520 Fence Is Not Square to the Table Surface

[BuckeyeDennis]

I have to wonder if the straightness and squareness issues you're measuring are really due to damage, or to normal dimensional variations in extrusions. My 520 fence isn't precision-machined, it's just an aluminum extrusion that has been cut to length.

I just checked my own 520 fence with a 24" machinist's rule, and found that none of the surfaces are perfectly flat. But I apparently got a pretty good one, as I didn't find any deviations of more than about 0.002", including down the full length of both sides (the worst side was 0.0015" concave).

Here's a link to a document with standard tolerances for aluminum extrusions. The tolerances vary with material temper and geometry, but a shape like the 520 fence is apparently allowed a standard straightness tolerance of at least 0.050" per foot of length!

https://www.mkmetal.net/images/uploads/ ... erance.pdf

So I'm guessing that Shopsmith pays for tight-tolerance custom extrusions, but they're still unmachined extrusions.

[rlkeeney]

I have to wonder if the straightness and squareness issues you're measuring are really due to damage, or to normal dimensional variations in extrusions. My 520 fence isn't precision-machined; it's just an aluminum extrusion cut to length.

Obviously, this is a guess on my part. It is possible that it came this way from the factory and no one noticed it until now. As I mentioned earlier, I bought a used one, and until now, I have never used it for anything other than ripping. I ripped a handful of of 1/2x1/8 inch strips of cherry, and it worked great for that. Everything was dead square and straight.

[DLB]

You mentioned earlier in the thread that the extrusion itself, despite some questionable characteristics, measured square. The top of the main casting (Fence Base) is a machined flat surface, roughly 1-5/8" wide, that mates with the inside of the top of the extrusion. For the fence to be square to the table, this machined surface must be parallel to the table rails (and table itself). It seems to me that it would take only a small error here of a few thousandths to create the out-of-square condition you see. IIWM I'd measure this surface before ordering a new extrusion. I would trust my angle cube for this even though it technically may not have the required accuracy. This surface would also be a good spot for a permanent shim solution if the problem is the casting.

- David

[rlkeeney]

I can't buy a new extrusion from Shopsmith right now, and I don't see any suitable 520 rip fences on eBay. So I am giving up on my fence-straddling jig and have purchased a Shopsmith tenoning jig that I think will accomplish what I want. I'm not too concerned about the fence, and the aluminum tape should last a long time.

As for purchasing a used fence, this will probably not happen. The fence I have works well for 99.9% of what I use it for, and there is always the chance that I have overlooked something, and a used fence would have the same problem.

[dusty]

If the objective is total alignment to within a sew thousands of and inch - forget it. You might achieve that but you won't be able to maintain. I used to align (or attempt to) to within .005" but I gave up a long time ago. It was fun but making saw dust was more rewarding.

Besides - with my eye sight being whar it is I can no longer see a .005" error.

[rlkeeney]

If the objective is total alignment to within a sew thousands of and inch - forget it. You might achieve that but you won't be able to maintain. I used to align (or attempt to) to within .005" but I gave up a long time ago. It was fun but making saw dust was more rewarding.

Besides - with my eye sight being whar it is I can no longer see a .005" error.

I may never have noticed this if the cut had not been off enough for me to see it. That, combined with me trying to find out why my jig was tight on the back of the fence and loose on the front.

I sympathize with your eyesight issue. I have macular degeneration in mt right eye that gives me problems from time to time.

[rlkeeney]

My fence is square to the table thanks to several layers of aluminum tape. My jig is also square to the table. I had to tweak my Shopsmith alignment a bit and make a new zero-clearance insert for my 10-inch industrial Freud rip blade. It works perfectly. Today I cut a test piece with a 1/8-inch slot 2-5/8 deep, dead center in a 3/4-inch board. I have a stack of 10-inch blades, and not a one of them cut a 1/8-inch kerf.

Here is the test piece:

2026-05-27 16.29.18-1.jpg (147.65 KiB) Viewed 17396 times

All this just to cut the slot for the tail of this toy F-16. This is not my build. I haven't done it yet.

This is a rendered image from the plans.

In hand F16.resized.jpg (36.15 KiB) Viewed 17396 times

[dusty]

Sure do wish I could see this same piece cut without the fix (aluminum tape). Actuall, I wish I could tinker with your fence. I have two and I would call them both square to the table. However, mine are not perfectly straifht (end to end). What I find confusing about mine is that the miter bar moves freely in all miter slots.

[rlkeeney]

Here is a test cut with no attempt at centering. Measuring the right side with my calipers:

Top 0.382 inches
Bottom 0.312 inches
Difference 0.07 inches

Not much, but easy to see with my eyes.

2026-05-28 09.08.11-1.resized.jpg (115.04 KiB) Viewed 17344 times

[dusty]

Thank You. I can clearly see that the cut is not centered but what I don't see is the indication that the fence was not square to the table. Maybe I am looking for something that you did not mean to show. I gotta go back and read more carefully.

I am curious though. How do you measure (with what) to center the kerf. Not a real simple setup!