Spindle Sizes and Small Variations in Overhang and Offset

There is an RIAA standard from 1963 (and subsequent later revisions and metrications) which gives the record hole size as 0.286", with a tolerance of +0.001" and -0.002". This means that the hole can be between 0.284 and 0.287 (7.214mm to 7.289mm).

Why should this be? It seems a strange size until you remember that the standard was established in the age of imperial sizes, and the nearest fractional size is 9/32" which is 0.281" (7.14mm). This would have been the nominal size for a spindle, and therefore records were made to fit neatly on a typical spindle, not vice versa. If the record hole had also been specified at 9/32" it wouldn't be an easy fit, because if the hole size is within the range for spindle diameter, it is possible the record could be a push fit or an even an interference fit. It won't necessarily drop on or lift off easily.

So, some clearance is needed. For standard record hole tolerance, the spindle should probably ideally be around 0.282 (+0 and -0.002), 7.11mm to 7.18mm. From a quick internet survey of measured spindles, typical modern spindle sizes are around 7.1 to 7.2mm ie 0.279" to 0.283". In other words 9/32" +/- 0.002", which might be expected. This would give clearances from 0.001" to 0.004", with the occasional tight one.

However, with regard to protractors, these are small amounts. The error in the position of the nulls, if there is some play, is minuscule - variations in distortion in the hundredths of a percent (ie 0.01%)

For a protractor that uses the Dennesen principle, like the Acoustical Systems Smartractor/Unitractor, any variation in the overhang direction of 0.1mm would produce an error in offset angle in the order of 0.08 degrees, which, depending on the direction of the accompanying overhang variation, might  actually reduce average distortion for a Lofgren A IEC alignment, not that it would be noticeable.

Another consideration is that, because there will always be an error in set up,  there is a fifty/fifty chance of an error in offset being compensated by an error in overhang, thereby achieving  alignments that are virtually indistinguishable. Of course, if the errors aggregate  then distortion increases markedly.

However all these numbers are tiny and bear no relation to the variation in accuracy normally achieved in practice (with whichever alignment). Usually the overhang is easiest to set, and both the Mint arc protractor and the Uni-protractor, make much of how accurately their items are made, to within 0.01mm, or whatever.  Perhaps if you have an arm with a micrometer adjustment for overhang you could perhaps work to those tolerances with patience and time. 

However, the restricting factor would still be how accurately the offset can be set, especially on an arm where adjusting the offset also affects overhang. None of the protractor makers can give a figure for the accuracy of their device, only for how accurately the parts are made. This is because it depends on how well you can do it. If you could get within a tenth of a degree you would be an ace. Worrying about hundredths is silly.