Well machineing shims to a tolerance of 1 micron is no great feat these day's. so if it came to the nitty gritty,it could be done-until you get a good temp change in the system-then all bets are off.

I dont think a magnetic field is going to change much over 1 microm though,so near enough would be good enough lol.
When it comes to perminant magnet's,nothing would supprise me.
I have just built a coil that outputs DC only,regardles of the input.
By useing PM's in the right configuration,we get only a DC output,without rectifiers.

Fancy that- PM's as rectifiers.
Now thats pretty useful.

Sounds very interesting, could you give details 

Hey, TS... can't you tell the difference between an INCH and a MILLIMETER? I have micrometer calipers that measure to the tenthousandth of an INCH myself and I know how to use them. A ten-thousandth of a MILLIMETER is quite a bit smaller than that, since there are 25.4 millimeters to the inch.

"No exact significance was placed on the figure by anyone but you, TK."

Er... nope, someone put that exact figure into the PDF and, by implication, asked me to believe it. But I can't, without some evidence that the person citing the value can actually make measurements of magnet position to that precision and accuracy. Just a measurement thing... well, I can recall a  certain very expensive spacecraft that missed its target, Mars, completely because of a "measurement thing". Measurements and accuracy are important, and false precision is just that: false.

And I learn something every day. Don't you? Good grief, what if Overunity actually needs 22.4283 mm spacing instead of 22.4282 mm?

For the sake of the argument  lets just say we can vary the field strength on either or both magnets with a wire coil and a little juice and that will bring the desired fields into play.  This variance is already done in actual circuits so let's just conceed it and move on to the next situation.

I finally got around to writing a little program to calculate graph areas,  and used it to recalc. the areas of the input and output
strokes from the measurement sets.  The value of the "total gain" calc. came out as 19.193 % of the input as opposed to the
original figure of 19.131%

(The original area calcs. were arrived at by hand counting and estimating tiny graph squares/fragments  duh !).

Over all 'm, I'm fairly well pleased to find the accuracy of my original area estimations were so close,  (only 0.062% off ! ).

Please find the 2 attached files "SLBIG.PDF" and "ROBIG.PDF."  They contain  outputs from the graphing program I wrote.
The "beta program" doesn't do the graphics yet, (only the calcs at this point).

I will be republishing  the latest versions of the  "Work from 2 magnets"  files when they are completed.  (soon ? )


                        Cheers
                            floor

Was still learning measurements methods at this point.
  floor