Adjusting a rip fence?

[Lord Kitchener]

Good quality new table saw, straight rip fence, and yet there's a slight taper to the cuts. For instance, a panel 450mm long, ripped to a 300mm width, resulting width 300mm at the front (the part that went through first) and 300.3mm at the rear.

Any ideas?

 

[Myfordman]

First check and adjust the slot(s) are PERFECTLY parallel to the blade
then set the fence guide bar at 90 degrees to the slots
finally adjust the fence to be parallel to the slots on both sides and set the fence faces at 90 degrees to the table surface.
Set the blade angle stops at zero and 45 degrees too.

Nearly every saw will need fettling in this way whether it is new out of the box or has been moved.

Listen carefully as a test piece exits the cut. you should just be able to hear the teeth on the rear of the blade just kissing the surface cut by the front teeth.

Note that adjusting the mitre gauge is not a way of dealing with non parallel slots and blades

 

[Lord Kitchener]

This saw doesn't actually have mitre slots, and the rip fence is set parallel to the blade within my measuring capabilities (digital calipers depth gauge between fence and blade front and back).

Having said that, the manufacturers (and H&SE) say the fence should toe out slightly for safety reasons.

It seems to me that just as one knows which way to shim a worktop if it's sloping by checking which way it slopes, one ought to be able to know which way to adjust a rip fence by observing in which direction the resulting cut slopes away from parallel, but nothing I've read on the subject so far helps.

 

[Myfordman]

The slight toe out recommendation comes from the safety risk of kick back if the fence toes in at all. Best of all if it can be set spot on parallel to the blade.

The adjustment method depends on how the fence is designed and built. Mine is biesmeyer clone, so easy enough to do one side of the fence with the adjuster screws. If the two sides of the fence are not parallel, as mine wasn't, it can be trimmed with paper shims and similarly to get the fence faces 90 degrees to the table surface.

 

[woodbrains]

Hello,

As said above, a slight toe out is not the optimal situation, but safe. Dead parallel is best and a toe in should be avoided at all costs. BUT, toe in or out, the cut should still yield a board of parallel width. It will just do it with a wide saw kerf and usually a ragged one. It is called heeling and toeing and saps saw power since the blade has to remove more material, and increases noise. Go for dead parallel blade to fence set up if at all possible.

The only way I can see a saw not cutting parallel boards is if the rip fence is creeping along the rails slightly under pressure of the board being pressed against it. Try clamping the fence on with a g cramp or something, to see if improves things. If this solves the problem, can the rip fence lock be adjusted for increasing the pressure? Is the fence locked at the rear? Is 0.3 mm in 450 actually worth worrying about, or even measurable really?

Mike.

 

[RobinBHM]

I doubt that measuring a 0.3mm difference on a table saw is giving much information. Saw blade deflection, feed rate, flexing of the fence, dust between fence and timber could all create a 0.3mm difference. A scientific experiment would need quite a few pieces cutting and recording the range.

Most accurate ripping is usually done by using a rip fence that stops just beyond the gullets of the sawblade, at the front where the blade direction is downwards. If your test cut wasnt with saw set up like this, then I would try this first.