How do you Saw large panels Dead Square?

[dzj]

A handy gadget, Steve.
Although, I doubt that any extrapolation method will rid the OP of his 0.5mm woes.

 

[Phil Pascoe]

If doing Pythagoras triangles, note that most tape measures are not accurate enough. For 0.5 mm/m you need a EU Class II (printed on the tape near the start) at least, preferably Class I. Most of those in common use are class III or unclassified. Cisco are the only Class I measures easily available: http://www.thetapestore.co.uk/tapes-rul ... e-measures.

Many squares aren't, either. They can be checked by drawing a "square" line to a straight edge (e.g. factory edge of an MDF sheet) then reversing the square and repeating. The lines will usually deviate, and the average of the two is the true right angle.

No such thing as exact; there is always a tolerance (error). "Dead on" just means "I can't see the error". I agree with others that 0.5 mm/m in a table top is entirely within tolerance.

Keith

Fisco. (sorry, it's the pedant in me). Your point about inaccuracy is fair, but to a woodworker to a large extent as long as the same measure is used for everything it doesn't matter too much.

 

[mickthetree]

Excellent vid Steve! Off to make one of those doohickys now!

 

[bugbear]

Am I missing something here? Hold a straight edge against your square, mark the edge and cut it. Even if you are a couple of mill out it will not be noticeable on a metre wide table ( this is one reason why your table top will overhang the base by say 10-15mm… a few mm difference in size will not be noticeable)

Agreed; I yield to no man in my love of precision, but designing out the need for precision also carries much merit.

I recently read Lovell's book on the making of the Jodrell bank telescope, and at one point
they redesigned the steelwork holding the main parabola shape. Clearly, the steelwork
needs to be stiff enough to hold the parabola accurately, but (clearly...) it can't be infinitely
stiff. Lovell went back to fundamental geometry so that the unavoidable
movements of the steelwork distorted the parabola in ways that had a minimal
effect on the focusing - in other words, he didn't reduce the amount of flexing,
he reduced the effects of the flexing on the quality of the telescope.

This allowed him to reduced the weight of the dish, which (in turn)
reduced the requirments of the support for the dish, and the mechanisms
that moved the dish.

Wondrous!

BugBear

 

[MusicMan]

[/quote]

Fisco. (sorry, it's the pedant in me). Your point about inaccuracy is fair, but to a woodworker to a large extent as long as the same measure is used for everything it doesn't matter too much.[/quote]



Phil, yes of course Fisco. Damned autocorrect! I like pedants

Normally you are right that using the same measure for everything takes out tape measure inaccuracies. But not if you are doing Pythogoras 3,4,5 or whatever, since the inaccuracies are non-linear (in fact from the manufacturing process, they tend to be cyclic). So the 3 may be a bit low, the 4 about right and the 5 a bit high, which would screw up the right angle. Using the same tape measure would only work in this case if the errors were linear (like an accurate tape that has been uniformly stretched).

Keith

 

[moosepig]

You can make a reasonably accurate square with a tape measure and eliminate cyclic and other inconsistencies in the tape by using the same measurement all the time. Mark out a rhomboid (i.e. try and mark a square) which must by definition have 4 sides of equal length. Draw lines between opposite vertices - these MUST cross at exactly 90 degrees if your 4 sides are the same length, regardless of whether it's a square or a rhombus. Cut along these lines - if you can do it accurately enough you end up with 4 pieces that have 90 degree corners.

 

[bugbear]

Of course, you can make your own "scale" during the geometrical construction of a 3-4-5 triangle.

The trick is to use a straight edge to draw the diagonal (longest side) first. Then set your compasses
or dividers to roughly 1/5th of the length of this line (but it must obviously be less than 1/5, not more).

Mark 5 divisions (6 marks) on this line. You now have both your diagonal, and a (theoretically...) perfect scale.

Set your compasses to 4 units, strike an arc.

Set your compasses to 3 units, strike an arc.

The intersection of your two arcs is the third vertex of your triangle. Join the two ends of the diagonal
to this vertex and you have an accurate right angle.

It's all greek to me. :lol:

BugBear

 

[MusicMan]

Ingenious solutions still emerging!

Keith

 

[Lons]

It's all greek to me. :lol: BugBear

:lol: :lol:

Always used to make me laugh when my students said "what do I need maths for? I'm going to be a joiner / brickie not an accountant". (Few of them could spell accountant either :wink: )

 

[Steve Maskery]

Here is how I use a Square of Thales to set my saw track square to the edge.
You need a SOT, a straight board with a notch cut in it and the straight edge you are trying to position, in this case my saw track.

I'll grant you it is a bit of a three-handed job, but the track is held in the notch and then adjusted until the SOT touches at three points.

I'm sorry I appear to have cut off the BL point, but I'm not going out to re-take it, it piddling down here. You get the picture. Well, most of it.

 

[Sgian Dubh]

However, you can get somewhere near "I can't see the error" without any measuring device at all.

A neat device, Steve. However, I feel a need for pedantry: you say measuring the diagonals of a rectangle proves (or disproves) square, which is true. However, I think it's worth pointing out that a regular trapezium should also measure the same across the 'diagonal' corners.

Not something to get too excited about normally, I know, but there it is. It may matter if, for example, one rail out of two in a four member frame is a few millimetres longer than the other one, although there may be other clues to this error, e.g., triangular shaped gaps at the shoulders. Slainte.

 

[Steve Maskery]

Did I say that? I don't remember. If I did I was mistaken, for exactly the reason you describe. A traditional window-cleaner's ladder has two equal diagonals...

 

[bugbear]

Did I say that? I don't remember. If I did I was mistaken, for exactly the reason you describe. A traditional window-cleaner's ladder has two equal diagonals...

You're in the clear you say that iff the stiles and rails are the same length AND the diagonals are the same, it's square.

BugBear

 

[Eric The Viking]

You can use a framing square to build a big try square, just by bolting hardwood to either side of one leg. Once done, getting it square is easy, and you can always unbolt the hardwood whenever.

On Class 1 tapes, I bought one recently very inexpensively: Advent Master Precision (5m). It was less than 10 quid incl. postage, AND IT IS BRITISH MADE.

When it arrived, I used it to check my other tapes: all bar my 'site survey' 50m fibreglass tape were pretty much dead on over 5m. None of them were expensive (I'm cheap, me). Thus, although it's ergonomically a better tape than the others, it wasn't strictly neccessary. And no, I didn't check that it was 20 degrees Celcius at the time ;-)

HTH,

S.