How square is square enough?

[Dan McGonigle]

Started on a workbench build recently. Dimensioning 12/4 poplar for the base. Face joint - thickness - edge joint - rip. Pretty standard. I squared my jointer fence to the bed using my 12” Starrett combo square. Similarly, I squared the table saw blade to the table using the Starrett. When all was said and done, I found that I had two 90 degree edges and two edges that were slightly out of square. These pieces are 2 3/4” by 3” and 4” for reference. After thinking it over, and checking all my setups agains, I came to this conclusion - when having to rip thick stock, I typically rip it on my bandsaw, 1/16 or 1/32 oversize, then clean up the edge on the table saw. I know this seems like extra work, and it is, but I hate pushing thick stock over my 2hp table saw and hearing it bog down and struggle. My bandsaw equipped with a 3tpi resaw blade goes right through 3” stock with ease. So I concluded that ripping 1/32” or 1/16” off the edge of a board is causing my table saw blade to deflect away from the workpiece, considering there’s nothing on the other side of the blade to counteract the deflection. To test this, I ripped my workpieces thinner, just by a few thousandths. This time, in an attempt to reduce the amount of deflection, I pushed the stock through very very slowly - ending up with some burn marks - but alas - the piece came closer into being perfectly square. Using my Starrett and feeler gauges, I found that some pieces are out of square by 0.002” over 3”-4”. I’m satisfied with that. I wish it was perfect, but it may just be the limitations of my machines. I often have to remind myself to not become obsessive over chasing “square”.

Bottom line. 0.002” over 3”. Is that square enough for you?

[Brian Holcombe]

Plane it to size.

A combo square is not uncommonly out of square, so if you are using that to square your machines then you may be doubling your error.

[Ryan Yeaglin]

Plane it to size.

A combo square is not uncommonly out of square, so if you are using that to square your machines then you may be doubling your error.

While this might be true of a common square from the hardware store, A Starrett square is a precision square. Although dropping one will take it out of square.

.002" of an inch is less than common machine shop tolerance =/- .005, So I would say it's fine unless your building a space shuttle. Wood moves way more than that from seasonal changes in humidity.

[Jim Becker]

Ryan, yes Starrrett squares that are kept carefully can be pretty accurate, but please trust that Brian knows what he is talking about when it comes to machine setup including in a machine shop! Check out his metrology thread when you have the chance. A combo square due to its nature can have some variance. I agree that on the wood side it's not such a big deal because as you note, wood is an ornery thing. But cast iron and steel are different and that little bit of variance cane absolutely affect machine setup. A fixed, high quality square is better for that task as a result.

OP, I also tend to dimension width using the planer.

[Phil Gaudio]

You are approaching the practical limit to woodworking measurement. Yes, this would be square enough for me.

[glenn bradley]

Generic answer is that if it is square enough for what you are building, it is square enough. Just to confirm your bandsaw ripping protocol . . . You joint a surface flat, then joint a surface square to that. You then put the flat surface on the table of the bandsaw and the perpendicular (square) surface to the fence and rip at the tablesaw. I tend to true-up bandsawn surfaces at the planer using the opposite flat side (that was riding the bandsaw's fence) as my reference surface but, there are several ways to get to where you want to go.

Even more important in avoiding an endless trip down the black hole of perfection; wood moves. Even if your machines are perfect to a machine-shop level of alignment, the resulting material may not reflect this perfectly ;-) Don't get me wrong. I am one of those guys that aligns to .001" whenever possible. I like to have my machines as close as I can get them. That way I know if there are deviations it is technique or the simple movement of stress release in the material. This is why I (and others I am sure) approach milling material as a multi-step process. I mill to oversize and let things rest. I then mill to final dimensions very shortly before I assembly things.

[Frank Drackman]

I layout all of the boards that I am going to glue into a panel, put the cabinet makers triangle on them, and mark each edge of the joints alternately out/in. I go to the jointer and run each edge away from the fence or toward the fence. This cancels any error that my jointer fence is not exactly 90 degrees.

[Patrick Walsh]

The answer is as my new boss saiz “it’s either square or it’s not”.. and I totally agree.

Out of square compounds. It’s not that hard to make things square unless you are building stuff that kinda trash to begin with. I’ve build miles and miles of cabinets “not furniture” face frame style and it’s out of square slightly like 90% of the time as there’s no real joinery. Just all screws and glue and aligning stuff the best you can by feel and with clamps..

[Andrew Hughes]

I remember a time when was starting out . Nothing was square my saw wouldn’t cut straight or square none of my tools could measure square. My bench wasn’t flat. I couldn’t find a square corner on anything.
But I endeavored to persevere now my cutoffs and work pieces are pretty good. And if I really try I can make something square big or small.
Sometimes smaller stuff is harder because I have to use hand tools.

[Robert Hazelwood]

I agree with Brian that the planer is a better tool for dialing in the width for this size part. I use the planer for this whenever I can.

To your general question, is 0.002" over 3" square enough, it depends. It will work for many things, but would be a problem for others.

[Dan McGonigle]

I agree with Brian that the planer is a better tool for dialing in the width for this size part. I use the planer for this whenever I can.

To your general question, is 0.002" over 3" square enough, it depends. It will work for many things, but would be a problem for others.

I had thought about using the planer to dimension the width, but the length of these pieces are so close to their finished length, any bit of snipe would be a big problem. I try to rely on hand tools when accuracy is paramount, like in small or delicate pieces. In reality, I could take a few plane shavings off these boards and probably dial them in closer to perfection. The parts I'm working on are for a workbench base, so I think the level of square is acceptable. It just annoys me when I triple check all my machines and end up with anything other than perfect. Like I said before though, it may just be the limitations of my tools.

[Robert Hazelwood]

I had thought about using the planer to dimension the width, but the length of these pieces are so close to their finished length, any bit of snipe would be a big problem. I try to rely on hand tools when accuracy is paramount, like in small or delicate pieces. In reality, I could take a few plane shavings off these boards and probably dial them in closer to perfection. The parts I'm working on are for a workbench base, so I think the level of square is acceptable. It just annoys me when I triple check all my machines and end up with anything other than perfect. Like I said before though, it may just be the limitations of my tools.

You may want to see if you can dial out some snipe from the planer. For the most part on my humble DW734 snipe is very minimal, difficult to see or feel, and comes out with a pass or two with the smoothing plane (which I would be using on the surface anyways).

Another thing that occurred to me is to check that the planer is cutting parallel. A few thou out across the width could cause some squareness issues that you might be blaming on the TS. Although your theory about the TS blade deflection on a skim cut makes sense. What kind of blade? I would venture to guess that a full kerf dedicated rip blade would be least prone to deflecting like this.

[Dan McGonigle]

You may want to see if you can dial out some snipe from the planer. For the most part on my humble DW734 snipe is very minimal, difficult to see or feel, and comes out with a pass or two with the smoothing plane (which I would be using on the surface anyways).

Another thing that occurred to me is to check that the planer is cutting parallel. A few thou out across the width could cause some squareness issues that you might be blaming on the TS. Although your theory about the TS blade deflection on a skim cut makes sense. What kind of blade? I would venture to guess that a full kerf dedicated rip blade would be least prone to deflecting like this.

Maybe I’ll run a board through the planer and measure thickness on each side with a caliper, see if it’s tapering along width at all.

TS blade is a Freud Premier Fusion 40T. I have a Freud thin kerf rip blade but I don’t use it too often, although It’s the better blade for ripping thicker stuff. I’m using a Grizzly 2hp table saw, The accuracy of this saw is not as high as a high dollar PM or SS or something. I accept that because that’s just what’s financially doable for me.

Regarding the comment that combo squares commonly are not square - I do have a set of “engineers” squares, that I use for set up sometimes. These squares are “square” visually if testing then by drawing a line, flipping, and drawing another line. They’re also square if standing up back to back with my Starrett. Those are about the only way I can check them. I’m not validating the accuracy of these tools, I’m just saying that I’ve checked them to the best of my ability, which to some may not be accurately checking them at all.

[Mark Hennebury]

You need 3 squares and a reference flat surface to verify squares.
Stand square #1 on the flat surface and check square #2 and #3 against it, if they fit perfectly together, that tells you that 2 &3 are both the same angle and that they are complimentary angle to #1, but not what the angles are.

Now check #2 and #3 against each other, if they fit perfectly, then all squares are 90.
If not you will see the error, and it will be half of the error between #2 and #3

Maybe I’ll run a board through the planer and measure thickness on each side with a caliper, see if it’s tapering along width at all.

TS blade is a Freud Premier Fusion 40T. I have a Freud thin kerf rip blade but I don’t use it too often, although It’s the better blade for ripping thicker stuff. I’m using a Grizzly 2hp table saw, The accuracy of this saw is not as high as a high dollar PM or SS or something. I accept that because that’s just what’s financially doable for me.

Regarding the comment that combo squares commonly are not square - I do have a set of “engineers” squares, that I use for set up sometimes. These squares are “square” visually if testing then by drawing a line, flipping, and drawing another line. They’re also square if standing up back to back with my Starrett. Those are about the only way I can check them. I’m not validating the accuracy of these tools, I’m just saying that I’ve checked them to the best of my ability, which to some may not be accurately checking them at all.

[Bob Jones 5443]

How long are the parts? Why not use a cambered hand plane to get them square?

Wait: who let the Neanderthal in?