OMG, 30 y/o go-to ruler incorrect

[Mike Cutler]

I probably have 30 or more measuring things. Rulers, tapes, Mylar and metal strips, squares, etc. I use my Starrett 24” and 18” rules as my baseline for accuracy as they cost me the most but really have nothing else to use for a baseline. It can be difficult to figure out what is the most accurate device to use for comparison. For years I used my squares made by Bridge City Tools but kept wondering why my other squares were always off a little bit in comparison. Turns out the BCT squares were all off! Price does not reflect accuracy it turns out.

Peter
You're probably pretty good with the Starretts. They have always made nice tools. I have a 6' Starrett machine rule that lost it's NIST CERT, thus making it 6' straight edge and not worth anything.
It was heaed for scrap metal dumpster at work and instead found a way to my shop. That's my "go to" reference.

It's odd, and unfortunate, that your BCT square is not square. John was really anal about machining accuracies?????

[Max Neu]

This reminds of a saying I always think about;
"A man with 1 watch will always know what time it is,a man with 2 watches will never know what time it is. "

[David Utterback]

"If the error is evenly distributed along the entire length of the measuring device, then any portion of that length will be precisely proportionate to the total error. "
I agree with the above statement, HOWEVER I never have found it to be true in real life.
Have found that somewhere the tapes, rules, etc. that are off (don't match as I don't have NIST traceable to check with) are off in different areas, some even were true to others at longer distances and didn't match somewhere in the middle.
Best I have found when wanting matching rules/ tapes, etc is to lay them out in the store and match them up ALL the way out to the end. Sure get some funny looks when stretching 30' tapes out 6-10 wide

Your observation about variance over the length of the ruler is exactly what I have. Undersized at 1 - 3", variably over between 3 and 10", etc. etc. What I do not understand is why this type of error occurs in the manufacturing of linear measuring devices. I know they cannot all be NIST traceable, but how hard can it be to set up machinery to reproduce rulers that are systematically close to accurate. My ruler has engraved markings. I do know what kind of machinery is used to manufacture them. Apparently, there was some slop between the engraving head and the substrate holding mechanism.

[Brian Holcombe]

I check calipers against Jo blocks then check rulers against calipers. I compare machine fences with test cuts against long calipers (24”). Throw away anything that can’t be calibrated and is out of spec enough to cause error.

Doing this, in addition to squaring fences to a standard in my shop, became and incredible time saver. So nice to never second guess, and just go around and check machine fences and wear items every few months. I can’t remember the last time I needed to interrupt my work to check things or second guess squareness or accuracy.

I still cut to a knife mark when fitting things but that’s a different animal and I may be able to eliminate most of that by adding a DRO to more tools. Similar to scribing for machine work, it’s useful but the DRO is actually more accurate.

These precaution are cheap by comparison to screwing up something expensive.