Checking My Combo Sqaure

Viewing 15 posts - 1 through 15 (of 17 total)
  • Author
  • #308909

    I just purchased a a twelve inch Starrett combo square and I want to check it. I have done the line across a board thing and it seems ok but I want to be sure.

    I saw that Paul recommends using 2 boards. Why the 2 boards? I own his newest book and it still isn’t clear to me. I also saw the article on this site by Paul also recommending using 2 boards to check the square.

    I’d like to try this because I trust that Paul knows his stuff and this will be good to do once I understand the procedure; why cut the board? Why not just leave it and then check the square by flipping it after you’ve made a mark?



    Whilst combo squares – good quality ones – are perfectly adequate for most everyday woodworking purposes, they are not precision instruments.
    For that you need certified engineers’ squares which are extremely expensive and over-the-top for woodworking, in my opinion.

    The narrow-parallel-lines test, done with a sharp marking knife at about 1mm width for the full length of the rule section, will show up any significant inaccuracies to the naked eye, but with combo squares, be sure to do it on a dead straight wooden edge, checking all the combinations that you can load the straight edge in the yoke of the combo.

    One useful gadget that is often overlooked, is the digital angle block. This will give you repeatable accuracy to a fraction of a degree, if you need it.

    You put the block on one surface, hit the zero button, then replace it on the adjacent angled surface. A right angle will show up as 90.00 degrees. Anything else is not square and requires your judgement as to its acceptability.

    Not expensive, and worth investing in for getting into tight spots that you can’t reach with a conventional square.

    Good luck


    Thanks for the response. I still don’t get the 2 board thing. I have a feeling its very important and probably solves the issues of squaring up. Just want to understand Pauls method.


    Me too.

    I don’t see where two boards fits in with checking a combo for squareness, unless it’s a way of double-checking by using one board to outline any discrepancies in straightness with another……

    The job can be done with one board, as long as it has a good straight edge to register the flat of the combo stock and is wider than the ruler section is long. Ply or MDF will do very well…. it just needs one edge to be planed dead straight.


    I have rechecked and rechecked this thing and it is out of square over 10 inches by a full 32nd. Brand newish Starrett. It’s been sitting around for a couple years in the box and put away so there’s no way I can return it; there’s a 100+ down the drain I guess.

    Maybe someone will clarify the 2 board technique so i can check it that way.


    Using two boards is developing a straight edge if you do not have one

    You do not have to have both boards if you have a known straight edge.

    Paul is developing a perfectly straight edge with the two boards so he can measure from a straight edge

    A check with a piece of Heavy card stock or paper will proof the square as easily as milling two boards to get a straight edge
    He has also said that MDF and ply factory edges are straight.
    You can also verify your square by marking the edge of a square board then turning the square 90 degrees and checking the angle with the 3,4,5, method 3 up 4 over and 5 from point to point or any combo of the numbers. .

    Richard Kelly

    Here’s a slightly cumbersome test for an Engineer’s Square – it’s easier with a big wooden square – but if you are careful it does work.

    By Pythagoras’ theorem of 3-4-5 for a 90deg triangle, if we have 4″ of ruler sticking out of the Square, and if we measured 3″ along the casting and scribed a tiny line, the distance from that line to the tip of the rule must be 5″ EXACTLY.

    So using a pair of sharp dividers, set them to exactly 3″ using the engraving of the rule to ‘catch’ the tips precisely. Then put the tip into the corner of the rule/casting and scribe a tiny line on the casting at 3″ along.

    Then using the graticules on the rule again, set the dividers to 5″ EXACTLY, and see if distance from the tip of the rule you set to 4″ to the scratch is indeed 5″. If it’s not then it’s not square.

    (Make it a little easier by having more then 4″ of rule sticking out- say 5″ – then again you can line up with the graticule at 1″ with the dividers, as opposed to trying to catch the end)

    If the square is 1/32″ out over 10″ as in the OP’s then you will be able to spot the error.

    Fiddly as I say, but it is surprising how carefully this can be done as the graticules on the rule make it easier to set dividers very accurately.

    This does work a treat on nice big items like old big carpenter’s squares where you can scale up to amplify the error and also where even a good Engineer’s Square is too small to test squareness over a long distance.


    The nice thing about a combination square is that it can usually be trued relatively easily. It’s annoying that a square from a manufacturer like Starrett isn’t spot on, but at least you can do something about it. Look on YouTube for tutorials.


    I can’t clarify the two boards as I’m not quite sure of it myself but if it is out of square then you can correct it. If you pull the rule out you’ll see it seats on two faces either edge of the slot it runs in. You just need to file one face to bring it back into square. Go carefully and remove a tiny amount at a time, your 1/32″ translates to a very small amount by the time it gets to the stock.

    Having said all that I am surprised that a Starrett square is out but I suppose there are exceptions. You haven’t wasted your money anyway, just a little fettling needed.

    • This reply was modified 4 years, 9 months ago by cragglerock.
    Hugo Notti

    Another way to check for a square angle dates back to the ancient Egyptians and is perhaps even older, but the Egyptians left evidence:

    If you make a triangle with side lengths of 5,4 and 3 units (inch, millimeter, parsecs, whatever you like), the angle between the line with 3 and 4 units will ALWAYS be 90° – square.

    So, when you measure 3″ on one side of your square (handle) and 4″ on the other side (ruler), the distance between these points should be exactly 5″.

    Why? Because 3² + 4² = 5² (9 + 16 = 25), and Pythagoras showed, that a² + b² = c² on all triangles with a square angle between line a and b.

    Of course, this leaves you with the question, how accurately you can measure the distance, and how accurate is your ruler. It is always a chicken-egg problem.

    So here is one suggestion: How precise is your working with chisels, saws, planes etc.? does 1/32″ really matter on a length of 10″? Paul often says “we strive for perfection”, which also means, “we aren’t perfect yet”.



    Thanks for the responses. Much appreciated. I have checked this thing at least a dozen times, if not more, and it is indeed out of square over the 10 inch length by a full 32nd; this is a brand new Starrett 12 inch combo square. I have tried it on numerous boards and edges, factory and fresh from the saw.

    I think that one poster here on this thread has helped cleared up the 2 board technique; I think the purpose was simply to create a straightedge; that makes sense and it was something I considered but wanted to be sure. Thanks for the help everyone.

    I looked on YT for help on correcting this issue but couldn’t find one that dealt with a Starrett. I found cheap brands where they filed down the lands, but not a Starrett. In fact I don’t even see lands, except simply for the entire area where the square slides into. I’ll keep investigating and try to salvage this thing; so disgusting when you spend so much money to avoid these issues and the issues remain.



    Might be worth contacting Starrett and explaining the situation. They might do something if appeal to their sense of pride 😉


    That would be my next option . Most companys don’t like to have bad products. I think they are guarenteed to 2 tho over the length of the beam. It might just be a warrenty issue.

    • This reply was modified 4 years, 9 months ago by deanbecker.

    I fixed it! Just took some very light sanding with fine sandpaper with a thinner steel rule as a guide. I took the square apart completely so I had access to the lands and sanded one side very lightly. I checked for square after every sanding; it took 4 times to get it so square that my eyes cannot see the difference of the very sharp pencil lines.

    I checked inside and out over the 10 inch accessible length. Looks great now!

    Thanks for the help everyone.

    BTW, I had the option of a cast iron head or hardened steel head, I chose the cast iron because it was cheaper. Thank goodness I did that because I’m not sure I could have sanded the hardened steel if it was out of square. Of course perhaps the hardened steel head is better quality and would never require touch up.


    Participant Is a quick video on bring one back to square. I have had to true several older squares I have had up. including flattening the beam.

    Tuscloosa, Alabama
    Lung T'an Hu Huesh Kung-fu Woodshop

Viewing 15 posts - 1 through 15 (of 17 total)
  • You must be logged in to reply to this topic.