Realistic square cut accuracy expectations?

[Hans Friedebach]

Whether using a table saw or a track saw, what accuracy do you expect every day for 90 degree cuts on sheet goods? - for a 30" cabinet side, for example.
What accuracy do you regularly achieve and how do you measure it?
tell me if you're a working Pro or serious DIY and whether your response is basd on your table- or track saw usage - or no difference

I am asked this question by different people and want to be able to offer a wider experience based "consensus" - not just my own opinion
Hans

[Steve Eure]

When it comes to cutting anything, I'm, as my bride would say, "anal", when it comes to accuracy. Because I have a small shop, I generally cut my sheet goods outside on a table I made using a Bora Saw guide and a Craftsman circular saw. Using the guide is extremely accurate if my measurements are precise. Even with that, if the part is smaller, I'll cut it a 1/2" or so over and then take it to the tablesaw for precision. I generally get a better cut on the tablesaw because of having a better quality blade on it.

[James Zhu]

If the table saw or track saw cannot make square cut, it is useless for making fine furniture.

[Peter Kuhlman]

I assume you refer to the panel surface being square not the edges being square to the top.
I have no problems achieving a square cut with my Festool TS55 on my MFT/3 or on my Bridgewood cabinet saw with an Incra fence system. I mean well under 1/16" of square and most often better than that. It all depends upon your setup and the repeatability of your tools. Once set up I no longer verify the accuracy as both tools hold their calibration unless I make some major changes. I use an 18" triangle from Woodpecker to check setup on the MFT and a dial gauge for my table saw to check blade and fence.

[Cary Falk]

I expect better results on my table saw than my track saw. I use my track saw to break down plywood or the occasional odd angle cut. I have a JessEm sliding table on my table saw. How accurate am I? As close as my measuring device. What do I use to measure? A Woodpecker precision 26" square for the big stuff and a 12" for the smaller stuff. I am a hobbiest not a pro.

[Brian Holcombe]

Ive never gotten a finished cut off a track saw, nor do I expect one. They're usually square enough to work but definetly not comparable to a better piece of machinery or hand tuning.

Im working on a kitchen currently and one of the benefits of a track saw is that you can cut out of square with greater ease.

[Sam Murdoch]

For sheet goods - ply parts as you describe - 30" cabinet side - I generally use my track saw. Always rip off a new edge not the factory edge. Such a straight cut with the guide rail is typically more than straight enough. I find that it is essential, when making long track saw rips, that the supporting table be as absolutely flat as you can achieve so that the rail is not deflected in the length of the cut. My work surface is an 1-1/2" piece of MDF on strong back frame and solid horses (was hoping for a nice FELDER lift table :rolleyes:) and with a sheet of 1/2" insulation as my sacrifice board under my cut sheet.

The difficulty with the track saw/accuracy is in establishing square cuts on both ends of those 30" sides. By square in this case I mean tolerances of 1/64" or less. No parts more than that are considered accurate enough. I use the Woodpecker 26" square and careful layout. For multiple cross cuts of the same dimension I use a pair of guides that clamp to my rail and register off the end of my 1st square cut.

I don't hesitate to use a track saw for accuracy with plywood cuts or for tall doors that might be unwieldy on the table saw but I very rarely use the track saw for sizing doors or most other solid wood square cuts. Table saw with very accurate homemade sled is my preferred method. I wish I had a sliding table saw but I don't.

Full time professional but in no way a production shop guy is how I characterize myself. I take more time for set up than most others would tolerate but I work with my system and I do good work IMHO :).

[Hans Friedebach]

thank you all for your prompt and detailed responses.
I am hoping to get more data to provide a more statistically relevant sample.
I am curious at the reports that you can't get an accurate finished square surface cut with a track saw. I know a number of seriously professional remodelers who do it all the time. Have any of you had any experience like the one shown on this YouTube :https://www.youtube.com/watch?v=AmNyPvsfSCo

[Kyle Iwamoto]

I don't own a "track" saw, I have a guide that I use to rough cut the ply.
I expect my table saw to cut a square edge every time I use it. Assuming you start with one square edge. If this requires a trim cut then that's what I do. IMO if you set up your table saw, and it is a good one, you should be able to expect a near perfect cut every time. I'm not a pro. I do have a 52" saw, so this makes things a bit easier to rip 30".

[Frederick Skelly]

I guess I don't quite understand the question. I mean, I cut sheet goods square to the accuracy of my best square. I setup the table saw to that standard and check it regularly. If I happen to make a cut an it isn't square, it gets tuned. With the tracksaw, I cut oversize then put it on the tablesaw and cut to size. My joints align and fit tightly, so I think I'm getting it right?

[Bill Dufour]

You need to be much more specific. People here are saying it is square but I do not think anyone here is using a scraped in square to measure it. Same as people measure the flatness of a table saw with a old level not a precision straightedge . That kind of accuracy is not practical in wood since the wood moves with changes in humidity and temperature much more then the gauging stuff allows. I have checked my tools but not with precision meteorology gear since what i have is plenty good enough for the quality of my work and my machines.
i do not think anyone ever worries about holding measuring tool in their bare hands when measuring wood. It is an issue with precision metal working. There is a good reason all quality micrometers have a plastic handle on the arm.
Bill D
I have read that a difference in 1/1000 of an inch is the smallest you can see in a flat wood surface so anything that is good to 1/1000 or less is plenty good enough in wood. Holding a micrometer in your bare hands can easily expand it by 1/1000 in a few minutes.

A starrett 12" square is about $700, but it comes with certification.
http://www.starrett.com/metrology/product-detail/metrology/metrology-products/precision-measuring-tools/squares/Solid-Squares/20-12%20W~SLC

[glenn bradley]

When setting up a tablesaw sled I shoot for about .001" over 18" or so. Everyone has there "close enough". .001" over the length of my fence, blade parallelism or right angles over 18" or so works for me.

[Martin Wasner]

Running a sheet on the tablesaw, without a sled or slider, all you're doing is making a smaller version of the full sheet since you're working off of two edges rather than one. If the sheet isn't square, your part will not be square. Sheets are rarely square.

I don't expect much from a track saw, it's completely up to the operator setting the track down where ours going to end up

My panel saw was set up with .005" in a bit under 4' of cut. Where it's at presently for accuracy, I don't know, but I haven't noticed anything being off anything noticeable since I moved it. That number was determined doing a five sided cut to figure out which way to adjust it. Measuring for square you can 3,4,5 two edges, or just measure corner to corner, which is my usual go to method.

[Andrew Hughes]

I cut some very good Birch plywood with my portercable saw boss and a straight edge.The parts came out perfect and that was the best day of my life.
The proof was obvious when I assembled everything.Square and true
I still remember that day.:rolleyes:

[Frederick Skelly]

I'll be the odd man out again. I just don't believe that wood needs to be square to less than about 1/64". I'm sorry, I just don't buy that as "required". I've built many nice projects with tight joints, working as I do.

Now if somebody wants to work to that standard "just because they do", I respect that.:D
Fred

[Doug Hepler]

Hans,

I note that you joined SMC this month and that this is your second post. Welcome to SMC (and perhaps to woodworking as well). The angular accuracy of a crosscut (e.g., deviation from 90 degrees) depends on the length of the sides and the accuracy that you need. I usually want 1/32" accuracy. I get the crosscut sled, miter gauge, whatever, as close to 90 deg as possible using what is called the five-cut method.

I invite you to look at Chapter 18 in my book, Notes and Reflections While Shaving Wood (http://plaza.ufl.edu/chepler/). It's a free pdf, and there is no registration or any such thing.

Doug

[andy bessette]

When setting up a tablesaw sled I shoot for about .001" over 18" or so. Everyone has there "close enough". .001" over the length of my fence, blade parallelism or right angles over 18" or so works for me.

Yeah, right. Come on.

[Wayne Lomman]

Tolerances depend on the total dimension. If we are talking about sheet goods, the best way to check for square is by checking diagonals. It's simple geometry - if your sides are parallel the diagonals must be equal. If the diagonals are within half a millimetre, the job will work very well. Any slider will cut this accurately and repeatedly as I can attest from many years using one. In the absence of one I clamp a fence and rip with a portable circular saw.

If we start talking about dimensioning timber for chair components for example, this is where you need a precision try square and a vernier for checking sectional accuracy.

Back on sheet goods, use the 3:4:5 triangle to reset a square line if you need to start again after a wobbly cut. Or use the same ratio to set out your new shed slab etc. Geometry and maths principles like these are free, can't get stolen and don't wear out. Cheers

[Robert Parent]

For me, ripping on a tablesaw I have always been able to get a part as accurate as needed. Crosscutting a part is totally different ballgame.... Smaller parts are usually not a problem but the large the panel the more issues there are. After years of trying I have yet to discover the magic. I now have a CNC router so anything large in size or any part needing accuracy gets cut on the router, problem solved.

BTW: I just picked up a mini track saw but have not yet cut anything using it. I am expecting it to be about the same as a tablesaw assuming one takes care measuring and uses a good blade.

Robert

[Brian Lamb]

I routinely cut panels with my Felder slider that are within a few thousandth of an inch of being square. I use a precision square to set my sliding table fence, and you can use it to set the beam on a track saw if desired and achieve the same results.

And I beg to differ with those that say "it's only wood, it's close enough". If you do a stack of 8 drawers and you want the gap between drawers to stay consistently at say 3/32" (the thickness of my Starrett adjustable square blade that I use as a spacer). If you are even .010" out of parallel on each drawer front then by the time you stack 8 of them up, you are .080" taller on one side than the other, which is completely obvious to the eye. Out of square shows itself also when you have a row of doors and you are trying to maintain the same gap between doors. The human eye will detect as small as .005-.010" variance especially when it's got multiple gaps to look at and compare.

[David Kumm]

For those who don't know, Brian makes a precision square for this purpose. I use it every time I crosscut on a slider to verify the fence is square. I'm not affiliated but the square is worth a plug. Dave

[Frederick Skelly]

I routinely cut panels with my Felder slider that are within a few thousandth of an inch of being square. I use a precision square to set my sliding table fence, and you can use it to set the beam on a track saw if desired and achieve the same results.

And I beg to differ with those that say "it's only wood, it's close enough". If you do a stack of 8 drawers and you want the gap between drawers to stay consistently at say 3/32" (the thickness of my Starrett adjustable square blade that I use as a spacer). If you are even .010" out of parallel on each drawer front then by the time you stack 8 of them up, you are .080" taller on one side than the other, which is completely obvious to the eye. Out of square shows itself also when you have a row of doors and you are trying to maintain the same gap between doors. The human eye will detect as small as .005-.010" variance especially when it's got multiple gaps to look at and compare.

Brian, I think I understand your example. With all due respect, I think it's really kind of worst case and somewhat unusual. For example, I'll never build a stack of 8 drawers. (Be it a large or a small piece.) But if I do, I think overall movement of the piece will have a greater effect than my cut being 1/100" out of parallel - especially on a large piece. And I think I can tune the gaps between the drawers with a hand plane to correct 1/100". Remember, I'm not making interchangeable parts - each drawer only fits one slot.

Best regards,
Fred

[Mark W Pugh]

.........if your sides are parallel the diagonals must be equal. ...........

Not necessarily - look at a Rhombus. The opposite sides are parallel to each other, but the entire figure is not square. This is the problem we're trying to solve.

[Robert Engel]

Squareness is actually more critical in cabinetmaking than furniture making.

To answer you're question on a 30" panel 1/32 is the maximum error I would tolerate and I wouldn't be happy with that either. I rough cut and use a panel cutting sled on tablesaw.x

"Square" is determined by how accurately it is laid out not the type a saw system. All any track saw does is cut where you've placed it.

I build a lot of cabinets like Andrew I use a simple guide with a circular saw and clean up the cuts on the TS.

[Brian Holcombe]

I'll be the odd man out again. I just don't believe that wood needs to be square to less than about 1/64". I'm sorry, I just don't buy that as "required". I've built many nice projects with tight joints, working as I do.

Now if somebody wants to work to that standard "just because they do", I respect that.:D
Fred

It depends on how much time you can afford to spend adjusting. If I cut something out of square then I spend time adjusting it that I would rather be spending on the next phase of the work. In some cases this is expected in others it is simply avoidable rework.

[Prashun Patel]

I have a set of squares of various sizes that I once checked relative to each other to convince myself they were all square enough.

For any panel, leg billet side, or cut end, i use these, looking for light and feeling if it rocks. If it passes its good enough for me.

For better or worse I am too lazy to measure my degree of squareness or flatness. Good enough for my eye has been good enough so far.

[Kevin Jenness]

Better stated as "if the sides are parallel and the diagonals equal, the parallelogram is a rectangle." As Wayne says, for typical cabinet parts less than .5 mm difference between diagonals gets you in the ballpark. There is a reason why professional cabinet shops use sliding table saws, vertical panel saws, beam saws and CNC routers to produce consistently sized and squarely cut parts. Once set up to meet the required precision and used sensibly the results are predictable and further work is unnecessary.

[Frederick Skelly]

It depends on how much time you can afford to spend adjusting. If I cut something out of square then I spend time adjusting it that I would rather be spending on the next phase of the work. In some cases this is expected in others it is simply avoidable rework.

I agree Brian. I have time. You do this for a living. I get that.

[Brian Lamb]

Brian, I think I understand your example. With all due respect, I think it's really kind of worst case and somewhat unusual. For example, I'll never build a stack of 8 drawers. (Be it a large or a small piece.) But if I do, I think overall movement of the piece will have a greater effect than my cut being 1/100" out of parallel - especially on a large piece. And I think I can tune the gaps between the drawers with a hand plane to correct 1/100". Remember, I'm not making interchangeable parts - each drawer only fits one slot.

Best regards,
Fred


Hi Fred, well, I understand your point but I find I run into this issue all the time. I tend to make a lot of projects with flat doors and in a style where everything is tight and symmetrical. This picture below is pretty typical of my projects, it has 6 doors and 6 drawers, is over 7 foot tall and 7 foot wide, a linen closet in the master bath. All the doors and drawers have a gap of approximately 1/8", if any of my panels were off, you would see it immediately. As for fine tuning, you really can't do that after the product has been finished and you don't want to be touching up in the field, so better to be right the first time.

364990

This is another example, not a very good picture, but this entertainment unit is over 9' long and with 4 doors and one drawer. The gaps on this one are tighter, 3/32" and the door edges have to be flush to the sides of the box, and hang under the overhung top so gap there is critical. Again, even .010" would accumulate over the distance and create issues.

364991

[Martin Wasner]

Full overlay and long runs demand almost perfection. The cumulative errors will drive you bonkers otherwise. You have a little fudge room, but not much. I don't do a ton of full overlay stuff, but squaring and sizing five piece doors and drawer fronts is a massive pain. I'm looking forward to having a CNC and just building the fronts oversized and just throwing them on the CNC to get them to finished size and squared up.

[Frederick Skelly]

Hi Fred, well, I understand your point but I find I run into this issue all the time. I tend to make a lot of projects with flat doors and in a style where everything is tight and symmetrical. This picture below is pretty typical of my projects, it has 6 doors and 6 drawers, is over 7 foot tall and 7 foot wide, a linen closet in the master bath. All the doors and drawers have a gap of approximately 1/8", if any of my panels were off, you would see it immediately. As for fine tuning, you really can't do that after the product has been finished and you don't want to be touching up in the field, so better to be right the first time.

Brian, you do beautiful work. Thanks for the examples and for clarifying. I see your point for the things you build and I agree it's better to get it as right as you can the first time.

Best regards,
Fred

[Sam Murdoch]

Better to get it right the first time - or the second time - at the least, just before it all goes together, rather than after the fact. Words to do woodworking by... I am in agreement :).

[Brian Lamb]

Brian, you do beautiful work. Thanks for the examples and for clarifying. I see your point for the things you build and I agree it's better to get it as right as you can the first time.

Best regards,
Fred

Thank you! This is almost all the type of work I do in cabinets, even in kitchens I get long runs with maybe several drawer banks and lots of doors in a 10'-15' run of cabinets. The furniture work becomes easy in comparison, at least for square and perfectly on size, because you are just generally making a stand alone piece... if the dining room table is out of square .030" then who's going to know? :D

[Andrew Hughes]

Its a small world.
Don't know if you mind me posting this Brian.
Heres one of Brian's Indicator holders I bought from him.Nothing like it out there.I use mine to set my jointer knives on my Oliver.

[Brian Lamb]

Hi Andy,

You got the last one I had in stock, hence the shortened butt on it, was actually a setup part as the blank wasn't quite long enough. I don't mind you guys posting, I just don't want to get in trouble for advertising, hence I didn't mention product in my previous posts.

[Andrew Hughes]

Hi Andy,

You got the last one I had in stock, hence the shortened butt on it, was actually a setup part as the blank wasn't quite long enough. I don't mind you guys posting, I just don't want to get in trouble for advertising, hence I didn't mention product in my previous posts.

Aww cool hoping I wasn't out of line.I am interested in a second red one put me on your list please :)

To the Op sorry for high jacking your thread.

[Rod Sheridan]

On a cut 30" long I'll be 3 to 4 thousandths of an inch out of square if I use the crosscut fence on the outrigger.

If I just use the short fence on the sliding table I would be less accurate.......Rod.

[Steve Demuth]

Fascinating how this thread got parallelism all tangled up with square. They are very different problems. Even a casually tuned cabinet saw will produce parallel edges to within a few thousands of an inch - as long as your fence is flat, your reference edge is flat, and your technique maintains a tight conformance of reference edge to fence you'll get a parallel cut. Just to prove the point, I detuned my fence to be 1/64" off parallel with the blade (which already has about .004" runout to start with) and cut a 3' long piece of mdf. It mic'd out at parallel +/- .002", except for the leading 1" or so, which was about .010" narrower than the rest. I obviously rotated the mdf every so slightly at the beginning of the cut. But in order to make my pointI wasn't using a featherboard or fence hold-tight - had I been I'd have gotten within .001 - .002" on the entire length.

Getting the narrow dimension truly square to parallel edges on a large panel is much harder, unless you have a slider or very high quality panel saw. I'm happy with 1/64" on that, and often would settle for 1/32" - because, notwithstanding Brian Lamb's beautiful work, and informed opinion, there aren't a lot of cases where it matters much. Unlike parallelism, where common construction techniques very definitely DO accumulate to unwanted visual effect, there aren't a lot of situations I've encountered - other than rail and style or any other joinery where you've got to "close the circle" - where that level error accumulates or is noticeable (to anyone other than the builder, of course).

[johnny means]

IMO, anything meant to be square should be ,for all practical purposes, square. For example an entry door should not run out any amount that can be measured with a tape measure or ruler. This is easy to achieve using any Pythagorean triple.

[lowell holmes]

Before track saws were invented we would rip a 12"+/- wide strip from a sheet of plywood and nail a joined 1x4 into the center of it. We would rip the plywood on
both sides using circular-saw with a plywood blade. Then straight cuts could be made by clamping the guide to the piece being trimmed and ripping with the piece
with the circular saw. I have a 4" Porter Cable circular saw with a good blade that will make that cut in a heart-beat. I used to build houses and job built my cabinets.
I still have the 4", 6" and 8" "skill"-saws.

It is easier to rip a sheet of plywood using this set up than using a 10" table saw. I find that ripping an eight foot length of plywood o a table saw to be dangerous.
I do it if I need to, but avoid it as much as possible.