PDA

View Full Version : Straight Edge (or not)



Ed Calkins
04-16-2009, 12:33 PM
Well I just used my new SawStop PCS on a project and could not be happier with the saw -- of course my previous table saw was a vintage (1949 or so?) Craftsman 8". During the set up of this tool and recent previous tasks I needed a straight edge of significant length (at least more than 12"). I used a 48" level I have but I am not sure how straight it really is. I settled for leveling the table extensions etc. on the saw to less than .005" as I think that is close enough and probably my straight edge is not even at that tolerence. I am concerned about this somewhat as I purchased a Starette 12" combo square a few years ago and then found that the 12" combo square that was my "standard" for the previous 10 years was not accurate when compared to the Starette. I am thinking a shop standard straight edge of 24" to 36" would be a good resource to have available for checking things that really need to be flat or straight. Any opinions on the need? And if needed, any opinions on what straight edge? Thanks, Ed

Brian Kent
04-16-2009, 12:38 PM
I have and like the LeeValley aluminum.

Steve LeGrue
04-16-2009, 12:40 PM
Lee Valley/Veritas makes several very affordable straight edges in both alluminum and steel. I have checked them against my $200 Starrett and they are all dead perfect.

Steve LeGrue

Clifford Mescher
04-16-2009, 12:41 PM
A few years ago I tried to tune up my jointer with a 4' level and got very frustrated. I purchased a Starrett straightedge and haven't looked back. It is great to have a "standard" to fall back on. Clifford

Chris Kennedy
04-16-2009, 1:21 PM
+1 on Lee Valley.

Mark Roderick
04-16-2009, 2:04 PM
It's very convenient to have a straight edge in the shop and those you can buy at Lee Valley are very good.

You can also make a straight edge of any length from plywood and hardboard, materials that won't change very much with changes in temperature and humidity. Cut the plywood about three inches wide and the hardboard about two inches wide. Put a groove along the edge of the plywood and insert the hardboard in the groove.

The trick is to make TWO, then lay them next to each other. If the edges of the hardboard don't mate perfectly, figure out where the high spots are and use a hand plane to take them down. When the two pieces of hardboard mate perfectly even when one of them is flipped end to end, you know you have two perfectly straight edges.

If you haven't used them for a while and wonder whether they're still straight, just put them edge to edge again and you'll know right away.

Paul Fitzgerald
04-16-2009, 3:03 PM
I love my Lee Valley 50" Aluminum Straightedge and would recommend it to anyone.

Paul

Wes Grass
04-16-2009, 4:39 PM
"When the two pieces of hardboard mate perfectly even when one of them is flipped end to end, you know you have two perfectly straight edges."

Actually, you need three. Because with 2 you'll never know if you made a large radius curve that matches up both ways.

Cary Falk
04-16-2009, 4:46 PM
I have a 3' aluminum one from LV and like it. Their offering are pretty reasonable.

george wilson
04-16-2009, 4:48 PM
I have a 4' Tubular Micrometer Co. precision straight edge 1/4" thick,with 1 edge beveled.For serious work,I've got 3 cast iron flaked and scraped straight edges 2" wide on the edges.But,for REALLY serious work,I have a 30" black granite master straight edge.Also have 3 black granite master squares for checking machinist's squares. It is good to have your own " Bureau of Standards."

Paul Johnstone
04-16-2009, 4:50 PM
I used a 48" level I have but I am not sure how straight it really is. I settled for leveling the table extensions etc. on the saw to less than .005" as I think that is close enough and probably my straight edge is not even at that tolerence. I

To answer your question.. I think for leveling a tablesaw cast iron extension, that is certainly good enough. I know mine are not that close, and there's no problem for me.

glenn bradley
04-16-2009, 5:25 PM
+1 on LV. I have the 36" steel which seems to be long enough for most things without being a burden for smaller efforts.

Phil Thien
04-16-2009, 6:36 PM
Most quality levels make pretty darn good straightedges.

If you see mill marks on the edges of your level, that is a good sign.

But I think they are typically straight enough for table saw setup.

Doug Shepard
04-16-2009, 8:13 PM
+1 on LV. I have the 36" steel which seems to be long enough for most things without being a burden for smaller efforts.
+2
I've got the same one. It's heavy though. I've also got an 18" Starrett for smaller stuff.

Brian Backner
04-16-2009, 8:48 PM
Wes is correct that to make a truly straight edge, it takes a total of three edges being worked against each other. Machinists, specifically those engaged in "scraping" the precision surfaces of various machines to close tolerances, refer to this process as the "Automatic Generation of Gages."

The process is very old, and incredibly laborious (think scraping steel or granite away by hand using only a carbide tipped scraping tool), but the techniques can create straightedges, flat surfaces or squares/angle bars accurate to 0.000001" (yes, that's one millionth of an inch). One of the things I had to do in school about a hundred years ago was to create a 3" cube of steel accurate and parallel to 0.0001" in all dimensions - all by hand. Took close to a hundred hours to accomplish - I still have it.

The actual process used is called the Principle of Symmetrical Distribution of Errors, whereby you compare A to B, adjust B to match A, then compare B to C, then adjust C. Then compare C to A, adjust A, then start over again. I can no longer remember the math, but by the time you do ten or twelve iterations, the error, i.e., any variation between the three workpieces drops to less than 0.0001". Typically one of the three, the "Master," will be stored in a protective case and will only be used to test the other two, which are then used for actual measurements in the shop.

Getting even more accurate is merely a matter of more time and labor. As an example, consider that VERY large surface plates can be generated in this manner. First start with a relatively small set of three plates - 12 x 18 x 3" thick cast iron or granite. Using the above principle, you can scrape them to whatever accuracy you like or need. Once you have the set of three to the desired accuracy, take the two working ones and use them to scrape a much larger surface. Start at one corner, flatten it, then start to move out into the remaining field of the larger plate half the width/length of the the smaller ones. Back in the 1800's cast iron plates that were in the hundreds of square feet were generated in this way.

This is obviously way beyond anything a woodworker needs to be concerned with, but the principle can used to make a set of 8 or 10' long straightedges to tune up that Oliver aircraft carrier you just bought. It also makes for fascinating reading. THE bible of these techniques was written in 1955 by Edward F. Connelly - Machine Tool Reconditioning. The book is rare and difficult to find, but can usually be gotten through interlibrary loan via your local branch.

Brian

Greg Crawford
04-16-2009, 8:54 PM
Another fan of my 50", aluminum Lee Valley straight edge. Best bargain in precision straight edges that I could find.

Ed Calkins
04-16-2009, 10:20 PM
Thanks to all for the info -- I will be purchasing a good straight edge asap, not sure how I lived to be 67 without it ;). George and Brian, thks for the insight on serious standards and processes. Not sure how many appreciated your input but I sure did. I spend the last few years of my Air Force time as a "Precision Measurement Equipment Laboratory" tech. It started with a 7 month school in Denver and then I went all over the world calibrating measuring equipment used by the Air Force (and other services) -- lots of it was electronic oriented so we had standards to calibrate instruments that measured voltage, frequency, etc. It opened my eyes to what is meant by a "volt" or an "inch" or "60 cycles" and how there is always many levels of "standards" to take the measurement to whatever level is called for. One of the most interesting jobs I have ever had. Thanks again to all. Ed

Ron Bontz
04-16-2009, 10:42 PM
+3 for the 36" steel straight edge. I considered getting an aluminum one but glad I didn't. Things tend to get bumped in my little shop.:)

Chris Padilla
04-17-2009, 12:48 PM
I have the LV 50" AL straight-edge. It is wonderful for tuning up the jointer since you can just graze the jointer blades on the AL. You don't want to do that with a steel straight-edge, which I also have: a 36" Starrett.

bridger berdel
04-19-2009, 2:12 AM
The trick is to make TWO, then lay them next to each other. If the edges of the hardboard don't mate perfectly, figure out where the high spots are and use a hand plane to take them down. When the two pieces of hardboard mate perfectly even when one of them is flipped end to end, you know you have two perfectly straight edges.

If you haven't used them for a while and wonder whether they're still straight, just put them edge to edge again and you'll know right away.

Nope. all you prove with that is that the concave edge of one matches the convex edge of the other. to get your own edges straight without a reference edge you need to make 3. when all 3 match with all 3 flipped both ways you know you have it straight.