Introduction: Checking a Square Before You Buy

Picture of Checking a Square Before You Buy

This is my old framing square. It has taken a lot of abuse, mostly from things I have done in attempts to make it more accurate that did not turn out as well as intended. I have thought about simply replacing it.

I looked at new framing squares on-line. Reviews were mixed. Some found the square they purchased to be square. Others found the same makes and models not to be square.  

This Instructable will show an easy way to check a square accurately in a store before buying, and it will offer a surprise to consider when looking for a very accurate square.

Step 1: A Little High School Geometry

Picture of A Little High School Geometry

"Two lines perpendicular to the same line are parallel to one another." is a first-year geometry theorem. No one was watching me in the store, other than the security cameras, so, I pulled three identical squares from the rack and spread them out on the floor. Their labels did not wrap around the square, which would have compromised my test but the labels were only stuck on the surface of the squares. I used one square as a straightedge. I slid the other two squares against this side. Then I slid them toward each other. I was careful to be certain the squares rested firmly against the straightedge. If the squares are really square, the vertical edges should meet along their length. Notice the gap between the edges where the two squares meet each other, especially how it forms a shallow "V" that becomes wider at the upper part of the photo. These squares are not square. See the second photo from a close-up photo of the gap.

I had checked framing squares this way in two other stores, but had not thought to take a photo. These squares are shorter than a framing square. This particular store had only one framing square on the rack, so I chose to test these shorter squares. The framing squares I checked in other stores all had a "V" gap, too; but, the gap on the squares shown here was the most severe.

One lesson learned is that squares off the rack in all price ranges may be accurate enough for framing houses with 2 x 4s, but those I checked are not good for precise work, like making furniture, without making adjustments with an anvil and a ball peen hammer. (This method for adjusting a square makes a dimple that brings the legs of the square nearer to each other or pushes them apart, depending on where the dimple is placed in the corner.)

Step 2: Surprise!

Picture of Surprise!

This is called an "L" Connector. They are in the portion of big box home improvement stores where you find corner brackets and similar things for connecting joists. I had long wondered about them and they are surprisingly accurate as squares. The legs of the larger "L" Connector are 12 inches in length each.

Step 3: Test

Picture of Test

I used three "L" Connectors to apply the test I showed in step 1. All of the edges met very nicely. See also the second photo. Whereas there was quite a gap between adjoining edges in the second photo of step 1, here the edges meet very tightly. The apparent gap at the very end of the two "L" Connectors is not due to any lack of squareness, but the edges are made with a little recess on each. The edges are still parallel.

If you choose to purchase an "L" Connector for use as a square, do not take what I have presented here at face value, but pull three from the rack and lay them out on a flat surface as I have shown here and check them, yourself.

Step 4: Applications

Picture of Applications

I chose to buy an "L" Connector just as a reference check for my other squares. It would suffice for an inexpensive square with very good accuracy. It could easily be adapted for use as a try-square. See this Instructable. If done carefully, the legs of the square could be extended by placing the legs against something straight. Bring a straight bar against the straightedge and clamp it to the "L" Connector. Drill holes and rivet or bolt the pieces together very firmly.  

I did make one little modification to the "L" Connector. The inner corner of the "L" Connector came with a round corner. That would make it difficult to use as shown in the photo. I used a file to make the corner square.

I am still thinking about whether I will work at improving my old framing square or replace it. The "L" Connector used as a square has given me a better idea than I had before of the exact problems that need correction. 

UPDATE (April 25, 2013): I was at Harbor Freight last evening and tested their framing squares. They were accurate, so I bought one. Today I was also at an Ace Hardware and their framing squares were accurate, too. My old framing square will be used with welding projects that need to be somewhat square. Since it is far less than perfect in several respects, it will not matter if weld spatter encrusts it. 


MrOddjob (author)2013-04-24

Flipping the square to test the accuracy of the two lines, as Phil B and Pfred suggest, is pretty much the way I check my spirit level. Lay the level on a flat surface, mark the position then check the position of the bubble. rotate the level 180 degrees, ensuring it is in the exact same place that you marked. The bubble should correspond exactly to the position you first noted. If it does not then the level is out.

Phil B (author)MrOddjob2013-04-24

You make a very good point. I had never considered checking a level in that way until I was working on a wheelchair ramp project with a man older and more experienced than I. He consistently took two readings with his level, turning it 180 degrees between the readings. I generally do that now, too.

imcp1024 (author)Phil B2017-08-25

yep, puts another hat on measure twice.

darrenah (author)2016-01-28

Thanks this instructable is very helpful!!

Phil B (author)darrenah2016-01-28

Thank you for looking. I am glad you can use it.

metalart72 (author)2015-08-03

Wow! this is just great!

Phil B (author)metalart722015-08-03

thank you for looking.

chriscochran (author)2013-10-27

I,too have run into squares that were inaccurate off the shelf numerous times, even smaller ones like an aluminum speed square. There is a technique I saw in fine home building years ago for recalibrating a framing square that actually works, although it may involve a little trial and error. Take a tapered metal Punch and lay the square on a flat surface. Hit the punch to make a small indentation either at the outside or inside corner of the square. I haven't done it for a while so I don't remember exactly but I believe hitting the outside corner will close the angle of the square and the inside corner will open it. You may have to play around with it a little bit but it definitely does work.

Phil B (author)chriscochran2013-10-27

Thank you for your comment. Yours is a good point. About 40 years ago I found a 1913 print copy of Samuel Griffith's Woodwork for Secondary Schools. I believe it was that volume that mentioned using a punch to expand either the inner or outer part of a framing square's corner to close or open the legs of the square and acquire accuracy. (That book in the 1915 edition has been available on-line free as a PDF download, but I am not certain it still is.)

I once did an Instructable on checking a square for accuracy and making it accurate. At that time my favorite method for restoring accuracy was to dress the blades of the square with a hand file. I had always been hesitant to dimple the corner of a square with a hammer and a punch. I have grown past that and have successfully used either a punch and hammer or the rounded end of a ball peen hammer to spread or close the legs of the square and make it accurate. At the time I did this Instructable I finally decided to replace my first framing square after 40 years. To my surprise, the most accurate framing square I found was a $9 steel square at Harbor Freight. After I had used it a few times, I did place it on an anvil and strike it once with a ball peen hammer. Now it is very good. You can see how I use and simultaneously check that square now at step 6 of this Instructable.

This sounds terrible, but once I found a way to make a square really accurate I began acquiring more squares than I can use in a lifetime, although each was acquired for a different reason and purpose. The best squares I currently have are some I made from scratch. Here is my best, although I have tweaked it a few times after some discovery or accident that made it less accurate temporarily.

It is really nice to use a square and know that my practice in the use of that square is likely less accurate than the square, itself, is.

graydog111 (author)2013-09-29

Good advice Phil. Years ago my dad bought a 10 foot satellite dish and it was up to me and him to install it. He lived 125 miles away, so I told him to set the pole in 1 cu.yd. concrete PERFECTLY LEVEL. I wend down to mount the dish and the pole was crooked. We checked his level and it was inaccurate. :-(

Phil B (author)graydog1112013-10-08

A few years ago another fellow and I were building a wheelchair ramp. I had never thought about it, but the other fellow always turned his level end for end to check the reading he had just gotten. He was older and more experienced than I am. If your father had turned his level end for end and taken a second reading, you would have had a level pad. I now turn my level end for end to take another reading as standard practice, even though I know that level always gives me the same reading in both directions. Thanks.

oilitright (author)2013-08-16

Hey Phil after I messaged you I checked your profile and saw this Instructable and wanted to say 'well done' for letting folks know that there is a lot of stuff like this going on in items we might assume were accurate when they just are not. I worked for a company that made stuff that went into space early in the Apollo program. I worked in what was laughingly called 'quality control' doing calibration on instruments and the difference between alleged "standards" and what got passed was amazing.

Phil B (author)oilitright2013-08-17

Your comment about what passes for standards reminds me of a story told to me by a guy I know. He works in highway construction. They were building an arched overpass. Quality assurance checks needed to be made of the concrete poured so it would resist cracking with freezing and thawing cycles. The air content was too low, too low again, and again. They kept pumping air into the concrete. Then someone checked and the instrument that measures the amount of air in the concrete was malfunctioning. They estimate the concrete in that overpass has 14 times too much air in it. They let it go as is because there are redundancies built into the design.

jimbotheconflictor (author)2013-05-12

L connector as try square. Genius.

Thanks. I was just curious, and decided to check them. As you know, I commented on your Instructable about truing up a combination square. I gather you like a square to be square, too. I remember someone once telling about a Scottish king (I think) who was confined to prison. He loved clocks and spent his time trying to get a dozen or more clocks strike the hour simultaneously. I now have half-a-dozen square and it drives me nuts when they vary from one another. (I did not set out to have that many squares, but it was almost an accident.) Still, different squares work better in different applications.

Bill WW (author)2013-04-30

Too bad nobody was watching. It would have been interesting.

Phil B (author)Bill WW2013-04-30

People who sell tools sometimes know quite a bit about what they sell, but sometimes not. When they do not, they get a deer-in-the-headlights look when a customer seems to know something about the technical aspects of the merchandise. Sometimes they get defensive. I remember a journeyman machinist at Sears trying to return the Craftsman tools he bought for his classes because the instructor insisted on Starrett tools. The sales clerk kept insisting Craftsman tools were every bit as good. There was quite an impasse with Sears losing points on customer service. I once went into Radio Shack to buy an LM741 variable voltage regulator chip I needed for a project. I also wanted the right control resistor. I stood there at the counter and factored the formula on the back of the blister pack. Then I calculated what I needed while the clerk stood there with no idea at all what I was doing or what it meant.

Contrast that with a woman who phoned me from General Electric. I had to replace the main service panel in our house. When I turned everything "on" 230 volt motors would not run. Threads for a bolt in a bus bar had been compromised during manufacture. I wrote to the company. I thought perhaps the woman who phoned was in customer service. I asked about her familiarity with the breaker panel. She told me she was an engineer and the plant manager. They cared enough to send their very best. She also sent me a bag of new parts to replace what was compromised.

rimar2000 (author)2013-04-24

Your instructables are always useful, Phil. Thanks for sharing.

Bill WW (author)rimar20002013-04-30

Por supuesto que sería diferente en el hemisferio sur. ¿No es la geometría de 180 grados en Argentina? ¿O es 360 grados?
Feliz otoño, Osvaldo. Debe ser primavera aquí, pero todavía está frío.

Un abrazo.

Of course it would be different in the southern hemisphere. Isn't geometry off by 180 degrees in Argentina? Or is it 360 degrees?
Happy Fall, Osvaldo. It should be Spring here, but it is still cold.

A hug.

rimar2000 (author)Bill WW2013-04-30

Bill, according to INDEC (National Institute of Statistics and Census) trigonometric circle is 46 °, and annual inflation in Argentina is 10.2%. The Kirchner government is like that.
Bill, según el INDEC (Instituto Nacional De Estadística y Censo) el círculo trigonométrico tiene 46°, y la inflación anual en Argentina es de 10.2 %. El gobierno kirchnerista es así.

Phil B (author)rimar20002013-04-24

Thank you, Osvaldo. I hope your squares are accurate or at least accurate enough for what you need.

bptakoma (author)2013-04-28

Remind folks to check squares on a flat surface. it's possible that the concrete floor in your photos wasn't consistently level, and that affected the results. Top of a good quality table saw should be good. :)

I love math in the real world.

Phil B (author)bptakoma2013-04-28

You are correct. The flattest surfaces available in the stores I visited had some roll and were not completely flat. If you look closely at my photos, you can see shadow areas where part of the edge of one square rises a little above the square's edge next to it. A rolling floor is probably what one will find in most stores. One answer would be to try different sections of floor or rotate the orientation of the squares on the floor. Then check to see if the results change or not.

pfred2 (author)2013-04-24

I'm not surprised consumer grade squares are not terribly square. Accuracy costs more than most are willing to pay. Anyhow the test I've heard of for squares pits one square against itself. This eliminates the variable of another square's inaccuracy from coming into play. Scribe a line on a straight edged panel, then flip the square over and try to scribe another line over the first one. Any discrepancy between the lines is how much your square is off times two.   But you should know about that if you already know the dimple trick.

Phil B (author)pfred22013-04-24

phred, scribing the line you mention is similar to the paper I put down, as I mentioned to a couple of people. The big problem I have had with such a procedure is to find a straightedge that truly does not vary or wander the least bit. You are right about paying for accuracy. If you go to Amazon and price try squares, those that guarantee accuracy cost several times more than those at the corner hardware store.

nanosec12 (author)2013-04-24

I never would have thought to put 3 of them together to check the accuracy...but then again I always 'assumed' they were more precise in the first place.

Well Done

Phil B (author)nanosec122013-04-24

I did not think of putting three squares together for a check, either, until a few days ago. It is really quite shocking how far out a new square can be. Thank you for looking and commenting.

Galway (author)Phil B2013-04-24

I was always taught to check your square after purchase and periodically. This is an excellent rough in-store test. I am curious though, how do you know that BOTH of the upper squares in the photo above are out of square? Could it be just the one on the right? Or Left?

Phil B (author)Galway2013-04-24

Since they are the same model by the same maker, it is reasonable to assume they are identical. But, a good first check would be to lay one over the other and see if they have the same profile. If that is the case, a "V" gap between them would mean both are inaccurate. Even if only one was out of square, not knowing immediately which one would make me cautious about buying either one. It is like the story of a man who wanted to discourage people from stealing watermelons from his patch. He put up a sign that said, "Caution, one of these watermelons is poisoned." He came back one day and the sign said, "Caution, two of these watermelons are poisoned."

woodlina (author)2013-04-24

Phil . . . I have walked right past those L-Brackets for 30 years and never once thought about using them as a true square. or to test my old square. Your simple brilliance never ceases to amaze me. (( YOU DAWG YOU!)) This one is another home run! You have earned your engineering degree right in your garage! -Alan

Phil B (author)woodlina2013-04-24

Alan, thank you. Say nice things like that about me and you will make me harder to live with than my wife already finds me. My favorite story about walking past things involved Edison wondering why a black spot formed on the inside of the globe in his lightbulbs as they got closer to burning out. He did discover a galvanometer showed a faint current trickle between the filament and a piece of metal on the inside of the globe when the bulb was operating, but Edison thought it was all merely a flaw to be corrected so he could give his customers a longer lasting, better product. It was Lee DeForest who discovered he could put a thin wire in the space between the filament and that piece of metal, and he could either accelerate the stray current flow between the filament and the metal by adding a positive electrical charge, or diminish it by adding a negative electrical charge. DeForest called it the Audion. We know it as the electron tube and radio took a giant leap forward. My second favorite such story would be the Nevada prospectors who cursed the useless black sand that clogged their gold mining equipment. One day someone took some of that worthless sand to the assayer's office, learned it was silver, and The Comstock Lode was discovered. I was asked to speak at my son's 8th grade graduation ceremony. I told the Edison/DeForest story and encouraged the kids to be curious about things other bypass without notice.

rolfy12 (author)2013-04-24

Now I know how I can test my square, which I currently suspect of being off.

Phil B (author)rolfy122013-04-24

Thanks for looking. This Instructable describes in more detail than I have given in the comments how to use the square to draw what should be parallel lines. Once you already own the square, this is a good way to check it.

emerson.john (author)2013-04-24

I have adjusted them with a carbide tipped scratch awl, just tapped a few dimples in the appropriate spot to increase or decrease the included angle.I suppose the squares started off as truly square, then were slightly bent from true when the scales were stamped on. A nice, practical instructable, as always. Didn't even make me want a welder this time.

Phil B (author)emerson.john2013-04-24

You make me smile. Thanks for looking and for commenting.

knife141 (author)2013-04-24

Nice job, Phil. I never thought about checking my square to see if it really is square.

Phil B (author)knife1412013-04-24

A lot depends on what you need from your square. Years ago I converted my electric handsaw or circular saw for use as a table saw. I set my homemade miter gauge with a square. Then I placed a framing square against the miter gauge to set an end cut from a sheet of plywood as the rip fence. (See steps 14 and 18 in the link above.) Any inaccuracy was compounded and was noticed immediately as the blade heeled, slowed from binding, and smoked during a cut (if inaccuracies were really bad). It became important for my framing square to be accurate. Shortly after buying my framing square I discovered it was not really square. Unfortunately, even if a square is accurate when it is new, workshop accidents and rough handling can cause them to go out of square eventually.

I do have a word of caution. Chances are your square may vary a little from being square. It may not have mattered for your projects up to now. Once you know one of your squares is not really square, you begin to lose confidence in it and you can become obsessive about getting it squared up. Once you already own the square, making it accurate may mean using a ball peen hammer on an anvil, or it may mean carefully using a file. As I mentioned in response to Emcy Square, I like to tape some paper on a surface. Then put down something I believe to be straight and hope it actually is. This straightedge should not be on a higher plane than the paper or inaccuracies begin to lurk. I make two lines on the paper only 1/16 of an inch apart from one another. If the lines are not absolutely parallel through their entire length, I know the square is not really square.

I have a little combination square made by a reputable maker of layout tools. It was sold by one of the big box home improvement stores. I paid more than enough for it. It was off around 1/32 of an inch over 12 inches. I used the side edge of a thinner file to make a stroke inside the frame of it and check it until it was actually square.

Anyway, thanks for looking and for commenting.

EmcySquare (author)2013-04-24

Great !!
I knew the "3-4-5" rule, but that's definetly easier

Phil B (author)EmcySquare2013-04-24

I used the 3-4-5 rule very successfully years ago when my wife wanted to lay floor tile in adhesive squares and we needed some reference lines. I think I made the numbers 6 feet, 8 feet, and 10 feet. Any inaccuracies became a small amount compared to the grand scale of things. The 3-4-5 rule is a little more difficult to use when the scale is 3 inches, 4 inches, and 5 inches. If measurements are off somewhere by 1/16 of an inch, it matters. For smaller things like this, my previous favorite method has been to tape some paper to a surface, lay down something I am pretty confident is straight, and scribe lines next to one another while flipping the square for the second line. The lines should be parallel. This is quite accurate, but not really possible until after the square has been purchased.

About This Instructable




Bio: I miss the days when magazines like Popular Mechanics had all sorts of DIY projects for making and repairing just about everything. I am enjoying ... More »
More by Phil B:Make an Electric Motor Run AgainMotor Made New McGyver StyleCuptisserie With Four Spindles
Add instructable to: