Checking the accuracy of a 45 degree square?

[Quesne Ouaques]

I recently threw out two combo square heads because I found out they were not accurate. I bought a new one recently, and checked the 90 degree edge and all looks good.

But it got me thinking....

How do I check the 45 degree edge? I have never seen a reliable test for doing this. If anyone can give advice or point me to an on-line article, I would appreciate it.

Thanks

[Jason Beam]

Were the heads you threw out not adjustable? I've never had one that wasn't, but it's possible, I guess. There's usually a little bump type thing on each end of the slot the blade rides in. Adjusting them is just a matter of filing one of those bumps down a bit to get the blade to seat squarely.


As for testing 45's ... Best I can think of is a board that you KNOW is square on all 4 corners (A real square block). Then you should be able to reference the head on any side and align the blade diagonally across perfectly from corner to corner. This isn't nearly as elegant as I'd like, so hopefully someone has a better method.

[Robert Rozaieski]

You need a "square" square and a compass. First make a line 90 degrees to the straight planed board edge with your "square" square. Then pick up your compass and set it to an arbitrary setting less than the length of your 90 degree line. From the corner of the right angle, use your compass to make two marks, one intersecting each leg of the right angle (i.e. the board edge and the 90 degree line). Then connect these two points. This line will be exactly 45 degrees to the edge of the board. Check your 45 degree square against this line. It's a lot harder to describe than it is to do.

[Alex Yeilding]

You need a known right angle -- a square that you know is square, or a board cut perfectly square. You are going to mark on these, so if it is a square, you might want to put masking tape on the surface, or use layout fluid.

Scribe a 45-degree line from one leg of the square to the other. Then flip your 45-degree tool over and scribe the same line (or just observe it) from the other leg. If they match, you have a 45-degree tool.
45degree.jpg

[Zahid Naqvi]

draw a streight line. Measure 4 45 degree angles starting from one end of the line. The last 45 will tell you how accrate your 45 square is.

[Keith Beck]

I recently threw out two combo square heads because I found out they were not accurate. I bought a new one recently, and checked the 90 degree edge and all looks good.

But it got me thinking....

How do I check the 45 degree edge? I have never seen a reliable test for doing this. If anyone can give advice or point me to an on-line article, I would appreciate it.

Thanks

Believe it or not, the plastic drafting triangles you can get at art supply stores are pretty darn accurate for not a whole lot of money. That's what I use to set up many of my tools and check my squares against.

Keith

[Thom Sturgill]

Several good ideas here. For another, get a board with two sides parallel and draw a 45 from each side and check that they are parallel. Of course accuracy is only as good as the parallelism of the board.

[Jason Beam]

Several good ideas here. For another, get a board with two sides parallel and draw a 45 from each side and check that they are parallel. Of course accuracy is only as good as the parallelism of the board.

Thom,

I thought that might do it, too, but unfortunately they will ALWAYS be parallel, no matter what angle they're at. Because you're flipping it and rotating 180 degrees, you're only creating the same angle.

I wasn't sure that was right at first, so I drew it in sketchup and sure enough - ANY angle will draw parallel lines so the method won't work.

[Thom Sturgill]

Thom,

I thought that might do it, too, but unfortunately they will ALWAYS be parallel, no matter what angle they're at. Because you're flipping it and rotating 180 degrees, you're only creating the same angle.

I wasn't sure that was right at first, so I drew it in sketchup and sure enough - ANY angle will draw parallel lines so the method won't work.

You are partially right, you are rotating, which would give a check on parallelism of the board. Then flip and draw a line which *should* be perpendicular to the first set.
You have to check it with a square or by the 3,4,5 method and a ruler. Might just be easier to use one of the other methods, I was trying to come up with something as easy as the test for a square where you simply draw a line perpendicular to an edge and flip the square and check the line.

[Jason Beam]

Several good ideas here. For another, get a board with two sides parallel and draw a 45 from each side and check that they are parallel. Of course accuracy is only as good as the parallelism of the board.

You are partially right, you are rotating, which would give a check on parallelism of the board. Then flip and draw a line which *should* be perpendicular to the first set.
You have to check it with a square or by the 3,4,5 method and a ruler. Might just be easier to use one of the other methods, I was trying to come up with something as easy as the test for a square where you simply draw a line perpendicular to an edge and flip the square and check the line.

So you check for perpendicular, not parallel like you'd said originally? That way would work, except it's not quite as easy to check for for square at the intersection of two lines somewhere in the middle of a board, though. At least not terribly accurately, like you'd pointed out.

I came at it the same way as you did, I think. Trying to make a "draw two lines" quick test, but in the case of a 45, I don't think that there is an easier one than the one mentioned about using a known 90 degree corner for it. That method would be most simlar to the method for checking for square.

And for me, by that time i'd rather just use a drafting triangle. It's probably the quickest way to acceptable accuracy.

[Tom Grupstra]

If you use the same square for the whole project,everything will be out of square equally,no one will notice?....lol just kidding.