Checking straightness of a Track Saw Guide Rail

[Steven Owen]

A lot of people complain online about rails that aren’t perfect straight.  They claim the rails are out by as much as 16th of an inch.

The tool adjusting (measure within .001 of an inch straight edges from Woodpeckers and Lee Valley only come in length up to 48 inches.

What method or procedure is being used to determine the straightness of a rail?  Is it simply a matter of making test cuts and measuring the cuts for consistency or so there a more scientific approach to checking the straightness of a rail to make sure you didn’t buy an assembly line lemon.

 

[waho6o9]

I never thought of checking it, just clamp down to your marks and go for it.

Once in a while I check the diagonals and they're always on the money.

 

[Oldwood]

I just lay the guide on a piece of sheet stock as long as it is and mark a line along the length of it with a fine pencil and then turn it 180 degrees and line it up with the mark and see how it looks against the line. If there is a bow the difference you see is twice the error.

In the store when buying a Makita rail I just layed on the box it came in and used it to draw my line.

Not very high tech but gives you an idea how straight it is.

 

[Alanbach]

Maybe I am missing something here? As long as you make a fresh cut in Festools splinterguard that sticks down to the rail and the differences in parallel between the groove in the track (that the saw rides in) and the edge don’t matter because the splinterguard accounts for the difference. I get the concern if you change the blade and it no longer makes direct contact with the splinter guard of if you connect more than one rail together. Again, am I missing something? I don’t think it matters if the rail is crooked, it only matters if the groove in the track is crooked.

 

[Cheese]

If the rail is 8’ or less, I use an 8’ level against the male form track ridge. If the rail is over 8’, I use a chalk line against the same surface.

 

[lwoirhaye]

Some products like square steel tubing are quite straight.  You can go to a metal dealer and assess that by comparing the tubes, flipping them 90 degrees, checking for gaps and so on.  One or both tubes may be out of straight but if you do a few you'll find some that check out straight in relation to each other.  I use a 78" level and it's good enough for woodworking generally.  I bent one in moving and had to replace it... there was a bit of head scratching before I figured out it was bent.

For a test cut, a cheap pine 1x10, piece of mdf etc. can be cut down narrower on both edges and then measured to see if it is hourglass-shaped or has a belly in the middle.

I doubt they check them very well at the factories.  I think aluminum extrusion generally produces pretty straight parts and makers like Makita may feel the cost savings they get from whatever they do to make them cheaper is worth it.  Not all track saw users work to really fine straightness tolerances.  I have a mechanical edgebanding press so my cuts have to be very straight to get the banding on tight.  Most of the straightness concerns I worry about around that would not concern somebody edgebanding using other methods as much.  If you had a job to make a bunch of huge veneered panels for a boardroom installation you'd look like a dope if your cuts weren't very straight. 

 

[Corwin]

Maybe I am missing something here? As long as you make a fresh cut in Festools splinterguard that sticks down to the rail and the differences in parallel between the groove in the track (that the saw rides in) and the edge don’t matter because the splinterguard accounts for the difference. ... Again, am I missing something? I don’t think it matters if the rail is crooked, it only matters if the groove in the track is crooked.

Each feature of the guide rail's shape, be it the T-slot, the back edge or the raised 'hat' that guides the saw, are parallel with each other. From one rail to another you may find slight differences with the width of the hat, or the distance between the hat and the back edge of the rail, but within any single guide rail each feature is still parallel with every other feature of that rail. So, if the guide rail is crooked, then so is the "groove in the track" that the saw rides in -- aka the hat. And the splinter strip won't fix that issue.

 

[Michael Kellough]

As Corbin says, “the hat” or spine as I call it, is the only thing that counts in determining if your guide rail is straight, if making straight cuts with your track saw is the goal. Further, it’s really only the splinter guard side of the spine that is critical (the side the non-adjustable part of the jibs (gibs?) rubs against).

So, if you want to check straightness of the guidrail against another straight reference tool just hold that tool against that side of the spine.

In my experience that method is only useful to confirm that running over your guide rail with your truck had an adverse affect. I don’t know any straight reference tool that is the right height and length to conform well enough to the spine to get a meaningful reading (which would have to be in the .1mm range). Long extruded aluminum levels that cost half as much as a long extruded aluminum Festool guide rail are no straighter than the guide rails I’ve tested (and the height of said level laying against the spine of the guide rail is awkward at best).

The only valid way to test the straightness of a guide rail is to trim two stable seasoned boards (use boards about a foot shorter than the guide rail, plywood is okay) then turn one of the boards around (not over) and push them together to see how closely they match. If there is no measurable gap anywhere along their length you’ve got the rare perfectly straight guide rail. When testing my FS 2700 guide rail there is a gap of 1/100th inch in the middle of the boards (twice the actual deviation from straightness). I wish it were less but in practice it’s only a minor problem at worst requiring no compensation.