Wobbly table saw fence

[chillfok]

I bought a dial indicator in order to check my table saw's alignment since it's pretty new (1 month old). As I was running some tests I decided to check the fence and found it's all wobbly. When I mean wobbly I mean by thousands of an inch so no big deal (maybe) but I was wondering if this is normal. The worst spot deviates by 0.008". Watching the attached video should make my question clearer.

table_saw_fence.MOV

[Brendon_t]

I've never measured my fence for flatness but honestly don't think I would be worried about 8 though. I don't cut joinery along the fence and an 8 thousandth deviation along a rip wouldn't make me blink an eye.

[bleedinblue]

So by wobbly, you don't necessarily mean the fence moves, but that the fence is not dead straight. 

How does it cut?   If it cuts well, don't worry about it.

 

 

[..Kev]

Make a cut and measure your piece at each end to check the results.  You can pick apart any machine with a dial indicator.  Cut some parts and see if they fit up to your expectations.

[gee-dub]

Mine was a bit wonky as well.  The UHMW faces and aluminum extrusions can be just as bad.  Fortunately Saw Stop makes it very easy to resolve this if you are so inclined.

The steel tubes on t-square style fences are not milled and can show the deviations just as aluminum extrusions do.  My Biesemeyer had similar issues but, required the near-destruction of the faces in order to correct this problem so I never went there. 

We'll assume your sled is riding snugly in the miter slot as is not a contributor. Put a pencil mark on the face where ever it is farther away from the dial indicator than the closest point.  That is; fond your most proud point and consider that "zero".  Mark any area that is less than "zero".  This will be various spots along the tube.

Flip the fence over and use the provided ball-tip hex tool to loosen the face. 

As an initial check I measured the blue machined extrusion and it was fine.  I then mounted it to the tube "as-is" and the same deviations were present as with the stock face.  The extrusion conformed to the deviations in the steel tube just as the stock face had.

Shim the tube (I used masking tape) and re-attach the face to re-check.  It took me three or four tries to get within .001" all along the length. 

I checked the stock face while it was off and it is very consistent in thickness.  Now that I have the tube shimmed I will probably switch back to the stock face. 

The extrusion, while milled and wonderfully flat, makes a lot of my jigs non-functional as they expect the lower fence face.  As mentioned, the extrusion, although quite stout and very true in and of itself, was not better than the stock face until I shimmed the steel tube behind it.  We'll see ;-)

 

Edited May 29, 2016 by gee-dub
64

[Ronn W]

I recently trued my fence a little by shimming between the steel tube inside and the aluminum fence faces with pieces of paper. Depends on how your faces are attached to the steel tubes.  In my case the straightness of the aluminum fence faces were sensitive to how tightly they were fastened to the steel tube.

I decide to make my adjustments only because I was getting slight burn marks ripping maple or cherry with a thin kerf glue line blade.  Other than that my cuts were fine.

 

[chillfok]

Thanks everybody for the prompt responses! 

I have started woodworking very recently and I don't have enough experience to know if this will indeed affect precision (my engineering background makes me attracted to shiny measurement devices and gear of all sorts, hence the dial indicator purchase). 

So far this fence hasn't affected me and cuts seem to be parallel to at least 0.003" (with variable results). Cutting sheet goods on the other hand leaves me with incredibly imprecise boards but I think that's mostly if not all operator error. Small cuts are fine though. 

I will attempt going at it like gee-dub did. I think that will satisfy a good part of the dial indicator OCD-ness. 

bleedinblue: No. When I fix the fence in place it stays and there is no play. Also, the sled runners have no play either.

[gee-dub]

I just got back to a project I am working on for my own home.  Funny how our own stuff comes last ;-)  Anyway, this is the first rip cuts I have made since shimming the fence face.

My old hybrid had been dialed in and always gave me glue ready rips.  I have been meaning to tweak the Saw Stop and as shown earlier in this thread, I finally did.

The cuts are more what I am used to; glue ready, no hand plane required.  I only post this to add credence to all the good advice given to chillfok.  the extra effort is worth it and will pay dividends way beyond the time you invest.

 

[Andy Wright]

My Biessemeyer was a little wavy, and when I replaced it with a VSCT fence is saw an immediate improvements in my cuts. 

[chillfok]

Two and a half months later... I went and pulled the fence apart like gee-dub suggested. Shimmed some bits with blue tape and improved my fence quite a bit. It's not perfect like I'd want it but it's better than before. 

I did find something interesting after an hour and a half of taking the fence apart and putting it back again. Before I tell you my findings, here's what I did:

1) Marked a spot on the fence as "ground zero" and measured all the peaks and valleys throughout the fence relative to this spot.
2) Took the fence apart, that is, undid the screws from the bottom and pulled the plywood face from the tube.
3) Put it together again WITHOUT touching anything else, no shimming, no adjusting, no nothing. Just put the damn thing together.
4) Fully tightened each screw.
5) Repeat the measurements always making sure "ground zero" was in the exact same spot on the fence.

I repeated the steps above three times and noticed that I was getting measurements that had a 0.003" margin of error. I thought tightening the screws alone wouldn't have such a "big" effect on the plywood but that's what it is. So I suppose that the SawStop fence cannot be adjusted to an error less than that just because it's impossible to tighten each screw exactly the same every single time.

Thanks everyone for their help!