Compass and straightedge Construction has anyone used ...?

[zefrenm]

I ended having a lot of odd shaped scrap around the garage and started using compass and straight construction to build shapes for the current omni Top builds and Build and whole new set of jigs the showed i had another framers square and sheet rock square were a touch off square. before I purchase a beam compass I want to know if anyone out there drafts there Bfm builds using compass and straight edge?

[Gregory East]

The only construction trick I have used is checking diagonals are the same length to check if a panel is overall square. It wasn't.

I would throw away the square that isn't, replace it with one that is, and carry on with tape and straightedge.

[LelandCrooks]

Sheet rock squares are always out of square. They're not designed for precision work. If you have a carpenters square that's out, you can resquare it with a punch. Google it.

[Harley]

I want to know if anyone out there drafts there Bfm builds using compass and straight edge?

I do for setting up my sheetmetal patterns and anything new that is one off or similar complexity to a DR.

Here's the simple way to see if your square is square;

[zefrenm]

For the Dr even the Tuba 60 it seems to make more sense to build the parts from perpendicular lines off the midpoints than to put faith in measurements off thee assumed true side of a broad. I wish school did more with geometric construction than a Sidenote in algebra.

[Harley]

For the Dr even the Tuba 60 it seems to make more sense to build the parts from perpendicular lines off the midpoints than to put faith in measurements off thee assumed true side of a broad. I wish school did more with geometric construction than a Sidenote in algebra.

Yes you're right. Using compass is great in that you have perfect midpoints, also good for woofer baffle layout to position the outside perfectly etc etc.

It comes as second nature to me to do things that way because I trained as a draughtsman the old fashioned way - long before Auto-Cad came on the scene.

[Grant Bunter]

I find this quite interesting, basically because:
I made assumptions: my square(s) are in fact square, and my straight edge/ruler was straight, and accurate.
Doing so allowed me to work with constants. Even if they were flawed constants, they would be constants for everything I built.

History and education aside (in your case Harley), I wondered how much difference it would make?

I looked up straightening your square etc and noted you could calculate the error in the legs by using feeler gauges. It seems errors of 5 thou to 30 thou were not uncommon as a general rule.

Then I calculated a 1 degree angle "error" (compared to a straight line) over 8 feet (because that's a sheet length in ply) and came up with a 0.14" discrepancy over the 8', or in metric 3.55mm.
3.2mm is 1/8th inch, or the tolerance described by Bill (but we don't have many plans requiring 8' panels), so a 1 degree error over 8 feet is "a tad" over 1/8th of an inch.

What does that 1 degree error equate to in thou?
Turns out to be between 139 and 140 thou. Which is a lot more than the error in most squares.

Sure, ply isn't always square, but one corner is usually close enough. Since most panels are no where near 8', the margin of error is reduced.

I spent less time drawing, and more time building...

[zefrenm]

..... more than the error in most squares.

Sure, ply isn't always square, but one corner is usually close enough. Since most panels are no where near 8', the margin of error is reduced.

I spent less time drawing, and more time building...

wouldn't one degree of error start adding up with every cut all 3 planes after every cut?

[Grant Bunter]

..... more than the error in most squares.

Sure, ply isn't always square, but one corner is usually close enough. Since most panels are no where near 8', the margin of error is reduced.

I spent less time drawing, and more time building...

wouldn't one degree of error start adding up with every cut all 3 planes after every cut?

Sure zefrnm,
It would.

A 30 thou error in your square is nearing 1/5 of the 1 degree I was describing earlier, so over an 8' length of timber, 3.55mm becomes 0.7mm or near enough.
For a T48 side panel at 48", that becomes 0.35mm.
Statistically it's a non issue and a mere 10 percent of the allowance Bill makes for error.
I don't think you could look at a cab and say it has that amount of error.

You could introduce the same amount of error by drawing everything up with a ruler and marking just to either side of a mm mark, or not have your saw running exactly true.

You have to take all the factors into account. I reckon that if you draw up a panel, and everything goes seemingly perfect, and you had errors in all measures, once cut, of less than 1mm, well I would be quite happy with that, rather than thinking I needed to cut a new panel again...

[LelandCrooks]

That's why I use a pretty expensive set of aluminum squares that are adjustable. I check them every couple of weeks, using my table saw table. Temperature throws them out. And practically every panel of BB I buy is out of square. Sometimes not enough to bother, but sometimes they're pretty bad. I use the sliding table on my jet to square up the panels, but I have to rip up the sheet first, it won't make that big of a cut. Lots of waste when I have to do that.

You can do the same without a sliding table. Nail a guide board to your cut piece that's dead flush with the edge closest to you, sticking out past the other edge of the 90. Use it to guide through the saw with your fence. Now the opposite side and the edge closest to you are a true 90, if your fence is right.