Digital Fraction Calipers are everywhere!

[JR]

. The same measurement on a caliper that reads to the 64th with a resolution of 1/128 would be 2 and 3/64 +/- 1/64 or in decimal equivalent somewhere in the range of 2.03125 to 2.06250. You're looking at a .002 window of error with the decimal and a .031 window with the fractional.

I don't think that's correct. If the resolution is to 1/128" then the 3/64" measurement is accurate +- 1/128"(if the measurement is off by greater than 1/128" it will show as the next increment of 1/64" in the desplay). In decimal terms, that's 0.047 +- 0.008. I guess that makes the margin of error =0.016.

Or, I could be totally wrong.

JR

[Brian G]

I modified/co-mingled the concepts of several fraction/decimal/metric conversion charts so that I can easily match one to the other without having to think too hard.

DecimalEquiv.pdf

I printed several and have them in strategic spots in the shop (by the drill press, near the TS, and with the planer.

[AndyF]

Nice!

That'll come in handy.

[JR]

I printed several and have them in strategic spots in the shop (by the drill press, near the TS, and with the planer.

Nice, Brian! I won't need "several", though. It's all sort of one spot.

JR

[lcm1947]

Oh WOW! That's a great chart. Thanks Knuckles!

[LCHIEN]

Accuracy vs. resolution

The way I see it is

1/64 = .015625 so if you only have accuracy to .001 and a resolution of .0005 you will never be able to accurately measure .015625

1/64 = .015625
1/32 = .03125
1/16 = .0625
1/8 = .125
1/4 = .25
1/2 = .5

It seems to me that to measure to 1/8 inch accuracy, you need accuracy to .001 (actually I think .005), to get 1/16 you need accuracy to .0005, to get 1/32 you need accuracy to .00005.

Like I said I could be all wrong on this, this is just my interpretation of the numbers.

Accuracy and resolution are two different things, although usually they have similar values for a tool or instrument, there's little use in an instrument that has high accuracy but no resolution or conversely an instruent with high resolution but no accuracy.

Accuracy is a property inherent to the instrument.
In this case, for a digital readout, resolution is a property of the display, and the display mode. The digital readout in the 1/64th mode (display resolution of .015") has much less inhererent resolution than the inch scale (.0005" on mine). also the mm readout (with 1/10ths mm or about .004") is in between.

IIRC the claimed accuracy of these kinds of calipers is on the order of .001" and that's because there's an internal grid they count as it goes by to determine where the caliper jaw is. The internal grid has to be linear and accurate along its whole length e.g. spaced both evenly and so many per inch and not change (stretch, shrink or deform) with time or reasonable temperature variations.

To put it in another perspective, lets take an ordinary Steel rule.
It's made of stainless steel and has a temperature coefficient of expansion small enough that at our shop temperatures it does not vary more than .001" in length (I didn't look this up, I'm just assuming for example)

Now we have a machine that can engrave markings on it very precisely, say within .0005" of the desired location, we engrave a mark at 1.0000" from the end, and and another at 2.0000" and another at 3.0000" and so on.

Now we have a great ruler, or almost, anyway. It is accurate to .001" or better, but when we go and try to use it, we find it only has resolution of 1". Try and build some fine furniture with that puppy! In reality this is not useful and as I said at the begiinning, a ruler accurate to .001" but with a resolution of 1" would be mighty useless.

You can see the other example coming, they put 1/100th" markings on the ruler but were very careless and stretched it so that 1" marking was actually at 1.10" and the 2" marking was perhaps at 1.90 and the 99 marks between them were more or less spaced evenly but who knows?
So this ruler has .01" resolution but maybe .1" accuracy; I wouldn't have it.


To resolve JA's misunderstanding, when it reads 1-1/64ths, the doesn't mean its at precisely 1.015625". what it means is that the jaw is some where around 1.008 and 1.023 AND the exact point at which it changes to the next reading may vary from the ideal crossover point by .001" (the accuracy of the instrument).

[radhak]

...
To resolve JA's misunderstanding, when it reads 1-1/64ths, the doesn't mean its at precisely 1.015625". what it means is that the jaw is some where around 1.008 and 1.023 AND the exact point at which it changes to the next reading may vary from the ideal crossover point by .001" (the accuracy of the instrument). The reading is so coarse as not to have any resolution.

So in the realm of woodworking, if we stick to fractions, the variation is 1/1000". When we are looking for measurements at the level of 1/64s, isn't 1/1000" far finer than what we need (or can handle)? ie, if my tenon measures 1 3/64" wide, and i seek to make a mortise for it, it needs to be around that number in width, where variation in the 1000ths may be negligible. And since the mortise might need to be a 64th (or two) more, we are talking increments of about 15/1000, so the small variation might just get absorbed. Am i making sense?

I am only trying to ascertain if buying one of these callipers is gonna help, or are they so out of whack i might just have to dump them later. (Of course, at my level of expertise, anything finer than 1/8 is more than my hands can accurately handle )

[Lance]

Wixley addresses the resolution/precision issue

Wixley is the only on of these that explicitly addresses the issue of having fractional measurements with digital precsion to the 1000ths. They ALWAYS show the decimal value, and pop up the fractional equiv when you hit that measurement.

See what I mean at:

http://www.wixey.com/calipers/techno...x.html#precise

Some people still prefer the fractional dial calipers, marked at least to 64ths, feeling that they can eyeball 128ths and even 256ths with reasonable accuracy.

Rad, I would recommend that you either go dial or Wixley. That is a researched but not experienced opinion. I just got a digital one, not wixley, for X-mas, and haven't decided whether to keep it or take my own advice.

-Lance

[whitecobra]

I agree with Lance
We just installed Wixey's electronic fence on our BT3K (BTW it takes less then 10 minutes it is all but made for the BTU perfect fit easy to use and ONE neat system) (no affiliation with them)

Anyway it is set up and reads either metric or English to 0.001 PLUS fractional at one time

So it would say 3.312 and then right next to it pops up the 5/16 mark

VERY nice and does deal with the concern of someone else determining accuracy

As for Loring's description I learn more from this guy weekly then in 20 years of being a doc. THANK YOU SIR

Dr D

[LCHIEN]

Wixley is the only on of these that explicitly addresses the issue of having fractional measurements with digital precsion to the 1000ths. They ALWAYS show the decimal value, and pop up the fractional equiv when you hit that measurement.

See what I mean at:

http://www.wixey.com/calipers/techno...x.html#precise

Some people still prefer the fractional dial calipers, marked at least to 64ths, feeling that they can eyeball 128ths and even 256ths with reasonable accuracy.

Rad, I would recommend that you either go dial or Wixley. That is a researched but not experienced opinion. I just got a digital one, not wixley, for X-mas, and haven't decided whether to keep it or take my own advice.

-Lance

I didn't realize Wixey had that feature, and while neat in some ways does not address the problem perfectly. As we measure our work it is seldom precisely 17/64 although we might intend it to be, and if we WANT to see 17/64 (because that's the fraction its closest to) but its not exactly 17/64 (within +/-.001" anyway) then it won't do that job.

I can think of several different modes to operate in that might be more useful, but they start to become a setup nightmare with only a couple of buttons to use.

[siliconbauhaus]

I agree with Lance
We just installed Wixey's electronic fence on our BT3K (BTW it takes less then 10 minutes it is all but made for the BTU perfect fit easy to use and ONE neat system) (no affiliation with them)

Anyway it is set up and reads either metric or English to 0.001 PLUS fractional at one time

So it would say 3.312 and then right next to it pops up the 5/16 mark

VERY nice and does deal with the concern of someone else determining accuracy

As for Loring's description I learn more from this guy weekly then in 20 years of being a doc. THANK YOU SIR

Dr D

Wow...cool item. Shame it costs more than my saw did

[25]

Wow...cool item. Shame it costs more than my saw did

Yeah but you can just move it to your next saw .

[Lance]

Picking the nit's nits!!

I didn't realize Wixey had that feature, and while neat in some ways does not address the problem perfectly. As we measure our work it is seldom precisely 17/64 although we might intend it to be, and if we WANT to see 17/64 (because that's the fraction its closest to) but its not exactly 17/64 (within +/-.001" anyway) then it won't do that job.

You are absolutely right. That's one of the reasons people still like the fractional dial's: more contextual and complete information, with slightly less quantifiable analog precision (Mk I eyeball).

Still, if I measured 0.262" right after seeing the "0.250 1/4" reading, I would be pretty sure it was between 1/4" and 17/64". Seriously, I'd probably just scroll to the next marking that would allow me to fit the joint later (i.e. bigger for a tenon) or just transfer the measurement directly (best option, most of the time.)

For me this would be easier/faster than converting from decimal. The more I think about it, fractional dial would be the best option for me personally. I think it definitely runs to personal preference and what you are used to.

-Lance
(Definitely NOT challenging Loring's unassailable engineering awsomeness!)

[LCHIEN]

...(Definitely NOT challenging Loring's unassailable engineering awsomeness!)

QUANDO OMNI FLUNKUS MORITATUS?

I'm not challenging your Latin. I had to look it up:





When all else fails play dead.

[Jeffrey Schronce]

I installed the Wixey height gauge from Rockler today on my 15" Delta Planer. Took about five minutes total. Involved drilling one simple hole into the base of the stand. When I looked at their site people installed them on the 15" Delta with lots of aluminum angle etc. I found no need for that. Incredibly easy and accurate set up. I am very pleased with the product!