so measurements are relative/comparitive rather than absolure and provided readings are taken at much the same time temperature can be ignored.

For example take the machiming of a bore to be a close sliding fit on a ground shaft which is sitting on the bench. The first thing I did was to set the lathe up to turn ded parallel over the length of the bore by doing a turning test on a test boobin as normal. By small adustments of the jacking screw at the tailstock end of the Myford I was able to get both ends of the test piece to within 2 tenths over a 3 inch length. Now I cannot turn to as high an accuracy as that but I can measure to it, so I now have a lathe that will at least create a parallel bore. Whilst I was at it I turned the ends of the bobbin to be as close as I could get to the same size as the ground shaft so I had a gauge. Now I have already said I cannot turn to tenths but what I ment was I cannot guarentee the final size but with the two ends of the bobbin I could have 2 attempts. The first end came out around 5tenths under size and a tiny tweek brought the other end to better than 2 tenths. This was mostly luck. Select on test as we say in electronics. Taking a couple of thou off the first end gave me a usable test gauge.

I was now able to bore the casting to receive the ground shaft using the test gauge to check the under size fit. Quick final set of the cut and the final bore was produced which ended up as far as I can tell within half a thou of what I wanted which will do nicely.

If I could only have measure to a resolution of a thou I would have really struggled to get a good bore.

regards Martin

[not done it yetParticipant @notdoneityet]

This problem could easily be sorted by the OP – cut a bar in slightly descending increments and use the practical experience of taking measurements to determine the larger incremental changes first and later sort out the finest of diameter changes (after the use of some emery to remove the odd tenths?).

[colin hamiltonParticipant @colinhamilton16803]

Blimey I can't wait to post my next car boot bargin!! Thanks for the advice. So for the last two digits do I just see which two are aligned and that's the answer?

[SillyOldDufferModerator @sillyoldduffer]

Posted by colin hamilton on 22/11/2021 12:53:31:

… So for the last two digits do I just see which two are aligned and that's the answer?

Maybe. The easiest way to find out is to experiment. Get some drills and measure a few shank diameters. Drill shanks aren't accurate, but they're close enough to show how the vernier is read, and whether any arithmetic is needed.

¹⁄₁₆" look for 0.0625

⅛" look for 0.1250-ish (0.1245 to 0.1255&quot

³⁄₁₆" look for 0.1875-ish

¼" look for 0.2500-ish

Although they look simple micrometers aren't easy to use properly. Takes practice to read an ordinary thou micrometer and yours is ten times more sensitive. It's likely that readings will be all over the place at first. The clutch helps, but getting consistent results really requires the operator to develop a sensitive feel by practising on accurate standards such as gauge blocks. Some people get a feel really quickly, a few never get it, and most need several hours practice. Not good to learn on a wonky micrometer.

A thou micrometer will usually detect size errors in jobbing drill shafts because the steel rods used to make them aren't super-accurate. A tenths micrometer should see variations up, down, and around a drill shank, and might detect temperature changes too. If you measure a ³⁄₁₆" drill and get exactly 0.1875", there's a good chance the operator is torquing the micrometer to get the 'right' answer – observer bias. Unfortunately, in the tenths zone it's hard to tell the difference between correct measurements and errors caused by poor technique and faulty instruments.

Measure a few things and tell us what you get. We should be able to work out what the vernier is doing from a few examples.

Dave

[peak4Participant @peak4]

Here you go, a little experiment to illustrate the difficulty in trying to measure to too many decimal points, regardless of units.

I recently picked up a Moore and Wright Micro 2000 micrometer, which is probably about the limit of what one will see in a home shop.
It uses a Moire diffraction grid, on a pair of glass prisms, for the measurement, and a damped spring loaded anvil for consistent pressure whilst reading.

1st photo is a ½" gauge block, where the block and micrometer have been at a stable shop temperature of 7°C for some time. (The 7°C probably accounts for why I've not been out there much)

I picked the ½" block as we know the exact measurement we should be expecting in mm without using a calculator.
The second photo is warming the block in my quite cold hands for about 30 seconds; the block was still cold to the touch. (It might be interesting to play with a non contact thermometer).

Both measurements were consistently repeatable; the anvils and block are still in good enough condition to wring together.

It looks like the readings are a bit hard to read now the photos have been compressed by the forum software.
Top one is 12.699, bottom 12.700

Bill

Edited By peak4 on 22/11/2021 16:42:25

[Michael GilliganParticipant @michaelgilligan61133]

Nice demonstration, Bill

… it reminded me of this, from one of my favourite books :

.

MichaelG.

[mark costello 1Participant @markcostello1]

We stand on the shoulders of giants.

[not done it yetParticipant @notdoneityet]

Posted by peak4 on 22/11/2021 16:40:09:

Here you go, a little experiment to illustrate the difficulty in trying to measure to too many decimal points, regardless of units.

I recently picked up a Moore and Wright Micro 2000 micrometer, which is probably about the limit of what one will see in a home shop.
It uses a Moire diffraction grid, on a pair of glass prisms, for the measurement, and a damped spring loaded anvil for consistent pressure whilst reading.

1st photo is a ½" gauge block, where the block and micrometer have been at a stable shop temperature of 7°C for some time. (The 7°C probably accounts for why I've not been out there much)

I picked the ½" block as we know the exact measurement we should be expecting in mm without using a calculator.
The second photo is warming the block in my quite cold hands for about 30 seconds; the block was still cold to the touch. (It might be interesting to play with a non contact thermometer).

Both measurements were consistently repeatable; the anvils and block are still in good enough condition to wring together.

It looks like the readings are a bit hard to read now the photos have been compressed by the forum software.
Top one is 12.699, bottom 12.700

Bill

Edited By peak4 on 22/11/2021 16:42:25

My suggestion of self enlightenment could be easier than relying on machining or emery – just measure and record all digits, then warm the item by a few degrees and record the new digits. Done at the same time with objects multiples in of the smallest item would clearly provide increases of multiples of that smallest increase – if done as a carefully controlled experiment.

A 1” inch by 2” block would likely be sufficient if the operator is able to measure both dimensions without causing a change of conditions. Results could easily be confirmed, as accurate, by comparison with the documented coefficient of expansion for that material.

We (should) all know that machined items should be allowed to completely cool before taking measurements – particularly important for close-tolerance fits.

[SillyOldDufferModerator @sillyoldduffer]

Posted by not done it yet on 23/11/2021 06:55:41:

Posted by peak4 on 22/11/2021 16:40:09:

Here you go, a little experiment to illustrate the difficulty in trying to measure to too many decimal points, regardless of units.

…

Both measurements were consistently repeatable; the anvils and block are still in good enough condition to wring together…

…

We (should) all know that machined items should be allowed to completely cool before taking measurements – particularly important for close-tolerance fits.

Bill and NDIY make a good point about needing to account for temperature when taking high-accuracy measurements, but it's just one of several factors that reduce the accuracy of very fine measurements.

My feeling is good condition thou/0.02mm micrometers are on safe ground for practical purposes because their limitations (about ±0.01mm), hide a raft of errors that become significant when attempting to read tenths (0.002mm). Measuring tenths and better takes one across the border into difficult territory, where the operator has to get everything right. Owning a suitable micrometer in good order is just the first step.

Failing to account for sensitive error sources such as temperature means the measurements are unreliable. There's much more to measuring accurately below 0.001" than buying a micrometer, even a good one.

I suggest there are two good reasons for buying un-calibrated knocked about second-hand tenths micrometers:

• Collecting for interest. (I collect old slide-rules if I see them in charity shops, but rarely use them in anger.)
• As a way of doing a bit better than thou, but not expecting more than halving or quartering the difference. Even halving the difference requires the instrument to be checked against a gauge block.

The acid test is turn a selection of rods to various random tenths accurate diameters and have them measured by independently by someone who doesn't know what to expect. Both parties have to get the same answers. You can't mark your own homework!

Dave

[Author]

Posts