Working Towards 1/64's Accuracy

[Jeff Wright]

At age (nearly) 60, I look back over my lifetime and see myself as having been (and now being) two woodworkers. Thirty years ago, I accepted cutting tolerances of a 1/16 or even an 1/8 of an inch, with almost no true understanding of wood movement. Among my projects then were building two kitchens and a few standalone pieces of furniture. Today, with having had a thirty year absence from woodworking, and with an entirely new cast of tools and machinery, I have set as a goal to work towards an accuracy of 1/64 of an inch. While wood's movement itself is often greater than that tolerance, I figured I would at least shoot for that kind of accuracy even if I end up with something less fine.

Two inexpensive tools that are helping me achieve that goal are calipers (at $26.00, it gives me metric, imperial decimal, and most useful . . . fractional readings) and Lee Valley's Wood Movement Reference Guide, a wheel that helps one determine a specie's wood movement under changing temperature and humidity environments (available for $6.50 at LV).

The caliper has found a place in my apron pockets along side of the 4" starrett square. The Wood Movement Guide Wheel reminds me of the need to pay particular attention when picking through the wood pile at my local hardwoods dealer to select grain patterns most likely to expand and contract in a manner suitable to the intended design. The wood design guide wheel includes a 9-page instructional pamphlet that told me more about the nature of wood movement than any other resource.

Hey, I may never achieve my goal of 64ths, but you gotta have a target in mind, right?

[Gary Keedwell]

Jeff,

As a machinist in my other life, I strive for very tight tolerances myself. I have had other woodworkers mock me...saying I don't need to be so accurate because the wood is going to move anyways.
Well, I have joints that are still tight after many years. I do make allowances for wood movement....table tops is one example. However, I have picture frames that are as tight as the day I made them.
Gary K.

[Jeff Wright]

Jeff,

As a machinist in my other life, I strive for very tight tolerances myself. I have had other woodworkers mock me...saying I don't need to be so accurate because the wood is going to move anyways.
Well, I have joints that are still tight after many years. I do make allowances for wood movement....table tops is one example. However, I have picture frames that are as tight as the day I made them.
Gary K.

Gary, it's great to know I'm not alone! :D

[Luciano Burtini]

I just have to comment... although I may not always acheive such tight tolerances, there are two things that I do ALL the time when building furniture and cabinets. First all thicknessing is done with a vernier caliper and all measurements are done with a rule.

Using a tape for these two jobs has caused me no end of grief. The chosen tools need not be the best quality ones available, but stick with a decent quality ones. Tapes are notorious for having repeating errors (they are usually stamped using a large wheel so any error in the stamping is repeated at the circumference of the wheel). Likewise, I stay away from silk screened rules which often are inaccurate and almost always have very thick markings. Use etched rules and decent calipers (Starrett, Mitutoyo, B&S, M&W, Rabone, Union as well as many other brands work very nicely) and your first source of errors virtually disappears. Of course, this does little to address operator error.

I too have suffered the wrath of fellow woodworkers who have seen my 8" starrett calipers sitting on my thickness planer ;)

[Pete Bradley]

I'm a machining and accuracy fanatic too, but think a lot of "accuracy" in woodworking is about either avoiding measurement or knowing what the eye sees as accurate.

For avoiding measurement, I mean matching or marking one part off another. It's often a lot harder to get two independent pieces to the same 64th or smaller, but you can match one piece to another more accurately than that.

Knowing what the eye notices is also important. A gentle curve can look straight, and a handplaned surface may look flat even though it's not perfect. Leave a 64th gap between two pieces it looks like junk, but match them exactly at a slight angle and no one will notice. Handling minor blemishes is another example. I'm working with an amazing piece of figured maple that has a few small pits in it. I can't count the number of times well-meaning people have said "just mix some sawdust with glue and it will be a perfect match". Not after it's finished it won't. However, I'm finishing with shellac rubbed out to a low gloss, so I can carve a gentle depression in the surface to the depth of the divot and sand it out, and the slight wave disappears completely in the figure of the wood. A high gloss might make it a little more obvious, but still probably less so than filler.

Pete

[Mike Cutler]

Another anal "Tight Tolerance Person" here. :cool:, ;)

[Jeff Wright]

I just have to comment... although I may not always acheive such tight tolerances, there are two things that I do ALL the time when building furniture and cabinets. First all thicknessing is done with a vernier caliper and all measurements are done with a rule.

Using a tape for these two jobs has caused me no end of grief. The chosen tools need not be the best quality ones available, but stick with a decent quality ones . . . .

Where possible I try to eliminate measuring altogether. As an example, I just got a couple sets of these bar gauges:

http://www.leevalley.com/wood/page.aspx?c=2&cat=1,43513,43553&p=32585

The use of story sticks also comes to mind.

I was sensitive to measurement tool errors when I calibrated the fence ruler on my MiniMax 410 combo machine. I compared the rules on both the fence and my various measuring tools to see how closely they aligned. I determined it helpful to try to use the same measure tool throughout a project rather than switch midstream.

[Steve Rowe]

At age (nearly) 60, I look back over my lifetime and see myself as having been (and now being) two woodworkers. Thirty years ago, I accepted cutting tolerances of a 1/16 or even an 1/8 of an inch, with almost no true understanding of wood movement.
...

Hey, I may never achieve my goal of 64ths, but you gotta have a target in mind, right?

I have discovered as my age increases and my eyesight decreases, that 1/16" looks more and more like 1/64".:D
Steve

[Per Swenson]

First off a confession.

Being Norwegian I am a true square head and

a obsessive collector of measuring devices.

I don't use 'em.

My eyesight is horrible from staring at the sun to much in the 70s.

Nope, story sticks and gauge blocks.

Gauge block and the touch of my fingers in my case is infinitely better

in the accuracy department then anything I can see or hope to remember.

Per

[Jim Becker]

Like Per, I've been embracing the use of story sticks a lot on my projects at this point...and I try to avoid the tape measure as much as possible.

[Jeff Wright]

. . . My eyesight is horrible from staring at the sun to much in the 70s. Nope, story sticks and gauge blocks.Gauge block and the touch of my fingers in my case is infinitely betterin the accuracy department then anything I can see or hope to remember.Per

I agree. At age 59 I can understand why they force retire airline pilots at age 60. The eyes just aren't what they used to be!

[Don Bullock]

I find it very ironic reading this post. I just sent in an order for some tools for measuring to WoodCraft because today they had free shipping.

Jeff, I too am in the same boat after more than 20 years away and I'll be 60 next Saturday. I find that my tolerance for error isn't what it used to be either. I just wish I had kept up with the skills necessary to create things to the tolerances I want. Yes, the bar gauges are on my list of things I need too. No, you are not alone.

Per, I have a small set of gauge blocks and never thought of using them. Just mentioning them gave me an idea for several uses for them. I could have used them on my last project. I used my fingertips instead, which did work. I checked the accuracy with a ruler. As for story sticks, I forgot all about them (all the years away from the hobby combined with CRS syndrome). After finding an article on them on the Fine Woodworking sight, it all came back. Thanks.

Steve, I fully agree with your observation. Fortunately for me it's my distance that is going. I can still see close up without glasses.

Luciano, part of my order was some more rules. I don't know why I have always used a tape measure when I knew a rule was more accurate. I guess some things, like wisedom, may actually improve with age.

[Al Killian]

I have a question about measureing long things over 3 feet. If you dont use a tape, then how will you measure it then? I am fairly new and have yet to tackle anything of great style. Still trying to get all the basics down.

[Per Swenson]

Al,

Use a dedicated tape measure. Once.

Then cut a stick. Mark it with size and throw a dap of paint or tape on

it, so it is differn't than the rest of the sticks.

Use this to mark all subsequent pieces of the same size

cutting them and then checking with the tips of your fingers

with the original stick for exactness.

It is way faster and accurate then measuring each and every piece.

This also eliminates error.

The tips and palms of your hands are the most sensitve measuring devices

you own. Don't believe me?

Check your face a hour after you shave.

I forgot something, silly me.

It is more important when building something that the parts are identical,

for plumb, level and square then the actual size. I hope this makes sense.

Per

Per

[Mike Cutler]

I have a question about measureing long things over 3 feet. If you dont use a tape, then how will you measure it then? I am fairly new and have yet to tackle anything of great style. Still trying to get all the basics down.

I have a Starrett 6' metal rule to do this.

Per's method is quick and time tested.

Method I was taught in cabinet making, which I don't use, is to use sets of dividers and work from a common reference line.

A tape measure can be used, but...... once used it becomes your defacto standard,and has to be used in the same manner. It might have that little "push me, pull me" doo-dad on the end, but don't rely on it for anything other than basic measurements and general carpentry.

The best way to condition yourself is to begin to use a marking knife, and get rid of the pencil. Not always advice I take myself though.:eek:

[Jeff Wright]

I have a question about measureing long things over 3 feet. If you dont use a tape, then how will you measure it then? I am fairly new and have yet to tackle anything of great style. Still trying to get all the basics down.

Being a self-confessed sucker for woodworking gadgets (and boating and flying gadgets as well), I use the gauges shown here. They can be adapted for almost any reasonable length:

http://www.leevalley.com/wood/page.aspx?c=2&cat=1,43513,43553&p=32585

[Mark Singer]

I use a tape to rough out. Then as I crosscut to length , I use the first piece cut as the "story stick" and measure all from that. Drawing and drafting most of my life is an advantage in layout. Some tings I measure with a digital caliper such as distance from blade to fence and heigh of blade at the cabinet saw for ripping rabbets...etc. Some things I just eyeball....you need to know when. My main marking/measuring tools are combination square, double square, marking gauge, bar gauge for inside and outside measurements, marking knife, hard and soft pencils, I also use colored pencils for general marking, like matching boards, "planed face,",jointed edge" etc..... The best measurements come from the piece that is already cut. Cutting techniques are critical or marking accuracy doesn't matter. Cross cut sled....transfer mark with a square to the leading edge and set the blade right next to the line.
Joinery...scribe from one part to another as with dovetails then saw accurately. The properly tuned bandsaw and fence is a great place to make joints and templates.....you can just move the fence a touch and trial and error if need be. You can set stops to limit each cut consistently.
It is far easier to make a tight joint the first time then it is to try to fix it and it always shows.

[josh bjork]

I'm a hack but my stuff got a bit better when I started checking my framer's square for square every day. Also checking things for square after they come off if the saw. My equipment isn't elaborate by any means and there is a lot of user deviation.

[Pete Bradley]

I have a question about measureing long things over 3 feet. If you dont use a tape, then how will you measure it then?

A lot of times, the exact length is not as important as having it match other parts. For example, if you're cutting aprons for a table, hold the two pieces together and cut them to length. If you marked and cut the two pieces separately, it would be very difficult to achieve the same level of accuracy.

Pete

[pat warner]

A connection like this (http://patwarner.com/images/fish.jpg) will not tolerate a .003" error.
The parts just jam on assembly. The T&G runs along the entire glue line
A 1/64th is a mile. Keep it tight.

[lou sansone]

it depends on what you are building. story sticks work for a lot of things, but in the case of sash, you better be within 1/64 or you will see it when you assemble it. Although I have not built any yet, the gallery in a decent secretary would also need some pretty close measurements.

Lou

[Al Killian]

Thanks for explaining it to me. I am still working on getting a decent set of measureing tools to make life easer.

[Steven Wilson]

Accuracy and precision start with machine setup. It's hard to make a square box when you can't cut square! I was using story sticks, a knife for marking dovetails, Starret rules, a single tape measure (there is only one tape measure in my shop), well tuned planes, very sharp paring chisels for making things fit, stops for repeatability, and all the other usual things and was fairly happy with the results although sometimes a glueup would take a bit more force than it should. Well the next step up in accuracy occured when I replaced my American style tablesaw and got my sliding table saw dialed in. All of a sudden glueups went smoother, boxes were square without clamping the diagonal, and things just went smoother. What was the change? I'm now cutting square stock. It's very hard with the limited crosscut capability of a standard tablesaw (or a cutoff saw you make) to make square cuts with great precision. It may measure square but still isn't quite square. This inaccuracy doesn't show up in smaller pieces but is very noticeable with larger assemblies. The large cutoff carridge of a sliding table saw makes it fairly easy to dial it in so that square cuts are really, really square. Life is good when it's square.

[Matt Lentzner]

So if one is shooting for 1/64ths by length, what is the equivalent accuracy of an angle? 1 minute, or less?

Just curious,

Matt

[frank shic]

jeff, if you want 1/64 accuracy, you may want to consider investing in a digi-fence - although it's not cheap! i bought one last year and it's made my woodworking much more accurate and repeatable.

[Steven Wilson]

what is the equivalent accuracy of an angle? 1 minute, or less?

It depends over what distance. I was always able to get things square on my PM66 to well under a minute as measured with a Starrett protractor, however that isn't good enough for an 8' bookshelf. For an example, I use pieces of MDF roughly 2'x2' to align the crosscut carrige on my slider. Performing a 5 sided cut (cut, rotate, repeat) I can get the maximum difference in parallel between the 1st and 5th cut to be .003" or less. That mean's over roughly 8' I'm off .003" from being square. That's a whole lot less than 1 minute of arc and probably less than a second (I should calculate that sometime).

[Jeff Wright]

Accuracy and precision start with machine setup. It's hard to make a square box when you can't cut square! .... Life is good when it's square.

Steve . . . AMEN!!

[Jeff Wright]

jeff, if you want 1/64 accuracy, you may want to consider investing in a digi-fence - although it's not cheap! i bought one last year and it's made my woodworking much more accurate and repeatable.

Frank . . . a friend of mine recently installed the digi-fence system on his Euro combo machine. That will be "second semester" work for me!!

[Michael Cody]

I've always treated accuracy as a goal but precision is a must. If you cut 4 pieces of wood all the same length and miter all the corners so the angles total 90dg at the corner and 360dg for the entire square -- it will be square.

If the table legs are 27.75 inches or 27.57 inches, it doesn't normally matter as long as "ALL" of them are the same. I strive for ok accuracy --- but always for repetable precision. For me it's story sticks and blocks. Measure the first one close, make all the rest the same.

[Jeff Wright]

. . . For me it's story sticks and blocks.

Mike, I am familiar with story sticks, but can you clarify what BLOCKS are? Thanks.

EDIT: Ahhh, you are referring to SET-UP blocks!

[Jules Dominguez]

Most dimensions in things we make of wood, whether we design them ourselves or use a plan, are arbitrary and it doesn't make a bit of difference whether they're off a 64th or a 16th or even more. What matters is that all of the parts that need to be the same dimension are the same dimension, and that everything that needs to come together in a good fit does, and most of that is easily achieved by using the same setup to cut multiple parts.
When you design something, you decide you want 8 inch or 10 inch or 11 inch shelves, and it doesn't make a damn bit of difference if they're off one way or another as long as everything fits together well and the visual effect is what you intended. Hell, if you had had another glass of wine while drawing up your plans, you might have decided to make that 36 inch tabletop 37".

[frank shic]

jeff, it's not as simple as chucking a bit in the router, but the instructions are first-rate and the results are dead precise. just keep it in mind the next time you have an extra $200 lying around... i'm sure you'll advance to the second semester in no time at all!

[Wilbur Pan]

It depends over what distance. I was always able to get things square on my PM66 to well under a minute as measured with a Starrett protractor, however that isn't good enough for an 8' bookshelf. For an example, I use pieces of MDF roughly 2'x2' to align the crosscut carrige on my slider. Performing a 5 sided cut (cut, rotate, repeat) I can get the maximum difference in parallel between the 1st and 5th cut to be .003" or less. That mean's over roughly 8' I'm off .003" from being square. That's a whole lot less than 1 minute of arc and probably less than a second (I should calculate that sometime).
If I remember my trigonometry, 0.003" off over 8', or 96" will give you an arc of tan-1(0.003/96) = 0.00179 degrees = 0.107 minutes = 6.45 seconds.

Hope I did that right. :)

[John Bailey]

They say a machinist works to 1/1000th of an inch,
a furniture maker works to 1/64th of an inch,
a house builder to 1/8th of an inch,
and a boatbuilder - to the nearest boat.

My present projects are a 15ft. sailboat and an 18ft. kayak - they work out better with my old eyes.

John