Thread: Combination squares

Originally Posted by banana

are any of those deviations from square significant enough to turn up in a knife line test?

I think the more relevant question here is "Are any of those deviations significant enough to matter AT ALL?"

Given that we're working with timber, I'd say the answer is somewhere between not really and no.

If it helps, the formula to convert degrees to mm/m is 1000*tan(angle), so 0.05 degrees is 0.8 mm/m. I'd say that doesn't really matter a whole lot.

Originally Posted by elanjacobs

I think the more relevant question here is "Are any of those deviations significant enough to matter AT ALL?"

Given that we're working with timber, I'd say the answer is somewhere between not really and no.

If it helps, the formula to convert degrees to mm/m is 1000*tan(angle), so 0.05 degrees is 0.8 mm/m. I'd say that doesn't really matter a whole lot.

I agree.

What I was showing was there's no need to fork out out a whole lot of money to get a square square.

thanks, that makes sense. the rule of thumb i've learned from woodworking youtube is that if you can strike the same knife line using the same side of the square from both directions, it's "as square as you'll ever need it to be for woodworking". but i don't yet have the experience to have a feel for that in practice, nor the precision measuring tools to translate it into something quantifiable, so the extra context is helpful

Originally Posted by banana

thanks, that makes sense. the rule of thumb i've learned from woodworking youtube is that if you can strike the same knife line using the same side of the square from both directions, it's "as square as you'll ever need it to be for woodworking".

This assumes your reference surface the stock (the bit that sits up against the reference surface) is straight but a piece of newish MDF is usually straight enough.

The limits of human vision to be able to distinguish between two very fine lines with good eyes in sunlight is about 0.1mm but in practice it's more like 0.2 mm, but because a flip angle is used this translates to 0.1 mm.

Old codgers like me with dodgy eyes are going to be at least twice as poor or worse.

The final angle resolution then depends on the square's blade length.
For 100 mm long blade the angle is then arctan(0.1/100) = 0.057º
For a 300 mm long blade it's arctan(0.1/300) = 0.019º
This assumes there is no uncertainty in the zero position of the stock which is not meteorologically correct.
One way to bring that into the calculation is to not use the blade length but the stock length as the length factor in these calculations.
The stock length is shorter so the angle then increases but not usually by much.

Whatever is used these are plenty good enough for WW but may be a problem for metal workers.

The squareness of my cast iron reference block is arctan (0.025/150) = 0.0096º and this would not be good enough for some metal work operations.

Super precise right angle comparators are, like surface plates, made from granite and they come in various grades and sizes, the top end ones eg for checking CNC machine alignment are square to 0.00002º or better.

I have some Starrett master squares in various sizes and their stated accuracy is 1/60,000 or 10 times Bob's reference block, but you won't get much change from $600 for the 6" (I did not buy them new, I'm not that loaded ). They are designed for inspection/calibration and were kept for making sure machine fences and saw blades were square.

Pretty much all my cabinetry work was done with a $40 Irwin square and that was plenty good enough.

Originally Posted by Fergiz01

What's the consensus on fitting new rules with old stocks? I have a vintage Mitutoyo stock with a new metric Stanley rule. I've tried to find a new Mitutoyo rule but no luck.

At least in the United States, combination square rules tend to come in two thicknesses, which I think of as "carpenter" (thinner) and "machinist" (thicker) grade. It's enough of a difference that you can't fit a machinist rule in a head intended for a carpenter's rule, and a carpenter's rule inserted in a machinist-square head will be sloppy and may not lock because the clamping bolt slips out. Other than that, though, I've found the rules of a given type are pretty consistent. I've interchanged any number of rules and heads. There are a few of the right thickness but the wrong slot for a given head; but they're not common.

A PEC rule did not fit my Starret head due to being thicker. Anyone want to buy a metric 100mm rule?

Dont need to be that perfect for woodwork. Ive adjusted a few of my vintage ones to be square, they have held square every since. I compared it to a friends starret. Over 30cm, i couldnt tell the difference in the lines. Good enough.

Would anyone know of a replacement metric 12inch, yeah i know, rule for my late Dads Moore and Wright square, it would be 100 YO, the rule measures 2.24to 2.27 mm thick, thanks , Ross

Originally Posted by Bill Houghton

At least in the United States, combination square rules tend to come in two thicknesses, which I think of as "carpenter" (thinner) and "machinist" (thicker) grade. It's enough of a difference that you can't fit a machinist rule in a head intended for a carpenter's rule, and a carpenter's rule inserted in a machinist-square head will be sloppy and may not lock because the clamping bolt slips out. Other than that, though, I've found the rules of a given type are pretty consistent. I've interchanged any number of rules and heads. There are a few of the right thickness but the wrong slot for a given head; but they're not common.

Hi Bill,

I went and checked how the replacement blade fits into the stock, and you're right - it's too thin. Everything is all square when it locks down which is the important thing but now that I noticed this I cannot unsee it. I've fortunately found a replacement Mitutoyo blade so will go down that road. Thanks for the heads up.

Zac.

Sent from my Nokia 4.2 using Tapatalk