|
For anyone wit a strong interest in high precision, I strongly recommend reading Wayne Moore's "Foundations of Mechanical Accuracy". For years you couldn't find this book for under a few hundred dollars (it is often considered the bible of precision workmanship), but recently it came available online as a PDF.
Of course the degree of precision it delves in to is thousands of time the level we have any right to concern ourselves with in this field, but the basic concepts are fertile none the less. Acute awareness of just how much the most seemingly rigid forms can distort under the slightest load, basic principles of checking and referencing, calibrating our tools - you don't need to be able to level a 4' surface plate to ten millionths of an inch, but the more you understand and appreciate how these results are achieved, the more you'll find the same core principles working their way in to your methods. There's of course quite a gap between the methods used in calibrating the highest precision machines and what will bring realizable benefits to a flexible neck of wood and metal, but still the more you understand of the higher standards, the closer you'll find yourself to achieving them.
I got my start when I was a machinist apprentice at 18, and was tasked for several months with lapping surface bearings, and the methods of lapping, referencing, and gauging have stuck with me and strongly influenced the methods I've developed for fret work. A bit overkill perhaps, but when practiced and refined can be employed without any significant cost or time, so why not.
We do use straight beams in our shop, leveled and reference to a good surface plate (certified to 25 millionths of an inch some years back, though no doubt past due for recalibration today). Regarding sandpapers, the 3M gold rolls have proven quite reliable. For several years I used a 3M precision lapping film for final checks, 30 micron on a precision uniform backing. After years of stubbornly insisting on using this for my final reference though, its use has faded since I've finally admitted that I've never found it to reveal any notable inconcistencies left by the 3m 240 stage. For a non-precision, not guaranteed to be uniform sandpaper, this stuff does a pretty darn good job.
Of course much of the precision is due to the methods, where any minor deviations in the sandpaper are cancelled out and to some degree self regulated by how the beams are moved and positioned. It's not something simple to describe, but simply involves good awareness of how precision is affected and achieved. My precision leveling process can often involve as much as 3, 4, sometimes more sequential stages of leveling, crowning, and referencing. May seem like unnecessary redundancy, but with practice can bring notable refinement with very little added time. Grit and pressure are altered through the process, along with very intentional control of position and deflection of the neck, and awareness of inevitable deflection of the beams. Sounds like it may be hard and time consuming, and while it is both of these to learn, once mastered it is no longer either.
How much precision is enough? I view precision in two main categories - long and short range. Short range precision is simpler to quantify. With any two adjacent frets, any discrepancies must be multiplied by 18 at the saddle to allow the same clearance for the string on the following fret when compared to perfectly level. This means if fret 2 is .001" higher than fret 1, or 10 is .001" higher than 9, the saddle must be raised .018 higher to give the same clearance over 2 or 10 when fretted at 1 or 9. To put it in perspective, this means that a half of a thousandth variance would require a bit more than a 1/3 turn on your Strat saddle screws to allow for the same clearance. Not huge, but not quite close enough for me to call "perfect" either. With good tools and methods, it's not that hard to bring your short range tolerances down to within 1/10-1/4 of a thousandth of an inch.
Long range tolerances are a bit harder to quantify, as well as harder to define an ultimate "ideal". Aside from determining a preference for straight or relief, how relief is centered and shaped, or where fall away may begin, the neck is really too flexible to place objective final numbers to the shape. Gravity, the pluck of a string, touch of a hand - these all affect significant distortions on the neck. Shape can be strategically controlled in a relative sense though, with varying relief from treble to bass, or deciding whether relief should be centered at the 6-7 frets vs more focused in the first 7 and flattening out slightly from there up. Lot's of different approaches can bring benefits to different players and different necks, so this one is more a strategy and judgement to be developed by experience.
And finally, back to grits - I can use 80, 120, 180, 240, and occasionally still the 30 micron lapping film. It all depends on where you're starting from. If it's a neck I just fretted I may go straight to 240. If I'm leveling out a big hump in a fret dress or doing a partial bar refret, there's no sense in wasting time bringing it in to the initial ballpark, and I may start with a course file.
Longer answer than anyone was lookng for I know, but once I get started... Fret dressing starts out seeming simple. Work for enough demanding players and uncooperative instruments though, and whole realm of previously unconsidered demands and complexities can be revealed. Then comes the long hard learning curve of refining your methods and understanding before you reach the next level. Once practiced and thoroughly understood though, the simplicity and ease of controlling predictable results can be realized in a much more satisfying way.
_________________
Eschew obfuscation, espouse elucidation.
|
Login
Register
FAQ
