Tolerences...

[cwo4cno7325]

Hi, All,

I'm wondering what tolerence one uses when cutting pieces in general. I am currently making an end table from pine as a prototype for the 2 real ones for the LR. I can usually cut boards to 1/32 of an inch of what I had on my plans, but when assembled, I see small gaps or areas that I have to plane down.

I rather enjoy the "extra" woodworking, but am wondering if this due to being a poor woodworker, or if this is normal procedure. Same thing with right angles. I set the table saw up with a machinist's square, and a "tight eyeball", and get pretty good results, but can still get a 1/16 inch error on a 24 inch square board.

What do you guys get on the average?????

Thanks.....

[SonomaProducts.com]

I think with stable and well tuned equipment one can typically hold better than a 32nd over 24 inches. Larger spans become a challenge.

Regadring gaps and "plane to fit" type situations, the best approach is to "design in" methods of fit so you have a plan about how all critical fits will work and can be gauged as milling and fit-up progesses. Yes. miters often need to be trimmed to fit well but hiding edge gaps under moldings, utilizing rabbet ledges to let butts hide under an overhang and keeping a shoulder plane on the bench and having some filler on hand and learn about burnishing to push the edges of wood around when needed all server to make it all look good in the end.

For me a card> Hi, All,

[Limp Arbor]

Keep at it and you'll continue to get better.

I try as much as possible to 'production cut' everything. Doing things like cutting all of the similar rabbits or tenons at the same time gives you consistency when fitting the mortises or vicey-versey. Anytime you need to reset the fence, blade height, angle, etc. you are allowing for error.

As far as being out of square I use a panel cutting sled to square up panels. Once you take the time and make all your test cuts when building the sled you'll never have problems again. I use a simple one like this:

the runner on the edge of the saw) But some go with a double runner sled:

[dpb]

[e-mail] wrote: ...

...

Something's not set up to have that much error in 24" -- have you verified the miter slot is actually parallel to the blade? If it isn't, even though the miter gauge is perpendicular to much higher tolerance, the panel will "move" towards or away from the plane of the blade as it passes by.

--

[Leon]

Set your saw up properly, .005" is typically acceptable. Cut to exact measurements, if you are not worried about 1/32" you will continue to see gaps. Then there is practice that makes you better.

[Edwin Pawlowski]

I'm pretty tolerant of other's religious views, sexual preferences, but I've gotten into some discussions on politics.

[Joe]

paraphrased from Tolpins book "Workshops"

carpenters work to the nearest 1/8", furnituremakers to the nearest 1/64" and boat builders to the nearest boat.

That advice given, strive to continually get better and you will.

Personally, I'm constantly striving for my tolerences to get better or my eyesight to get worse. jc

[spaco]

I used to make all my own bee hives. There are 9 or 10 frames (of 4 pieces each) per super, so there are a lot of pieces to make over and over again. I was able to hold +/-0.003 in dry pine to make all the parts. Table saw and router where the main cutting tools.

Pete Stanaitis

-------------------

[e-mail] wrote:

[Phil Again]

This reply is a bit OT:

The goal is not the tolerances you want, but in finding the source of the errors. Eliminate, reduce, or control the sources of errors and the tolerances will take care of themselves.

I know that sounds like some Monthly Quality meeting in the lunch room BS, but it is true. Why is your saw giving you 1/16 inch error on a 24 inch board? Why are there gaps in your assembly? The longer you are in woodworking, the more you will find out about your tools and the sources of errors.

By learning to think through the process you are using to cut shape the wood you will find the sources and by trial and error learn how to control the sources of error. You will then learn a lot more about your hobby and yourself.

BTW: 100 years ago, in woodworking the greatest source of error was lack of practice in using hand tools. Nobody today has the time or inclination to serve many years as an apprenticeship learning to use hand tools. We trade money for power tools, jigs, fixtures, and carbide blades instead of spending time doing very boring and repetitive tasks. But even back then, thinking through a problem to reduce the causes of errors was an important part of learning woodworking, IMHO.

Now back to other replies that may actually help you.

[Phisherman]

Working with wood is very different than metal or plastic. Wood moves so we often use joinery that overlaps, slides past each other, etc. To check to see if two boards are the same length, I'd butt them up to a stop. If I can not feel a step on the other, I say, for all practical purposes that they are the same length. A 1/16" error is telling me the table saw is in need of a tuneup.

[Phisherman]

I'm very tolerant, of political parties, race, religion, gay marriage, abortion, etc. But I am very intolerent about the bailouts and lack of privacy in the USA.

[Tanus]

I don't think you can do better than this book for table saws. It will guide you through some of the setups the guys are talking about and help you tune your saw. Of all the woodworking books I own, this is in the top 5

Mehler describes the topic very well, and is a wealth of knowledge.

Tanus

[Pat]

I must be a carpenter.

[clare]

I always set my miters so that if there is an "open" edge it's on the blind side - outside mitered corners go for 89 degrees instead of 90, insides go for 91. That way the visible edge is ALWAYS tight.

[clare]

Well, like death and taxes, the latter is inevitable if you live long enough. SWMBO never seams to have that problem when looking at your work though!!!!

[clare]

Google "dave woodland" or "sawshop" for common sense saw setup information that will make your saw predictable and accurate.

Try

[dadiOH]

I feel I have failed in life if I can't keep it within 1/64 over eight feet.

I fail a lot so I just beat it into submission with a hammer.

[cwo4cno7325]

Thanks for a ton of good ideas! I think I have all tools well tuned up, and I may have overstated the 1/16 inch error. That 1/16 inch usually comes from the combination of a number of parts, each with a very small error that when added together may yield a total 1/16 inch error at most.

One thing that I have found that helps is to cut all the same size pieces at the same time, as mentioned above. They may not be exact in dimension, but they are the same!

The tolerences mentioned for house carpenters, and furniture makers at

1/64 is a good target. I think if I can get 1/64 from the power tools, then I can "fix" things with hand tools during assembly, to MY tolerence.

And practice! I just looked at one of my first projects, a wood box for camping, done on a Workmate, with a really bad circular saw. It looks "very authentic" and at least shows some improvement when compared to current projects.

Anyway, thanks for all the help.

Rich.....

[Tim Douglass]

Cumulative error is a bit stickier problem. You would think that the errors would tend to go at random directions, effectively cancelling each other out over the entire project, but I tend to find that they always are strictly additive, leading to sometimes rather significant error at the end.

No answers for you, but you are not alone.

-- "We need to make a sacrifice to the gods, find me a young virgin... oh, and bring something to kill"

Tim Douglass

[J. Clarke]

Statistical analysis of tolerance stacks is in itself an interesting exercise.