Table Saw Fence Scale...How could this happen?

[Jason Morgan]

I realize this is probably one of those slap you in the head DUH things, but for the life of me, I cant figure it out.

Ive been getting pretty anal about the setup of my table saw (PM64A with accufence) and recently started a project to build a router cabinet. I just checked and rechecked miter slot to blade...good...fence to miter slot...good...fence square to table...all good. I set the fence to 13" and measure from the inside of a tooth to the fence...perfect 13.00". I get to cutting up large PW panels and set the fence for 32" and make a cut. I measure it out of curiosity and its 32 1/16" :confused:

Then I set it back to 13" and its dead on. I try 45" and it off by 1/16th. at about 20" its off by a 1/32nd or maybe a shde more. How can this be? I dont remember it being off before, but I cant say Ive ever checked.

Am I going crazy? I dont even know what to adjust.

[Walt Pater]

I dont even know what to adjust.

How about your tape measure? I dropped mine once and it was 1/16" off on the pull measure, but only on one side of the hook. Is that the way you are checking the width of your cuts? Cheaper to replace the tape than your PM, I figure.

[Don Frambach]

Unfortunately, table saw scales and tape measures suffer from production innacuracies due to the way they are "printed". I read a detailed discussion of this problem a while ago and will post it if I can find it (unless someone else does so first.)

[Rob Blaustein]

I had the exact same experience recently. The problem is with the stick on rules--they are just not reliably accurate to that degree--1/32" out of 20-30" (and just for fun, check your distance at 1", I bet that's off a bit too). I have a Grizzly cabinet saw and had never gotten around to putting the tape on but I did order a Starrett tape instead. I installed it the other day, zeroed it/calibrated it at 5" inches and never checked it at 20". I cut a bunch of 20" wide panels and noticed I was around 1/32" too small. Now that's not too bad and I can probably reset the hairline by 1/64th so I'm 1/64" too big at 5" and 1/64th too small at 20" and it will probably be right on somewhere in between. Of course wood moves easily that amount, though perhaps not plywood, which is what I use a lot of. For now, I set the fence and double check with a tape measure (which also may not be that accurate). You can't win, eh (unless you get one of those really fancy fences).

[Kelly C. Hanna]

I never mount tapes on my TS's anymore...they are wildly innacurate compared to tapes and rules. I measure each time manually and go by that measurement only.

[Jamie Buxton]

Tape measures and the usual stick-down tapes are printed by a rolling press. If the tape squooges around a bit on the ink while it is on the rolling press, the tape comes out a little inaccurate. For more accuracy, get a steel machinist's rule. They're machined or etched, and are much more accurate than tapes. You probably have seen the little 6" ones. From machinist-supply houses like Enco, you can buy ones which are 36" or 48" long. You can cut them with tin snips if you need an in-between size. A typical spec for a 36" one is that it is accurate to +/- .005" everyplace along the rule. Generic 36" ones cost about $50. You can stick it to your table saw with double-stick carpet tape.

[Barry O'Mahony]

I get to cutting up large PW panels and set the fence for 32" and make a cut. I measure it out of curiosity and it's 32 1/16" :confused:Maybe, maybe not. More accurately, you have a panel that the TS scale says is 32" wide , and your tape measure says is 32.0625".

Which scale is more right? I have no idea. If you stick with using one, it usually doesn't matter. But if you're going to be using two different instruments to measure the length, you should know how they compare, unless you know they are both traceable back to NIST standards at a certain precision And you're not going to get traceability from printed scales: as another poster mentioned, they are printed off of rubber wheels and are notoriously "sloppy". There was a discussion about this a little while back on the thread about the Starrett Digitape; do a search.

[Dave Fifield]

This is precisely why I replaced my Jet Xacta fence/rail with a very precise Incra LS-TS (http://www.incra.biz/Products/TSLS.html) table saw fence system. I have never looked back.

I helped a bunch of Boy Scouts do some basic woodworking a while back. The major lesson I left them with was a slight variation on the old saw, namely "measure twice using the same rule, cut once". I showed them several measuring devices (tape rules, metal rules, stick-on rules) side-by-side to show how inaccurate they were. Some were up to 1/8" off over just 12"!!

[Norman Hitt]

Isn't it amazing, that with all the High Tech capabilities we have today, that unless you buy very High Dollar Specialized measuring devices, And many other things as well, that the quality and accuracy are not anywhere as good as the much less expensive things we bought 50 years ago? One of the most accurate measuring devices I have is an "Ancient" Lufkin 24" Wood Rule with Brass ends, model No. 34 V. It is accurate from one end to the other. I have probably 7 or 8 tape measures around here, and NO TWO of them match each other when laid side by side.:mad:

[Matt Meiser]

I had the exact same problem, only mine was worse. I bought a Starret tape so I can have a better chance of accuracy, but I've started using a metal rule to check critical cuts.

[George Summers]

One problem that no one has mentioned yet is the material that the tape is made of. Many of the stick-on tapes that you buy for a tablesaw are of mylar or some other type of plastic. As you lay these down, they stretch. I went through three before I bought a metal self-adhesive tape. The metal did not stretch when I laid it down and is very very close to being perfect all the way out to 36".

George

[Jason Morgan]

Thanks for all the comments. I didnt know how big a problem this was. I guess this has always been a problem and I just havent noticed it.

I think I am going to try to build a jig with a 36 or 48" steel rule so that I can easily set up larger cuts with. Like I said before, when I set up and measure at 10"-12", everything is dead on. Im ensivioning something like a big T-Square with a steel ruler as the long leg.

I also had recently replaced the face on my TS fence and thought that somehow that might be messing things up, but I couldnt see how.

Thanks again, Ill post some pictures of what I dream up.

[Kent Fitzgerald]

There is another possibility: if the fence rail along the front of the saw is bowed, the fence will go out of parallel as it is moved away from the blade. This in turn will change the blade-to-fence distance.

Check the fence parallelism at a few different settings. I doubt this is the entire problem, but it could be a contributing factor.

[Jason Morgan]

Kent,

Good call. Ill check that tonight. I could not think of a mechanism for variable table settings (assuming the tape itself is correct), but that is one.

Thanks!

[Andrew Shaber]

A few thoughts:

--> put your tape measure against your TS rail rule. You'll clearly see how close they are.

--> As stated above, try to use the same rule as much as possible.

--> only get anal about cuts where it is critical to be anal. The same setup 1/16" off can make 20 pieces exactly the same.

--> Story stick or cut to a correct fit. It doesn't matter if a cut is an exact length if it doesn't fit. I tend to avoid the tape measure and prefer fitting my pieces.

[Brandon Shew]

This is precisely why I replaced my Jet Xacta fence/rail with a very precise Incra LS-TS (http://www.incra.biz/Products/TSLS.html) table saw fence system. I have never looked back.

Me too. I love my Incra.


I would think that someone could produce a decent version of a tape rule though - regardless of the type of fence. Even 1/32" off over 120" is bad given todays precise manufacturing techniques.

[Barry O'Mahony]

One of the most accurate measuring devices I have is an "Ancient" Lufkin 24" Wood Rule with Brass ends, model No. 34 V. It is accurate from one end to the other.How do you know?

[Peter Gavin]

Does any one have a recomendation on where to buy the most accurate fence measuring tape? I have got to replace the one on my new fence rail, but would rather get an accurate one as long as I'm about it. I saw Lee Valley sells jig tapes that would do what I want, but at $2.50 a piece I wonder how accurate they would be.

Thanks in advance

Peter

[Michael Ballent]

Bies, and Kreg both sell them (through Amazon)... I am thinking of getting one of them, not sure which though

[Steven J Corpstein]

replace it with one of these?

http://www.wixey.com/fence/index.html

[Joe Chritz]

I have two measuring rules that get used regular.

One on my Table saw fence and a 12' stanley that measures everything else. I shop around until I find one that matches the tablesaw fence (they normally are so close I can't see them out) and use that for all measurements.

If using a tape remember to check the inside/outside measurements to make sure they are correct.

Joe

I like that digital fence. It would increase repeatability a ton.

[Rob Wilson]

This might work. Haven't tried it myself, but for a stick-on tape it has the advantage of being steel (and Starrett).

http://www.hartvilletool.com/product/11159

[Rob Blaustein]

This might work. Haven't tried it myself, but for a stick-on tape it has the advantage of being steel (and Starrett).

http://www.hartvilletool.com/product/11159

That's the one I got that I referred to in my post above (previous page)--and it's not as accurate as I would have thought.

[Barry O'Mahony]

Bies, and Kreg both sell them Does either Biesmeyer or Kreg publish a specification one their absolutle accuracy? Probably not. So how do you know they are any better than what you have now?

Any measuring tape produced by a printing process probably isn't going to have a high degree of absolute accuracy. Unless the manufacturer produces specs that say otherwise, you shouldn't assume better than half of the smallest graduation.

If you want better, you'll need to get something for which the manufacturer publishes accuracy specifications. These are usually engraved things like machinists' rules.


That's the one I got that I referred to in my post above (previous page)--and it's not as accurate as I would have thought.
Rob,

How do you know that the Starrett tape is the inaccurate one, and not the thing you're comparing it to? ;)

[Norman Hitt]

How do you know?
I laid it against a machinest friend's steel Starret Rule. (I know that's cheating, eh?):D

[Ken Massingale]

There is another possibility: if the fence rail along the front of the saw is bowed, the fence will go out of parallel as it is moved away from the blade. This in turn will change the blade-to-fence distance.

Check the fence parallelism at a few different settings. I doubt this is the entire problem, but it could be a contributing factor.

That's what I had with the Ridgid TS3650 new. The front rail was slightly bent outward. The scale was dead on with the fence near the blade and gradually became off till at 36" it read a sixteenth off. A new rail and it's dead on all the way down the scale.
HTH,
ken

[Rob Blaustein]

Rob,

How do you know that the Starrett tape is the inaccurate one, and not the thing you're comparing it to? ;)

A good point. Here's why I suspect the Starrett tape. I "calibrated" my hairline indicator on my fence by cutting a 5" piece of wood using the fence. I measured the width of that piece using my Starrett machinist square and checked it using 2-3 tape measures. The square and tape measures agreed with each other--obviously to within some margin that I don't recall, but it was likely within 1/64 to 1/128th of an inch (I also checked with a digital caliper, but can't recall how close it was to the others). By calibrate, I mean I then adjusted the hairline to match the measured width. Of course there are inaccuracies here as well due to parallax. THen I set the fence to 20" wide, cut a piece of wood, and it came out 1/16" too narrow, as measured using the same tape measures as above. Now you could say that if I measured my 5" piece to within 1/48", then 1/16" would be within tolerance, but I think I was better than 1/48" with the 5" piece. But I should probably check it out again before badmouthing Starrett. Sorry for the long answer.

[Hank Knight]

Jason,

My last tablesaw was a Powermatic Model 64 "Artisan" saw. I think it was the precursor of yours. It had a Vega-type fence with a round, chromed rear rail. The tape was fixed to the rail and was curved accordingly - in the opposite direction from a regular spring-loaded tape measure. It was never very accurate, so I didn't rely on it. One day the glue quit and the tape popped off the rail. I tried to find an easy replacement locally, but everything was either flat or curved int he wrong direction to fit on the rail. I finally got Powermatic to send me a replacement. IIRC, it took several phone calls, waiting, and finally a hissyfit on my part to get their attention, but they sent me a replacement tape. It was fine, on-the-money accurate and easy to read (My original one had gotten scuffed up and hard to read). I kicked myself for not ordering a replacement earlier. You might want to give Powermatic a call and see if they can offer you some advice - or a new tape for your saw. I hope you get better response from PM's Customer Service than I did.

Hank

[Steven Wilson]

The tapes on table saw fences are often inaccurate. I use to have a tape that had a 16" section that was 1/32" off. I ended up having the manufacturer send me a couple of tapes and I installed the one that was the best. When you install a new tape be carefull to lay the new tape on straight and to use consistent pressure in pulling on the tape. It's fairly easy to put a 1/32" warp in a tape when you're installing it.