Acceptable accuracy?

[Hugopuk]

I have made my first table saw sledge and getting totally obsessed with accuracy, I think it's the lockdown mentality finally pushing me over the edge.

Using the 5 cut method, on a board 27.5mm long the differences between point A and B was 0.05mm, divided by the 4 cuts = .0125 and divided by the length of cut = 0.00045.

My pivot point is 90cm so the adjustment is 0.041mm.

Try and try again was something I was told but all I am getting is frustrated, Please confirm I am mad for trying to get better, as trying everything on my squares says this is acceptable?

 

[That would work]

Crack on with it and make something

 

[MusicMan]

There's no such thing as perfect accuracy, simply accuracy that is fit for purpose.

I'd say that you have done very well. What is important is the error in a demanding situation. Across a 600 mm board the error would be 0.27 mm, about an average pencil line. Another way of looking at it is that it is 0.02 degrees. I can just detect by eye an out-of-square by 0.1 degrees.

It is up to you to decide what accuracy is acceptable for what you are doing. I would say that that is easily good enough for any woodwork joints, cabinets etc The only thing that might need slightly better accuracy, as Custard has pointed out in this forum, is the fit of a mitre picture frame.

And I am fussy. I ran a metrology lab once and developed a way to calibrate transducers to 0.00000001 mm (10 picometres). But not for woodwork

 

[xy mosian]

In imperial, old money inches etc., 1mm is 39.6 thousanths of an inch. Therefore 0.1 mm = 3.96 thousanths, of an inch. Your error begins to sound quite small doesn't it?
xy

 

[sammy.se]

I have made my first table saw sledge and getting totally obsessed with accuracy, I think it's the lockdown mentality finally pushing me over the edge.

Using the 5 cut method, on a board 27.5mm long the differences between point A and B was 0.05mm, divided by the 4 cuts = .0125 and divided by the length of cut = 0.00045.

My pivot point is 90cm so the adjustment is 0.041mm.

Try and try again was something I was told but all I am getting is frustrated, Please confirm I am mad for trying to get better, as trying everything on my squares says this is acceptable?

Yep. Quite Mad.
I'd be super happy with that. Now please make something very very square and show us the fruits of your labour

 

[Hugopuk]

Thanks for the affirmations I shall do as instructed and make something square .

 

[Droogs]

and then something curved with compound mitres (hammer)

 

[transatlantic]

out of interest, now perform the test again with different lengths of wood. You'll probably get varing results, as you've tuned the fence to the specifc scenario under which you're testing. Your fence probably has a little bow, and your work piece also a little bow, and it just so happens that these pieces give you the results you're seeing, as you've effectively tuned the fence to account for the bow.

If you were to use another piece of timber that had a different type of bow, you'd have tuned the fence slightly differently. Hopefully you see the point I am making.

 

[MusicMan]

Yes, you do have to start with a straight fence and a straight reference side for the test to work.

If you start with your fence, one edge of your work and one other piece, and you work them till all three fit smoothly, then they are all straight. (The Whitworth test from the early 19th C). You don't need a standard straight edge in order to make one.

 

[Woody2Shoes]

Another point - that I haven't seen made yet on this thread at least - is that you may get different measurements from the same piece(s) of wood over time because the wood moves about (assuming that your measuring techniques is accurately repeatable anyway!).