What do you consider "dead on"

[Richard Wagner]

when cross cutting items like stiles and rails?

After reading another post on how do you cut stiles and rails (make accurate crosscuts), I must ask "What constitutes "dead on". If cutting large panels, a small error can result in a major miscut but when cross cutting pieces no more that 6"-10" wide a small error is just that - small. If I check with a quality square and see no light, I consider my cross cuts to be "dead-on".

Am I wrong?

[Bill Huber]

Not in my book, that is the way I check it, if I see no light with my good square I say it's dead on. I have a Woodpeckers 1281 for the wider stuff and a Woodpeckers 640 for the smaller items.

[Richard Wagner]

Not in my book, that is the way I check it, if I see no light with my good square I say it's dead on. I have a Woodpeckers 1281 for the wider stuff and a Woodpeckers 640 for the smaller items.

I am not that good a wood worker. My squares, except for one engineers square from Woodcraft, are all old Stanleys.

[Jeff Duncan]

'Dead on' specifically for door rails is something I usually don't worry about until I'm coping. I always set my copes to do a full cut + about 1/32" which ensures a full cut. Then if I use my square I don't want to see any light.

It also doesn't matter, (to me), for door stiles to be 'dead-on' as I trim the doors to size after glue-up.

Using a square with a light source behind is an age old technique that you just can't go wrong using IMO.

good luck,
JeffD

[Todd Burch]

I use a "dead on" crosscut sled I made. I can crosscut up to 26" wide. To verify it was "dead on" when I made it, I used a freshly ripped (known parallel edges) ~24" x ~36" scrap piece of ply (what I had laying around, maybe it was MDF) and crosscut the right edge (majority of the piece was on left side of the blade), and then I moved the scrap to the right and put the freshly sawn edge against against the tablesaw fence. There should be no gaps. If there is a gap, either the sled is off, or the fence is not parallel ( // ) to the blade. Check and fix the fence-to-blade //ism first. Then, fix the sled if/as needed.

Once that is worked out, make a fresh cut as above, and then flip the scrap ply over, front to back. The side of the piece of ply that was against the crosscut sled fence is now furthest away from you.

Jam the fresh cut edge up against the fence. If your sled is dead on, it should be flush with the fence along the whole length. Doing this doubles any inaccuracy so it's easier to see.

And, as the others above have said, for stiles, it's not so critical. But rails you do want to dead on. If you make and adjust a sled per the above, when you do cut your (comparatively tiny) rails and stiles, they'll be dead on too.

Todd

[Richard Wagner]

Thanks for the comments, all of you. I guess I gotta double check the sled and do whatever might be necessary to make it "dead-on".

[glenn bradley]

I do as you do Richard. If a known good square shows a good cut, its good. For profiled rails and stiles, if there are no visible gaps where the profiles join, I'm good.

[James White]

Richard,

Here is a thread and a video that may help you get you sled dead on.

http://www.sawmillcreek.org/showthread.php?151810-5-Cut-Method-What-is-it-amp-How-to-do-it

http://www.youtube.com/watch?v=UbG-n--LFgQ

[Neil Brooks]

I like to think of myself as kind of the Anti- Norm Abram.

He'd nail the joinery (on the final televised footage, at least ;) ) and declare: "that fits pretty good."

Whereas ... my joints fit pretty good, which I declare to be "dead on."

[HANK METZ]

"dead on" is when you examine the cut and do not declare "I'm dead"

- Beachside Hank

[Sam Murdoch]

I just think that in most woodworking, accurate and dependable joints make the work better and more enjoyable. I don't care about being 1/32", or sometime even 1/16" off, as long as I am consistently 1/32" or 1/16" off . Dimensions are organic but tight joints depend on accurate squareness in the length or width or thickness. If I'm framing with 2x stock accurate is relative to the materials at hand. Furniture and cabinetry however, demand exacting tolerances. Out of square in any orientation chases you throughout the project. Even boatbuilding requires "square to each other" type joints regardless of the curve or other "not square" aspects. "Dead on" is the good work that doesn't require you to "fix it" sometime later in the project. :rolleyes:

[Bryan Cramer]

Sam is exacally right. I belive accuracy is two parts: the prescision of your tools and consistency. Cutting smaller parts square is eaier than a large panel. Therfore I alway register the fence of my crosscut sled to the same edge. Then if I am not cutting "perfectly square" I will be cutting conssistently out of square. I strive for conssistency because this technique cost nothing compared to buying new accurate squares. I also use the same square throughout my project because not all of my squares agree with one another (consistency).

[Phil Thien]

If I check with a quality square and see no light, I consider my cross cuts to be "dead-on".

Am I wrong?

No, not in my opinion.

You can get bogged down in the 457-cut method of squaring a sled. If a trusted (tested) square says the cut is square, then you're done.

[Dave Norris]

I'm probably asking for trouble with this but... doesn't "dead on" mean "dead on"? Off by 1/32nd is "off a little"

The proof is in the pudding, so....

I go by the cuts themselves. Set up a stop, make a cut, flip the piece, make a cut, until you have four equal pieces. Then set the pieces together to see if they will make a perfect square. If you set three pieces together, and the fourth slides together with them, you are dead on. ANY gap means you are off. By flipping the pieces, you will also see if your blade is at 90 to the table.

I do all my crosscutting on a Dewalt RAS, and i actually set it using this method but with saw set at 45. Slide three pieces together, and then slide last one in place. Open at back of miter and I adjust to the left, open in front of miter and arm goes to the right (or vice versa-it's late) this actually multiplies the error, so if you are off even a skosh, the last miter will be off. When all four miters are perfect, it's dead on.

I've never gotten four perfect joints using a square to measure my cuts. I have (and do) get four perfect cuts using this method.

[Terry Haslett]

+/- 2 nanometers and +/- 2 arc seconds. :):):)

[tom blankenship]

I consistently keep all cut deviations within 2 angstroms per parsec. :D

tom

[Richard Wagner]

I consistently keep all cut deviations within 2 angstroms per parsec. :D

tom
I was serious when I asked the question. I am sorry that you and terry could not have been equally serious with your responses. To all others who have responded, I appreciate the inputs. I do have some work to do in this area.

[tom blankenship]

Hello Richard,

I apologize to you for my meager attempt at humor. I felt that you had already received several good responses to your question. Your question is one that I struggle with as well.

Again, sorry to be annoying.

tom

[John Coloccia]

I was serious when I asked the question. I am sorry that you and terry could not have been equally serious with your responses. To all others who have responded, I appreciate the inputs. I do have some work to do in this area.

I find it's easier to do precision work when I take a deep breath and RELAX.....

...
...


:D

[ian maybury]

Must say I think the core issue is sled or whatever and saw set up. It's necessary to check the cut afterwards with a good square, but what do you do if it's off? The real solution is surely a set-up that cuts reliably square???

ian

[Will Blick]

Richard, this is one subject where humor is always injected, some can't help themselves....

In short, yes, if your square says its square, well, that is pretty good... ASSUMING your square is SQUARE! Otherwise, its only as square as your reference, this is why IMO, it pays to have excellent reference tools, they make life so much easier in the shop.

Cutting oversized pieces with many of the multiple cut techniques is good method to assure your cuts will remain square, as if they pass these tests, they will surely cut thin rails with more than enough squareness.

Dead-on IMO is relative... sometimes, you can cut to a range, precision is not that accurate, other times, it means everything, such as dovetail or box joints... gaps as small as .001" become apparent, and you think, gosh, I screwed up. Other times, 1/32nd, and your happy...

[Carl Beckett]

Richard, this is one subject where humor is always injected, some can't help themselves....

Dead-on IMO is relative... sometimes, you can cut to a range, precision is not that accurate, other times, it means everything, such as dovetail or box joints... gaps as small as .001" become apparent, and you think, gosh, I screwed up. Other times, 1/32nd, and your happy...

Im all for humor, I dont think there was any disrespect intended.

Lots of good answers - I will chime in with the 'its relative' position. First and foremost, its a subjective term/definition. For example, as a machinist I might have a different definition than I would as a woodworker. But what if I am both?? Then I have to qualify it as 'dead on in woodworking'. Vs 'dead on in machining'. But here, sure woodworking is implied but if the responder crosses different disciplines then they have to qualify the answer 'for woodworking'.

So you get some responses to the question as - 'what is a precision that is adequate to where the precision of the joint no longer is a factor' (indirectly being what 'dead on' pretty much implies. But here again, it depends on the purpose of the joint. What might be ok for one type of joint wouldnt be for another. And Im of the camp that I will take higher accuracy in repeatability than absolute accuracy of any one given joint/cut. Im a user of story sticks though..... (dead on for a tape measure would we very different limits than dead on for a small precision square. In the instrument world they talk accuracy in terms of "% full scale". This allows the precision to fit the particular range of the measurement, assuming that in practical terms a zero to two ton scale might need a different resolution/accuracy than a precision kitchen scale.

So maybe we could talk percentages (for squares, angles work well because there is a limit of 360 degrees so in effect already a % full scale).

Personally, I have been known to intentionally work with some error/slop in order to gain some ease of assembly or speed, if its something that I know will close up and function well later in the process (glueing and finishing steps can swell things)

So I will also add a 'whats not good enough?' question. Which is the key question for me. I dont usually ask whats good enough until something isnt going together they way I want, and then I start checking things. If I discover a problem - then that is 'not good enough' - but until that point its all good enough. Meaning whats good enough, is everything except the items that are 'not good enough'.

[Bobby O'Neal]

I agree with the square in the light method.