Table flatness

[john_001]

Two diagonals do NOT verify a plane.

True. The diagonals would only indicate a twist. It could still be flat from side to side or front to back at any point.

[dusty]

Two diagonals do NOT verify a plane.

OK...tell me how to do this.

[bainin]

With no limits on $$...you would use an optical profilometer like this.

https://nanovea.com/instruments/st500/

b

[HopefulSSer]

With no limits on $$...you would use an optical profilometer like this.

https://nanovea.com/instruments/st500/

b

I suspect a brand new 520 upgrade kit would be cheaper! ;-)

[JPG]

Two diagonals do NOT verify a plane.

OK...tell me how to do this.

Using a single corner point repetitively while checking at other points/lines/angles using a straight edge. A time consuming/exhaustive process.

That is for a table. Checking trunion mounts surfaces is more difficult unless thee has a known flat surface for reference.

[RFGuy]

I agree that sometimes we chase our own tail in the shop with certain alignments trying to make them near perfect when in reality woodworking is often an endeavor where a 1/16" of variance on a cut piece is acceptable for many applications. So, to me this begs the question of how much variation is acceptable for tablesaw flatness? I did a little digging online and found a Fine Woodworking magazine discussion on this and one of the commenters claimed to work at a company that sells tablesaws. See his comment below:

Excerpt:
"Where I used to work, we'd check tablesaws before they go out the door and anything under .010" was acceptable and under .005" was considered good."

Given this piece of information, I believe both HopefulSSer and David commented on experiencing upwards of 0.030" which is 3x the "acceptable" limit by a competing tablesaw manufacturer, if this online comment is to be believed. Also, this competing tablesaw manufacturer likely has a much larger tablesaw table than the Model 500, so comparatively this 0.030" Model 500 table error is proportionately much larger in comparison. So, does anyone know how much variation in tablesaw flatness is acceptable? I mean a 0.030" dip at the blade for a Model 500 main table is a 0.25° error, if my math is correct. Maybe not a big deal for some furniture and other woodworking applications, but likely a horrible error, e.g. if you are making segmented bowl sections, right? Depending on what you are building, I think David and HopefulSSer have the right approach in attempting to correct this large error in main table flatness. For me, I have always wondered about table flatness on my Mark V due to it being a thin aluminum table. Honestly I would have liked to see more ribbing or other structural supports underneath the table to stiffen the Shopsmith design. Every time I used my Mark V in drill press mode with the hollow chisel mortiser, the amount of main table flex was significant. With time and a lot of drill press usage, I could easily see a Mark V main table becoming warped - from normal use (NOT abuse). Something that would never happen on a standalone tablesaw with a cast iron table by comparison.

[Hobbyman2]

At some point understanding spring back and what causes it will define how or if a table can be re flattened, from my own experience spring back is caused by the metal stretching, bending it back does not eliminate the stretch ,it can actually exaggerate it and at some point , unless it is re heated and the molecules are realigned ,,, it will / can snap or warp some where else .if it is bent that bad the bend should register through the entire system including the legs , I wouldnt want to risk damaging any thing else like the carriage or carriage locks ,,, JMO maybe find another use for it , maybe mount a miter saw to it or make a router table out of it . To answer the original question has any one tried to fix one? hard to say , these things have been around for decades ,maybe mounting angle iron under it some how ??? I can not say I have ever seen one that was repaired . these are made from aluminum not cast iron . I dont think it is ruined I just think you would be further ahead to replace it . it would be a good time to upgrade and enjoy the benefits the new table systems have to offer .again JMO and good luck .

[dusty]

I was just watching a MGR video on setting the 90* stop of the table in drill press mode by using a sanding disc and feeler gauge. Clever way of doing it, but it wouldn’t work on mine (or at least no better than anything else I’ve tried*) because neither of my two 500 tables are flat. They’re both down a bit in the center, I presume from years of force from drilling.

I’ve very carefully put them in a hydraulic press and tried to push the middles back up but with only partial success. There’s a lot of springback and I’m afraid to press to far lest I crack the table. Given that 100% of the tables I’ve ever seen (n=2) have this problem, I presume it’s common. Is there a generally-accepted solution? I searched but came up empty handed…

Thanks!

(* I just try to average out the error)

Should have said this earlier BUT I doubt that your sanding disks are "flat" though they might be "good enough" to use for this purpose.

Also: If you really suspect that the main table is warped as a result of it being attachedit to the trunnions - disconnect it from the trunnions and mount it like a floating table (using the table rails and tubes and no trunnions).

[JPG]

- disconnect it from the trunnions and mount it like a floating table (using the table rails and tubes and no trunnions).

How does one do THAT with a M5/500?

[dusty]

- disconnect it from the trunnions and mount it like a floating table (using the table rails and tubes and no trunnions).

How does one do THAT with a M5/500?

I guess you're just stuck with what you got. The remaining solution is to be satisfied with what ever precision you can achieve.

Besides that - the stated measurements all through this thread are probable an order of magnitude better than the woodworking produced by any/all of us.