As a woodworker, squares are a bane of my existence. They constantly undermine my confidence and I have never come across a square I am actually happy with. It seems like no-one makes one good enough for me at least.
It's definitely a crapshoot. I tried to buy a Chappell framing square, supposedly guaranteed accurate to within 0.003 inches, but they were discontinued before I got there. Then I tried to buy a Johnson CS 12 (stainless steel) but they also disappeared before I could buy. (No information about how accurate they were; they may be available again.) Then, just to have something I could use for a couple of new projects, I bought a cheap Johnson CS 7 (white on black aluminum) from Lowe's. The damn thing turned out to be almost perfect! It was so good that I decided to put it away and buy an Empire white on blue aluminum for everyday use. That one is pretty good too!
@@practicalmetrology8599 well your video is quite interesting because I use a Johnson speed square (or is it rafter square?) among others. The 90 deg measured using the base of the square stood up on a flat surface is out because the extrusion was not quite flat (had a convex bow to it). I had to hand dress it to get it somewhere near but it had been throwing me off for weeks because I was measuring things with one side of the base and then checking with the other and thinking my work was off. I just can't understand why it is so hard for someone somewhere to manufacture an accurate square. I can understand obviously why a rafter square might have a loose tolerance after all it's not intended for fine measurement but even the engineers squares I have come across (many marketed as DIN rated etc are terrible and if you drop them then forget it). I hapen to think 0.001 over 4" is quite alot. It makes a difference if you're using a square to square up a shooting board for example. Anyway rant over sorry bout that.
Absolutely no need to apologize. I agree with you. I was not really upset with the 16 inch Empire combination square that started this whole thing for me in 2018 (the oldest video). I just thought it was important that other people be made aware of the issue. That problem did cause me to screw up a couple of parts, but the satisfaction of figuring out what was going wrong made up for that. But when I bought the little 6 inch Starrett and found that it was 0.010 out of square, I went ballistic. I'm still upset about that; as I said (paraphrasing) in my Amazon review, I am personally offended that Starrett's current management would throw away a 140 year tradition of quality. However, overall I have come to accept the situation. It is possible to purchase a cheap square that is accurate. But you have to test everything, and keep testing if you drop things. And dropping things is a bigger issue every day as I get older, so I keep my square testing setup handy.
Wow, thanks for the comment! Would have been perfectly OK to be critical -- I just hoping that someone would say something! I know this video could have been a lot better -- but I also found out that more effort would not have been justified by the level of interest.
I need about an 8' long straight edge. Now I just have to figure out what to draw the line on. Sure hope they don't mind me doing this on the floor at Home Depot:)
I am new to woodworking and have been wondering if the higher end tools (squares n such) are worth it...I think now I may have a way to test whatever I buy and can return it if it isn’t as straight as I feel comfortable about. Or maybe to my surprise, the stuff at HomeD, Lowes and Ace will be decent enough :) Thanks for the video!!!
What I said about lining up the edge with the belly-type error, beginning at about 6:25 was the best approach if you have a single bulge on the surface. However, if there is not a single bulge, then what you really want to do is to align the highest point on the edge, as I showed, but then to rotate the edge about that point until you get a second point of contact between the edge and the line. What you want to achieve is that the two highest points on the edge are in contact with the two highest points on the line. That is exactly the same thing that you do with the hollow bow error, but the two highest points aren't at the ends of the straightedge when there is a bulge. Those two highest points, in the case shown, represent the best straight line approximation to the edge, and the difference between the edge and the line is twice the error with respect to that line.
you realize that there is no 2D straight edge, they are all 3D, they all have thickness, so i think all of that is pointless, with the paper; i ,as a mechanic, check engine blocks and heads from time to time (i tend to trust no one, it is MY responsibility, I check ALL, myself, been VERY lucky for over 30 years...) and i need 0.0005 to 0.001 inch MAX. out-of-flat on a straight edge, and that is no way to check; is there a better way, with an indicator, maybe?
Thank you for your comment! You are correct; real straight edges are 3D, and that fact is important. That is exactly why many people advocate flipping the straightedge about a vertical axis -- so they can keep the same side of the straight edge against the plane. This can allow one to avoid the 3D issue, but as I've shown, that method is sensitive only to the "even" portion of the straightness error. Actually I do have more to say about this, and I hope that you are interested enough to stick around to read it. I haven't thought about this in several years though, so I need some time to put a more comprehensive answer together. I will do that.
Are you there? I have quite a bit further to say, including some practical suggestions for moving forward, but I'm not going to bother writing them down if no one is reading.
As a woodworker, squares are a bane of my existence. They constantly undermine my confidence and I have never come across a square I am actually happy with. It seems like no-one makes one good enough for me at least.
It's definitely a crapshoot. I tried to buy a Chappell framing square, supposedly guaranteed accurate to within 0.003 inches, but they were discontinued before I got there. Then I tried to buy a Johnson CS 12 (stainless steel) but they also disappeared before I could buy. (No information about how accurate they were; they may be available again.) Then, just to have something I could use for a couple of new projects, I bought a cheap Johnson CS 7 (white on black aluminum) from Lowe's. The damn thing turned out to be almost perfect! It was so good that I decided to put it away and buy an Empire white on blue aluminum for everyday use. That one is pretty good too!
@@practicalmetrology8599 well your video is quite interesting because I use a Johnson speed square (or is it rafter square?) among others. The 90 deg measured using the base of the square stood up on a flat surface is out because the extrusion was not quite flat (had a convex bow to it). I had to hand dress it to get it somewhere near but it had been throwing me off for weeks because I was measuring things with one side of the base and then checking with the other and thinking my work was off. I just can't understand why it is so hard for someone somewhere to manufacture an accurate square. I can understand obviously why a rafter square might have a loose tolerance after all it's not intended for fine measurement but even the engineers squares I have come across (many marketed as DIN rated etc are terrible and if you drop them then forget it). I hapen to think 0.001 over 4" is quite alot. It makes a difference if you're using a square to square up a shooting board for example. Anyway rant over sorry bout that.
Absolutely no need to apologize. I agree with you.
I was not really upset with the 16 inch Empire combination square that started this whole thing for me in 2018 (the oldest video). I just thought it was important that other people be made aware of the issue. That problem did cause me to screw up a couple of parts, but the satisfaction of figuring out what was going wrong made up for that.
But when I bought the little 6 inch Starrett and found that it was 0.010 out of square, I went ballistic. I'm still upset about that; as I said (paraphrasing) in my Amazon review, I am personally offended that Starrett's current management would throw away a 140 year tradition of quality.
However, overall I have come to accept the situation. It is possible to purchase a cheap square that is accurate. But you have to test everything, and keep testing if you drop things. And dropping things is a bigger issue every day as I get older, so I keep my square testing setup handy.
Thank you for the video. This makes life easier. Happy New Year!
This is terrific! Thanks for sharing the two wrong flips to demonstrate how you can be fooled with those methods - that's great teaching!
Thank you and others for the recent comments. I am glad that at least a few people are finding something of value here.
This is what i was looking for! Thank you sir!
Glad I could help!
Thank goodness I found this! FINALLY an explanation that makes sense! Thank you!
Wow, thanks for the comment! Would have been perfectly OK to be critical -- I just hoping that someone would say something! I know this video could have been a lot better -- but I also found out that more effort would not have been justified by the level of interest.
I need about an 8' long straight edge. Now I just have to figure out what to draw the line on. Sure hope they don't mind me doing this on the floor at Home Depot:)
Thank you, this has been very informative
thank you
I am new to woodworking and have been wondering if the higher end tools (squares n such) are worth it...I think now I may have a way to test whatever I buy and can return it if it isn’t as straight as I feel comfortable about. Or maybe to my surprise, the stuff at HomeD, Lowes and Ace will be decent enough :)
Thanks for the video!!!
Very well explained and extremely useful.
Thank you for making this video. This has helped me immensely.
Very glad to hear that!
What I said about lining up the edge with the belly-type error, beginning at about 6:25 was the best approach if you have a single bulge on the surface. However, if there is not a single bulge, then what you really want to do is to align the highest point on the edge, as I showed, but then to rotate the edge about that point until you get a second point of contact between the edge and the line. What you want to achieve is that the two highest points on the edge are in contact with the two highest points on the line. That is exactly the same thing that you do with the hollow bow error, but the two highest points aren't at the ends of the straightedge when there is a bulge. Those two highest points, in the case shown, represent the best straight line approximation to the edge, and the difference between the edge and the line is twice the error with respect to that line.
Very very good!
Brilliant!!!
I love you
you realize that there is no 2D straight edge, they are all 3D, they all have thickness, so i think all of that is pointless, with the paper; i ,as a mechanic, check engine blocks and heads from time to time (i tend to trust no one, it is MY responsibility, I check ALL, myself, been VERY lucky for over 30 years...) and i need 0.0005 to 0.001 inch MAX. out-of-flat on a straight edge, and that is no way to check; is there a better way, with an indicator, maybe?
Thank you for your comment! You are correct; real straight edges are 3D, and that fact is important. That is exactly why many people advocate flipping the straightedge about a vertical axis -- so they can keep the same side of the straight edge against the plane. This can allow one to avoid the 3D issue, but as I've shown, that method is sensitive only to the "even" portion of the straightness error.
Actually I do have more to say about this, and I hope that you are interested enough to stick around to read it. I haven't thought about this in several years though, so I need some time to put a more comprehensive answer together. I will do that.
Are you there? I have quite a bit further to say, including some practical suggestions for moving forward, but I'm not going to bother writing them down if no one is reading.
Thanks but this is ermm... Obvious.