Question about being squared

Joined
Jun 23, 2024
Messages
88
So this afternoon I cut some 3/4" MDF down for a workbench top that I'm making.  I used my Makita track saw and two 55" guide rails.  I also used my TSO guide square. 

First I cut 1/4" off one long edge as a reference edge.  I then measured 47" from my reference edge and made a parallel cut down the other long edge of the MDF.  Next, I took my TSO guide square (squared against my first reference cut) and cross cut another 1/4" reference cut, this time along the short edge of the MDF.  I then proceeded to measure 67" from this short side reference cut and made a mark.  I then moved my guide square and guide and lined it up on this mark, making sure it was squared to the original long edge reference edge and made my final cross cut. 

When I measure using the corner to corner method, one is a hair over 79 7/16" and the other is just a hair under 79 1/2".

97_2.jpg


97_1.jpg


Not perfectly square, but somewhere between say 1/32 and 1/16 off..

When I measure the lengths of the two short sides, they are off by 1/16.  I can live with that for this bench top.  What I don't understand is when I measure the long sides, one of this is 1/8 longer than the other.  What throws me off even more, when I put either my Woodpeckers 851 square of my large 48" drywall T square against all of the corners, they all show as square.

How is it possible that my diagonal is less than a 1/16 off and all four corners checkout as square, but that one long side measures 1/8 longer?

Square.png


 
Maybe your parallel cut wasn't parallel?

I usually cut the long rip first, square the short end, and dimension off of that. 

Your system is sound and your cuts are very accurate. Frustrating as it is, you're close to perfect......

All the Best.
 
When I measure the lengths of the two short sides, they are off by 1/16.  I can live with that for this bench top.  What I don't understand is when I measure the long sides, one of this is 1/8 longer than the other.  What throws me off even more, when I put either my Woodpeckers 851 square of my large 48" drywall T square against all of the corners, they all show as square.

It's hard to measure the true squareness of a 64" x 47" piece using an 8" square. You'd need a dark room and strong backlight, or feeler gauges, and even then may be difficult. Even your drywall square, if actually square, would need careful holding against an edge.

What's interesting is that despite being off ⅛" on one set of edges and 1/16" off on the other set, your diagonals are only about 1/32" off. That's some serendipity at play.

There could be all sorts of things at play. Maybe the Makita rails aren't straight enough, especially since you have a join for the rip cuts. Maybe the rails moved while cutting or when being clamped. Maybe just a sum of various little errors here and there.
 
By chance, are you going to make this an mft grid top? If so, you may not need to have these pieces square at this point. After you have an accurate hole pattern large nearly perfect 90 degree cuts become a much more attainable objective.

I always drill my holes on a slightly oversized piece, then trim to center things up at the end.
 
Just a thought, how have you got the hook part of the tape measure secured when you're taking your measurements?  (Any of the measurements).

Is the hook sitting at 90degrees to the edge with the tape sitting flat, or have you accidentally pulled the hook so tight it's leaning over the edge slightly?

When you're measuring the diagonal lengths have you got the centre of the hook on the corner of the board?  You can't actually measure a true diagonal with a tape measure because the true diagonal is right along the centre of the tape.

I've struggled with this sort of check.  If I'm checking diagonals I use a small METRIC tape measure which I position with the centre of the hook on the corner and then sweep the tape across the opposite corner to get range of measurements.  The range of numbers is normally within a range of a few mm's.

Regards
Bob
 
waho6o9 said:
Maybe your parallel cut wasn't parallel?

I usually cut the long rip first, square the short end, and dimension off of that. 

Your system is sound and your cuts are very accurate. Frustrating as it is, you're close to perfect......

All the Best.

It is quite possible that the parallel cut was not exactly parallel.  That would/could account for the short edges being 1/16" off, but not for the long edge being 1/8" off.  I do have a set of the TSO parallel guides, but unfortunately not the 50" version.  I guess I need to move that towards the top of my list of other things to buy (I've got several more workbenches I want to build in the near future).

smorgasbord said:
It's hard to measure the true squareness of a 64" x 47" piece using an 8" square. You'd need a dark room and strong backlight, or feeler gauges, and even then may be difficult. Even your drywall square, if actually square, would need careful holding against an edge.

What's interesting is that despite being off ⅛" on one set of edges and 1/16" off on the other set, your diagonals are only about 1/32" off. That's some serendipity at play.

There could be all sorts of things at play. Maybe the Makita rails aren't straight enough, especially since you have a join for the rip cuts. Maybe the rails moved while cutting or when being clamped. Maybe just a sum of various little errors here and there.

This morning I ran my measuremtns though the old A²+B²=C² formula and 1/32" on the diagonals comes out to nearly an 1/8" on the long legs at those lengths.  So the math does work out.  This morning I also marked the longer side 1/8" shorter, so that it measured the same as the other side and remeasured the diagonals using this new mark as the "corner".  Doing this made the new diagonal measurements a full 1/16" off.  That would imply that it would be even further out of square.  Seems odd to me.

I'm guessing that second cross cut moved on me while making it.  It was the only cut that had me leaning across the full width of my current 4x8 workbench I was cutting on top of.

makpacman said:
By chance, are you going to make this an mft grid top? If so, you may not need to have these pieces square at this point. After you have an accurate hole pattern large nearly perfect 90 degree cuts become a much more attainable objective.

I always drill my holes on a slightly oversized piece, then trim to center things up at the end.

This benchtop is not going to be and MTF style, it is for my new CNC machine.  But your advice will be helpful for when I use the CNC to make my 4x8 MFT track saw workbench.

bobtskutter said:
Just a thought, how have you got the hook part of the tape measure secured when you're taking your measurements?  (Any of the measurements).

Is the hook sitting at 90degrees to the edge with the tape sitting flat, or have you accidentally pulled the hook so tight it's leaning over the edge slightly?

When you're measuring the diagonal lengths have you got the centre of the hook on the corner of the board?  You can't actually measure a true diagonal with a tape measure because the true diagonal is right along the centre of the tape.

I've struggled with this sort of check.  If I'm checking diagonals I use a small METRIC tape measure which I position with the centre of the hook on the corner and then sweep the tape across the opposite corner to get range of measurements.  The range of numbers is normally within a range of a few mm's.

Regards
Bob

I feel like my tape measuring discipline is pretty solid, but I can of course be making some small mistakes.  I will also swing the tape to show the range of measurements.  This afternoon I'll put together my Woodpeckers story stick pro and measure the diagonals with it.
 
woodferret said:
Psst.... your rails aren't straight.

That could very well be the case.  I do however always check them with a Woodpeckers 33" straight edge after putting them together.  Perhaps I need to check them with a longer straight edge?
 
No, it isn't possible for all four corners to be square and get those measurements.

"When I measure the lengths of the two short sides, they are off by 1/16."

What are the lengths? to the 32nd?

Looking your pics of the diagonal measurements it looks like they are only 1/32" off.

So, I wonder if you remeasured the original long side. Is it still 67" exactly?

If you give us the actual measurements of all the sides and diagonals to 1/32" or better we can figure out what the angles are.
That might determine if you have an equipment problem or if it's a procedure problem.

A 47" crosscut that depends on the TSO guide rail square is pushing it. Especially using a Makita guide rail.
At 47" an error of 1/10th degree will result in a greater than 5/64" error at the far end. (fat 16th)

An 8" square, no matter how precise, and a drywall square, aren't good enough to determine if your corners are really square.

Given your setup, to make a square rectangular panel I'd use geometry to lay out the corners.
Success will depend on the guide rail being straight and the splinter guard being freshly trimmed.

When I need to set the guide rail to a line and I know the splinter guard is worn I stick a Post-It note on the bottom of the splinter guard and run the saw to trim it.
 
If your lengths are not equal, the diagonal measurement method is just chasing your tail. It means nothing.

That said, I'm in the same camp as the others.  Although my experience with Makita tracks is limited and anecdotal, it's also not good. I've never had two of them at the same time (to join them) but neither was straight. They were sent back, along with the saw, and the company bought a TSC55. No one has needed to make a longer cut with it yet, so that's the only track they have, but at least it's straight.

I would start by checking the tracks, individually. Only then start testing the connection.

Assuming that your 47" dimensions are dead-on, I would guess the right side of your diagram is square. The 67 1/2" length is long because the upper left corner is over 90. Maybe the rail/square were not placed properly? Maybe the "toe" of the saw pulled it out as it was cutting?
 
You don’t need a precision square to make square cuts.

You need high school geometry.

A 3,4,5 triangle can be enlarged to give you right angles in any size.  A hand held calculator and an accurate ruler will do the trick.

Then all you need to do is extend the diagonal.  It is slow, but accurate.  Sometimes we forget that when we measure diagonals, we are using geometry.  There are other applications where geometry is your friend.

However, when I made my MFT-style benches (3), and I worked with the best accuracy I could manage, each top had one hole that was out of alignment.  A different location for each top.  I resolve this by drawing a red circle around the offending hole, and I work around it. 

However, I don’t use my tops for measuring, just for hold-downs.
 
Michael Kellough said:
No, it isn't possible for all four corners to be square and get those measurements.

"When I measure the lengths of the two short sides, they are off by 1/16."

What are the lengths? to the 32nd?

Looking your pics of the diagonal measurements it looks like they are only 1/32" off.

So, I wonder if you remeasured the original long side. Is it still 67" exactly?

If you give us the actual measurements of all the sides and diagonals to 1/32" or better we can figure out what the angles are.
That might determine if you have an equipment problem or if it's a procedure problem.

A 47" crosscut that depends on the TSO guide rail square is pushing it. Especially using a Makita guide rail.
At 47" an error of 1/10th degree will result in a greater than 5/64" error at the far end. (fat 16th)

An 8" square, no matter how precise, and a drywall square, aren't good enough to determine if your corners are really square.

Given your setup, to make a square rectangular panel I'd use geometry to lay out the corners.
Success will depend on the guide rail being straight and the splinter guard being freshly trimmed.

When I need to set the guide rail to a line and I know the splinter guard is worn I stick a Post-It note on the bottom of the splinter guard and run the saw to trim it.

I'll try and get some better measurements/photos this evening if I have time and post them up.  The only precision measuring device I have beyond my 1/16 tape is the Woodpeckers story stick pro.  It only has a resolution of 1/16 or 1 mm. 

That is a good tip about the post it note, I'll have to give that a try.

Crazyraceguy said:
If your lengths are not equal, the diagonal measurement method is just chasing your tail. It means nothing.

That said, I'm in the same camp as the others.  Although my experience with Makita tracks is limited and anecdotal, it's also not good. I've never had two of them at the same time (to join them) but neither was straight. They were sent back, along with the saw, and the company bought a TSC55. No one has needed to make a longer cut with it yet, so that's the only track they have, but at least it's straight.

I would start by checking the tracks, individually. Only then start testing the connection.

Assuming that your 47" dimensions are dead-on, I would guess the right side of your diagram is square. The 67 1/2" length is long because the upper left corner is over 90. Maybe the rail/square were not placed properly? Maybe the "toe" of the saw pulled it out as it was cutting?

Revisiting my track saw setups might be in order as well.
 
I do want to understand what has caused this and how I avoid it in the future, but since this is a simple bench top to set my CNC machine on, is there harm in going ahead with edge banding (I'm thinking some type of hard wood) and putting the laminate on it?
 
Packard said:
You don’t need a precision square to make square cuts.

You need high school geometry.

A 3,4,5 triangle can be enlarged to give you right angles in any size.  A hand held calculator and an accurate ruler will do the trick.

Then all you need to do is extend the diagonal.  It is slow, but accurate.  Sometimes we forget that when we measure diagonals, we are using geometry.  There are other applications where geometry is your friend.

However, when I made my MFT-style benches (3), and I worked with the best accuracy I could manage, each top had one hole that was out of alignment.  A different location for each top.  I resolve this by drawing a red circle around the offending hole, and I work around it. 

However, I don’t use my tops for measuring, just for hold-downs.

If the measuring or cutting is off, does it really matter how you layout it out?  I think my issue resides in one of those two fundamentals :'(

I do plan to make two MTF style tables, but hopefully the CNC will get the holes correct and then I can use them to trim the table top square.
 
sawdust-samurai said:
I do want to understand what has caused this and how I avoid it in the future, but since this is a simple bench top to set my CNC machine on, is there harm in going ahead with edge banding (I'm thinking some type of hard wood) and putting the laminate on it?

What your CNC sits on doesn't matter ...  It being flat would be nice though... Although it could be a circle.  Your spoilboard won't have to be square either as you go through the tramming operation to align your axis to each other, not anything physical.  You have the set up and operations manual?
 
TSO square is great and close - but I find it is off a bit no matter what - especially crosscuts on 4x8 plywood - just not enough surface area - but it does get you dang close - I usually cut to 1/4" over and then use the table saw to make everything square.
 
woodferret said:
sawdust-samurai said:
I do want to understand what has caused this and how I avoid it in the future, but since this is a simple bench top to set my CNC machine on, is there harm in going ahead with edge banding (I'm thinking some type of hard wood) and putting the laminate on it?

What your CNC sits on doesn't matter ...  It being flat would be nice though... Although it could be a circle.  Your spoilboard won't have to be square either as you go through the tramming operation to align your axis to each other, not anything physical.  You have the set up and operations manual?

I was more thinking about the slight out of square effecting the edge banding.  I do have the manuals for my CNC, but I've not looked at them yet since I'm still building the table for it.

Hipplewm said:
TSO square is great and close - but I find it is off a bit no matter what - especially crosscuts on 4x8 plywood - just not enough surface area - but it does get you dang close - I usually cut to 1/4" over and then use the table saw to make everything square.

That's good to know.
 
I used a Festool track saw decades ago and way before any accessories such as available now were even thought of and abandoned the whole idea due to the accuracy problems it presented then. Thinking about it now, I would try to set up a reference corner and then measure the diagonal from that using two story sticks, would that be feasible? I can't think of one good reason to own a track saw but I can understand why they are popular so I can't try the idea.
 
sawdust-samurai said:
I was more thinking about the slight out of square effecting the edge banding.  I do have the manuals for my CNC, but I've not looked at them yet since I'm still building the table for it.

If there's a bow, it's going to be so small that internal wood tension of your edge banding dwarfs it.  Just keep it sane to 1/8 - 1/4 and glue/clamps it'll be a non-issue.  I tend to be lazy and just buy the 1/4 poplar from HD's hobby stack rather than ripping my own.  It's more stable and I don't have to deal with milling down 4/4.
 
Mini Me said:
I used a Festool track saw decades ago and way before any accessories such as available now were even thought of and abandoned the whole idea due to the accuracy problems it presented then. Thinking about it now, I would try to set up a reference corner and then measure the diagonal from that using two story sticks, would that be feasible? I can't think of one good reason to own a track saw but I can understand why they are popular so I can't try the idea.

I only have one story stick, so not sure I can try that method.  I also think I need to rule out the saw and or track from being the cause as well.  Otherwise, all the layout, marking and measuring in the world is not going to produce a square result.

woodferret said:
If there's a bow, it's going to be so small that internal wood tension of your edge banding dwarfs it.  Just keep it sane to 1/8 - 1/4 and glue/clamps it'll be a non-issue.  I tend to be lazy and just buy the 1/4 poplar from HD's hobby stack rather than ripping my own.  It's more stable and I don't have to deal with milling down 4/4.

Thanks for the additional info.  I was going to use some stock from either HD or Lowes as well.
 
Back
Top