2010-09-26

Big Choices

I've got a piece of wood that may be willing to become a guitar neck, and I believe I have a way to cut it into shape.  Now I have to qualify and quantify that shape.  Right now I have to decide lots of things:

  • Scale length, which is the approximate length from the nut to the bridge.  Fenders are typically 25.5", Gibsons are typically 24.75".  
  • Headstock format/angle.  Fenders generally have a straight headstock where the face of the headstock is simply milled down from the level of the fingerboard to provide the break angle for the strings, while Gibson and most others attach the separate headstock at an angle
  • Nut width, which is the width of the neck at the nut.  Use of a Floyd Rose locking nut, Fender LSR nut, or the various species of compensated nuts will tend to dictate nut width, but a traditional cut nut has no limits. 
  • Bridge width, which is the width of the string spread at the bridge.
  • Fingerboard radius, which is the curvature of the fingerboard.  Fingerboard radius is implicated in the choice of bridge and nut.  If I choose a flat radius (easier to implement and fret by far), I couldn't use most pre-constructed nuts, all of which have a finite and definite radius - there would be string height mismatch.  For instance, Fender LSR nuts have a radius of about 9.5", which works with a 10" radius.  Floyd Rose locking nuts come in limited radii; the originals were 10".  Gibson-style Tune-o-matic bridges generally have a 12" radius.
The taper of the neck, as seen from the top, will be dictated by the string lines connecting the nut to the bridge, leaving something left over at the edges for bending.  The scale length will dictate how far that bridge is placed from the nut, so it figures prominently in the geometry.  The choice of radius may determine which bridge and nut I use, thus indirectly locking in the neck taper.

Scale Length/Fretboard Radius

Headstock Angle
I thought I might angle the headstock back.  Fenders are a little weird to me with their use of string trees to increase the break angle of the higher strings.  There are a couple of ways to implement the angled headstock pattern:
  1. One piece of wood - potentially weaker because the grain breaks going around the corner
  2. Scarf joint - as strong as the joint and the glue used to make it
  3. Finger joint - I thought I'd make one of these, just to be different
However, there are two issues with either the scarf joint or the finger joint, neither of them trivial in my case:  cutting this stuff (as always) accurately, at a very low angle, and the wood consumption issue.  As you can see from the following picture, a 10 degree angle will use up at least 6" of wood.  I can just about spare that, but not a bit more.  I can't begin to comprehend whether cutting a 10 degree angle the wide way across a 1" board is doable with a band saw, or if I would have to use a hacksaw.  Either way, big adventure.


Most headstock angles range from 10-15 degrees or so.  If I want to use an LSR nut, which is intended for use with non-angled Fender headstocks, any kind of angle might cause clearance problems with the back edge of the nut, in particular the rubber dampers out back (I could surely just trim those?).  Floyd nuts don't care - there's a big ugly string tree behind it, pulling the the strings way down low to position them over the clamping surface.

On the other hand, my G&L S-500 Tribute has a non-angled headstock with staggered-height Sperzels and seems to work fine.

Nut Width/Bridge Width
Due to my choice of fingerboard, the nut must match a 10" radius and the bridge must be able to accommodate a 16" radius.  The former rules out very little, but the latter rules out Tune-o-matics among other bridges.  My mind swings wildly between a full-on Floyd trem setup and a simple, traditional Graph-Tech nut with a top-loading fixed bridge.  In between I've conjectured the following:
I don't like string-through-body bridges because I imagine that the sharp 90 degree turn made by the string at the bridge causes tension to get stored on one side or the other, which when released results in tuning problems.  I don't know this, mind you, I just made it up as a theory.  There's a guy I know who mostly plays Floyd-equipped guitars and never, ever uses the bar, and says its because they never, ever go out of tune.  My theory on that one is that the locking nut is responsible for most of that, and the spring tension compensates for changes in the guitar geometry due to expansion, contraction and moisture retention.  But trems are not in my opinion great for sustain, and when you bend a string generally all the other strings ring flat.  I could try a Kahler, but I'm scared of all the machinery - and they're just as expensive as a good Floyd.

Conclusion
So you see, I'm trapped in a nest of quandaries.  The text above is only a bare beginning of a description of what goes through my head when I think about this.  I don't have to hurry, but at some point I do have to choose.

2010-09-23

Thank you thank you thank you

I just cut the remaining gnarly edge off my laminate.  I did it with a little 1/8" wide many-tooth blade that was probably included with the saw - it has teeth like a hacksaw blade.  I did it without any modifications to the unit.


I had to put on a new bottom "tire"- that's the thing that goes around the 12" band wheel - because the existing one was flapping away from the band wheel at full RPM, which was a disaster waiting to happen.  New saws have flat tires of urethane on uncrowned, flat wheels.  This old Craftsman has a rubber tire stretched over a crowned wheel.  I order the tire from Sears Parts.  Getting the tire on was far less difficult than I imagined from reading internet accounts.  The perennial (to people who talk about band saws on the internet) "glue or no glue" question remains unanswered.  I didn't glue my new tire on for two reasons:

  1. The top tire (original, AFAIK) stays on without any glue
  2. The bottom tire may not be properly trimmed or tensioned yet.
Here's the gnarly bit, with the cut side facing right:


I have a new Lenox 3/8" 14TPI Neo-Type blade coming from the internet.  The little 1/8" blade has probably given about all it can to the effort, and the wider blade will "clamp in" to the blade guides far better.

Cost to date:
  • Bandsaw = free
  • 2x rubber replacement tires from sears = $33 with shipping
  • 1x (80 in.) 6 ft. 8 in. x 3/8 in. x .025 x 14TPI Standard, NEO-TYPE, Metal Cutting = $20 with shipping
Total cost so far:   $53 (this step) + $209 (previous steps) = $262 

Oh my.

2010-09-18

Band-Saw-O-Rama

It's a band saw.  It was almost free:  the man who gave it to me had it given to him 15 or more years ago by a man who said, "It's yours, but if I ever need to cut something I'll want to borrow it back."  I suppose this arrangement is transferable, at least to the extent that I don't want to cause problems for the man who lent me the saw.

I will call this arrangement "glending," as in, "I have a band saw that some guy gave me, but maybe he'll want it back someday - I can't give it to you, but I can glend it to you."  How prevalent is this practice?  I bet it's more common with tools than anything else.


There are two 12" diameter wheels on which the blade runs, so it's called a 12" band saw.  This means that you can put about 12" of material to the left of the blade before your work piece runs into the housing.


The blade that's on here is 1/4" wide, made for cutting wood.  After I got the saw into rough adjustment (Craftsman manuals are available online), I gave it a shot on my laminate.  Well, not so good.  It got about an inch in and gave up.  There are several possible reasons for this:

  1. The blade was dull already
  2. The blade was not very dull, but it was not of good quality
  3. The saw was poorly adjusted by me
  4. I fed the work too fast
  5. Plastic needs to be cut much more slowly
I did some internet research and I'm putting most of my bets on 2 (crappy blade) and 5 (special needs of plastic).  Because that's what I've really got here:  a chunk of plastic contaminated with wood.  If I can cut the plastic layers, the wood will go along with it (as long as the operation doesn't involve any liquids that might destroy the wood itself).  This is not just any plastic - it's got glass fibers in it, and the two-part epoxy matrix is pretty horrible in itself.

Band saw speed is specified in the US as FPM (blade feet per minute).  Wood-cutting band saws generally go 2500-3500.  Saws meant for cutting ferrous metals generally go 100-150.  Non-ferrous metals such as aluminum go up from there.  I've heard 700 FPM as an upper limit for plastic.  I'll aim for ferrous metal speeds, as that will make make for a very versatile saw.


Between the motor shaft (at the bottom of the picture) and the drive shaft (the bottom 12" blade wheel) I have to insert some reduction.  The motor is specified to run at 1725 RPM, no load.  The diameter of the pulley on the motor shaft is 2.5".  The diameter of the pulley on the driven shaft is 5".  
  • 1725 * 2.5 / 5 = 862.5 (this is the driven wheel RPM)
  • 862.5 * (12 * 3.1416) = 32515.56 (this is the blade inches per minute - the second part is PI * diameter, which gives circumference of a circle)
  • 32515.56 / 12 = 2709.63 (this is the blade FPM, which agrees with the Craftsman manual)
To get to 100 FPM I need a 27:1 reduction!  There are a couple ways of doing this:
  1. Keep the little pulley on the motor shaft, then go to a big pulley on an intermediate (jack) shaft, then go from a little pulley also on that shaft to a big pulley on the driven shaft.
  2. Find some kind of motor that goes slower.  A gear motor, or a DC treadmill motor, are two ways it's been done.
I'd like the saw to be convertible between 2700 and 100, for maximum versatility.  This is going to require some backwards engineering, where I will stumble in the footsteps of people who've done this before, moving   farther from my guitar again.

I am a monkey.


The Trudginger I Go, the Receedinger the Mountain

I got a free band saw, but I think I have to modify it extensively to cut this wood-composite laminate I've created.  This is a good time to reflect on the meaning of Building Things Backwards, and the process of acquiring "knowledge".  Here I am, just trying to build a guitar.  Yes, I got tricky with materials, but it doesn't affect the overall process.  My process so far:

1. I got involved with composites
     A.  I built a heat box to cure composites
           i)  I had to struggle with foam and reflective bubble wrap
           ii) I had to build a heater out of lights
           iii) I had to test it for temperature range and stability
     B.  I layered the composite
           i) I attempted to use weights to compress the stack slightly
                 a) I discovered the need for a clamping jig
           ii) I discovered the practical need for a temperature controller
2.  I had to cut this laminate
     A. I tried a table saw with a plywood blade
     B. I tried a router
     C. I tried a hacksaw
     D. I tried a band saw
          i) I tried a cheap, stock wood-cutting blade
          ii) I will try a bi-metal metal-cutting blade
          iii) I will try to slow down the band saw speed so as not to overheat the plastic in the laminate
               a) I will explore gearing down with pulleys
                     1. I will diagram the entire machine
                     2. I will discover what arrangement of pulleys will fit in the machine
                     3. I will learn the mathematics of pulleys
                          I) I will create more metal framing to support the new pulleys and jackshaft
                               i) I will explore making the pulley tension adjustable so the machine isn't intractable
                               ii) I will explore making the machine operate in both new and old configurations
               b) I will explore using a DC treadmill motor
                     1. I will search the internet
                     2. I will search tag sales
                          I) I will learn to identify motor size by treadmill model
                          II) I will have to bring home a treadmill
                          III) I will have to dismantle a treadmill, preserving the parts I need

Maybe I'm not saying this well, or maybe I'm saying something too pointlessly obvious.  My point, however, is that every time I hit a wall I have to dig backward in a tangentially-related discipline. Each time I do, I am going back in time to follow the tracks of other people who visualized these problems very clearly decades and centuries before I was born.  "Band saw" to me means more than just a free saw that I have to recondition and tune up:  it now means blueprints and engineering and pulley ratios and keyed shafts and pillow blocks and jackshafts.  "Band saw", on a bad day when you have to cut something very strange, could be construed to mean, "All of engineering and physics."

I am constantly amazed by this process of blockage, regression, elaboration and breakthrough.  I don't know if I'm getting closer to building a guitar, or farther. There is a part of me that is always discouraged at this process; that is also the part of me that would have settled for rediscovering fire after each lightning strike, rather than learning how create fire.

I am a monkey.

2010-09-07

Wow, This Stuff Is Hard to Cut Part Deux

So I had this idea that I might be able to cut this piece of basswood/fiberglass composite with a router and a small bit.  This was a Very Bad Idea.  The board (or more precisely, the epoxy in the board) nearly caught fire. 

I was able to cut the ends off with a hacksaw, so I got a handle that accepts reciprocating saw blades and tried to rip the board edges clean with a metal-cutting blade.  No dice; it will take forever.  Mind you, wouldn't care if I only had to do it once, but I'll probably use the same approach to round off the back of the neck.

Plan C is to get a bandsaw.  Bandsaws come in wood- and metal-cutting varieties.  The former generally run the blade at 1500-3500 feet per minute, while the latter generally run about 150 feet per minute.  Some saws advertised as being for both wood and metal compromise with a speed of around 500 feet per minute.  Which do I need?  Do I care to bet $500 on it ($300 used)?

I'm going to try to find an older wood-cutting bandsaw in disrepair that is good raw material.  I will modify it with some extra pulleys to bring the blade speed down, and make it a two-speed machine.