Solar Compass. Suped Up It Can Even Tell the Time!

Survivalist guides on the web do not show anything like this! 
Some of  their ways to use the sun to find direction seemed a little  unreliable and time consuming too.   Especially in deep winter and high summer when the suns path deviates most, I would not have too much faith in the methods.
I hope this design is much quicker and adds to our knowledge base.

Astronomers watch the sun or stars by setting up the telescope on equatorial mount. 
 They line up the axis of rotation with the axis of rotation of the earth .  Imagine a big stick going through the earth from the north pole to the south pole. 
To line up with that, you  tilt your axis towards the nearest pole at an angle equal to your latitude.  So you end up with an axis tilted along a north south line.  At the north or south pole it is pointed straight up and at the equator it is horisontal on a north south line.
The telescope is attached to this axis
 Then they rotate and elevate the telescope to  point the telescope at the sun or star.  Then they just have to rotate the telescope to watch the sun or star all day or all night.  No more elevation is necessary.
The amount they elevate is called the declination.  For the sun, this amount is know for every day of the year.  
 So here is my thought,  If we make the same angle as the earths axis, and we make an elevation the same as the declination of the sun,  keep those 2 things static, and adjust the other parameters till it points at the sun, you should have the axis angle lining up north south.
So you have a compass!

Well, if equatorial mount is good enough for telescopes, then it is good enough for me!
Right?     Not exactly!
I made the device shown.  It was pretty hard to make.  I had to drill the hole at 48 degrees.  Thats about the limit that my table drill will do. Then I had to find a shaft that would exactly fit the hole,  Then another hole in the top wood to take the shaft.  Exactly vertical.
Then 2 exactly vertical holes in the sides of the top piece to join 2 exactly vertical holes in the mount for the magnifying glass!
Everything had to fit tightly with no wobbles and all angles had to be measured accurately.
 Its too hard!  
I tried it out on a sunny break one morning and it was only accurate to about 10 degrees. 
But it was much quicker taking a reading than I expected.
I just couldn't measure declination accurately enough and the wobbles were nasty too.
The size of my table drill limited me on how long the piece for declination was.
So If I was to do it accurately, the drill holes  would have to go!

It took about 3 days to think it simple.  I had to lose the rotating parts because I could not make them accurately.  So all I kept was the plane of rotation (which is at right angles to the axis of rotation).  So the angle for the plane, how do you get it?   Just (90 degrees - your latitude).
I opened a wooden box to that angle. 
Now, declination.   I have a little square with angle markings on it and it is about the most accurate large device I have with degrees.  I will try that. 
What else. A spoke to mark the declination!
A clothes peg to keep the spoke at the angle!
How to line the spoke with the sun?  
Paper. I tore off a strip of drywall tape and pushed the spoke through it.
 Now to try it!
 

The only difference is that the declination angle is reversed. So we in northern hemisphere point the spoke down by 23.5 degrees on December 22nd.   People in the southern hemisphere point it up by that amount.  All that means is that you turn the square around. And attach the spoke with its screw end at the pivot point instead. 
I made this yesterday, January 2nd 2011 and tried it once.   It seems to work. 
I am not planning to do much with it in the future. 
So please feel free to share your tips and 
your lost experiences here.
Brian
PS  This can  give you 2 possible directions for north-south.  One will stay in the same place (it is real) and if you wait 15 minutes the other one will move a bit.  
There are probably simple rules to know which is which.  I am too lazy to figure them out.

On the top of the box, you can mark out the hours  (15 degrees per hour)
See the attached image   (It is for the northern hemisphere)
Set it up pointing north, push your square against the pin,  find the sun,  and read off the time!
A solar clock gives local time.   
That means that if you have a watch set to your timezone, you can even figure out your longitude with the thing.
All, in all, pretty useful. (At least as a mental exercise)

We need an easier to read declination map or table for the year.
Probably a table  is better.   (So you can read off your  declination for that day for every day of the year.)
 I think accurate to the nearest degree is fine.
You could make an ornate version. 
Personalize the thing!
Anyway, that is pretty much it for me.
Hope some people like it.
http://www.wsanford.com/~wsanford/exo/sundials/DEC_Sun.html has declination tables
for the year. 
As does http://freepages.pavilion.net/users/aghelyar/sundat.htm
Brian