View Single Post
  #36   Report Post  
Posted to rec.woodworking
Ed Bennett Ed Bennett is offline
external usenet poster
 
Posts: 107
Default How much runout on TS is too much

On Oct 17, 3:46 pm, Frank Boettcher wrote:
On Tue, 16 Oct 2007 10:29:16 -0700, Ed Bennett
wrote:

Some check the "factory"
alignment in a misguided (ignorant) attempt to judge the quality of
workmanship.


Ed Bennett


http://www.ts-aligner.com
Home of the TS-Aligner


An insulting and misleading statement. Leads the reader to consider
that there is no basis for quality judgement and comparison from the
alignment done on the assembly line by the manufacturer.


I don't know why anybody would take offense Frank. I really don't
think it's reasonable to expect a machine to maintain proper alignment
after riding around on forklifts, trucks, and rail cars. The
vibrations and thermal changes virtually guarantee that alignment will
be lost during shipping. If I were personally responsible for
aligning and testing table saws at the end of a production line I
would not be surprised or offended to learn that 99.9% of the machines
that I so carefully aligned arrived completely out of whack. It's
just physics. My Unisaw needed alignment, and it came with a bunch of
"shock watch" tags on it.

I think a manufacturer cannot verify that a machine is defect free
until they properly align and test it. And, the quality can be
clearly judged inferior if a machine cannot be properly aligned. But,
the state of alignment as delivered "out of the box" is pretty much
irrelevant.

The truth is unless the component parts are just terrible, all saws
can be "set" to very close to zero at 90 degrees and I suspect that
most manufacturers have assembly procedures that achieve that using
rather sophisticated set up tools. I know one does at least. As the
blade is tilted, it is exactly the "quality of the workmanship" of the
component parts that determines the reading at 45 degrees and the
difference between the two figures is an excellent indicator of the
quality of workmanship when comparing different units. The flatter
the table, the more parallel the boss plane to the top, the flatter
the cabinet top plate plane, the more accurate the trunnion/brackets,
yoke assembly and arbor assembly, the closer that 45 degree figure
will stay to zero out of the box.


I agree, there are certain aspects of alignment (like the tilt axis
parallelism to the table top that you mention here) that are dependent
on quality of manufacture. It would be incredibly easy (and
inexpensive) to implement in-process 100% inspection of every single
casting that gets machined. And, the use of the Meehanite casting
process would significantly reduce (eliminate) post machining
warpage. I suspect that only an exceptional manufacturer would do
such things. And, if they really did, then I would expect that none
of their saws would require shimming under the trunnions or between
the base and the table (or, such a small number that you would just
never hear about it).

Having heard of this problem from owners of all the most popular
brands, I suspect that they really aren't doing anything substantial
in this area. My 80's vintage "Proudly made in the USA" Unisaw needed
shimming (among things) before it would operate properly. So,
whatever Delta did before shipping my machine, it didn't help much.

While there are certain things you
can do to offset the tolerance stackup of some of those parts if
others are bad "you got what you got".


This might be how I "got what I got" with my Unisaw. If the
tolerances stack up so that the product (when fully assembled) can not
be properly aligned (without shimming, filing, or other
modifications), then any mechanical engineer will tell you that the
manufacturing process is poorly designed. Tolerances are *supposed*
to define the range of variability for which no defect can occur.
Unfortunately, too many manufacturers define their tolerances as the
range of variability for which an affordable amount of warranty
expense occurs. I'm in the "zero defect" camp, not the "acceptable
warranty liability" camp - which never made me very popular with the
bean-counter types. They were always glad that I only did the
numbers, not the decisions.

Ed Bennett


http://www.ts-aligner.com
Home of the TS-Aligner