Building a Watercolor Painting

I've taken up art.  I've decided I'm gonna back off on building/cutting/modifying objects for a while, except for the LCD monitor I'm going to attempt fixing in my next post. So armed with some supplies from the local art stores and a 1978 edition of The Watercolor Painting Book from the local library, I decided to do my own rendition of "Demonstration 10.  Winter Landscape".

That's it.  I call it "Early Snow" since I used way too much green in there.

Here's my Reeves paints, which I got in a set obviously.  The individual tubes are marked for color fastness and one or two are not excellent, just good.  A fact I will have to remember if I ever embark on a paint I anticipate keeping.  And still, even with 18 tubes of paint, all the instructions call for colors I don't actually have.  So I make it up.

Here's the paper.  Commonly available Strathmore, cut into quarter sheets to conserve badness.  There's enough bad art in the world and I don't want to create too much at a time.

Here's a nice shot of my brushes.  Nice because in focus.  I use them indiscriminately.  I think I need one called a "rigger", though, for fine lines (originally rigging on ships, so I hear).

And finally during cleanup, my palette.  The colors were pretty in the water, but don't show as well in the photo - nothing does, because my camera sucks.  I cleaned the palette out completely because I'd crapped up all the little color blobs so thoroughly that not one of them was uncontaminated with barf gray.

Building art is fun.  I'm going to do it again.

Third Attempt at a Homemade DIY Table Tennis Racket

Why I Need to Build Another Racket
My first racket, when combined with LKT Rapid Speed 2.0 rubber on both sides, weighs 220 grams.  That's 130 for the 7 ply fiberglass reinforced blade, plus 90 for the rubber and rubber cement used to glue it on.  I played with it for an hour and it's too heavy.  Okay for predictably looping the forehand, pleasant to smash with, but horrible for backhands, where the wrist is more involved.  And because I didn't build up the handle enough, I always had the impression it might slip out of my hand and kill someone.

Who knew building a ping pong paddle could be so hard?  I felt I had to do a few things differently:

1. Lay off the fiberglass.  The fabric I have is too thick, I think, and it brings in way too much plastic resin.  It's a theory.
2. Use clamps instead of weight to form the laminate.  I need to drive more of the glue out, thin out the glue lines.  Glue provides little strength on its own.
3. Less layers and a bigger, monolithic core.

I also rough-cut a piece of birch plywood to shape, to see how the weight came in.  

Left to right:  original 3 ply butt joined blade proof of concept, 80g; finished 7 ply blade, 130g; 2x LKT Rapid Speed with leftover rubber cement, 90g; rough-cut birch blade, 104g

Not good enough.  So I began to calculate.  The idea was to calculate the weight of a cut out blade based on a ratio from the 6"x10" whole board. That way, I could measure the boards going into the paddle and determine how it would turn out in advance.  I used the birch plywood and the remainder pieces as my sample, and calculated a bunch of stuff from it.

These calculations are all crap

The problem is that the birch plywood I started with was not cut to the final 10" length, but 12".  So my golden ratio of 0.57 was relatively useless.  And I didn't feel like doing it again.  I wanted action, not words!  I still think this is a good idea, though.  I really should do it.

So my next attempt was a 5 ply:  3/16" balsa core, two transverse 1/32" basswood plies, and finally two longitudinal 1/32" basswood plies.  All this wood came from National Balsa, which has been my choice of suppliers so far.  I didn't even pay for the balsa.  It came as part of the packing, to prevent my 6" wide 1/32" thick basswood sheets from breaking.  Sweet.

Here We Go
Per my new plan, this time I clamped rather than weighted.

Those are 1/2" granite tiles on an old desk, with a length of oak on top to keep the metal off the stone.  The glue is two part epoxy left over from my guitar project.  Next I needed a method to measure mass, or Earth weight.

This scale is a 500g model from Harbor Freight for ~$13.  It works consistently.

Once the piece cured and I cut it to shape, minus a couple millimeters per side to cut the weight further and make it easy to fit my old rubber sheets, it was about 87 grams.  Too much, so I hollowed the handle.

Blade with hollowed handle section.

I do all this work with a bandsaw, files, a Stanley SurForm cheese crater, and 100 grit sandpaper.  While I have a router, I do not have a router table.  That's why it looks totally unprofessional - I don't make templates and route around them.

Handles ready to be applied.

I decided I didn't need the end to be closed.  I broke it out completely.  The handles are more 3/16" balsa plus a layer of 1/16" basswood on top.  The completed blade is a good weight, IMO.

Completed blade, 91.2 grams

LKT Rapid Speed rubber reapplied and trimmed.

Assembled racket, 175.4 grams

Not bad!  Nearly 45 grams lighter than that beast I built the first time.

Playing Impressions
The completed paddle is in line with "real" paddles I've played with in the past.  Again, the handle isn't thick enough, thus the hockey tape.  I can flip the backhand much better than the 220gr monster, and it's only 5gr heavier than my friends preassembled Stiga Cannon.  It is of course much spinnier and hard to handle.

I am not a rater player, I haven't even come out of the basement yet, so I don't know how this blade and combination rates.  ALL?  OFF?  Who knows.

Now I have to get used to playing "real" rubber again.  The old monster blade, at 130gr naked, will be converted into a sandpaper paddle.

Relative Paddle Weights

• "Ping Pong, the Original" 4 pack of preassembled paddles: 135g - 141g
• Stiga Cannon preassembled paddle: 171g
• My second blade, the "beast": 130g
• My third attempt, hollow handle:  91g

Second Homemade Table Tennis Racket Complete

I put a couple coats of wipe on poly to seal the blade and prevent the handle from soaking up skin grease too fast.  I'm not a genius at table tennis but I have a feeling this would be classed as a very, very rigid blade.  Could be wrong.  OFF, certainly, I'd bet.  With 7 plies, two of them bias plies and two of them glass, and epoxy for the adhesive, it's definitely a beast.

I followed some instructions on the intertubes to apply the rubber to the blade using rubber cement.  I got some thinner for the cement, and I probably should have used it, but working quickly with the un-thinned cement went okay.  Apply one coat to the blade, two to the rubber sheets, let dry to tacky, apply, roll, put under a big book, trim.

It was the trimming that went horribly.  I used a fresh razor blade and it kept catching in the foam and the rubber.  I ended up using a pair of shears, going an eight inch at a time - and still, it's crap.  Functional, yes!  But ragged and crap looking.  I supposed I might get better with practice.

See?  Also this thing weighs about 10,000 pounds.  I figure the blade started out heavy because I chose maple for the core.

I'll let you know how it plays.  I expect to bring a lot more spin to our little table at work than anybody else, and I expect a lot of people to be irritated by it, including me.  I plan on visiting a TT club an hour away sometime in the next month to play some folks who are above my level, and maybe get their opinions on my racket.

Second Attempt at Homemade Table Tennis Blade

This time I'm going to do 7 plies, 5 of them wood, all glued with 2 part epoxy:

1. Center ply of 1/16" hard maple, lengthwise grain
2. Two plies of 1/32" basswood, 45 degree grain (from a 12" wide sheet)
3. Two plies of 8.9oz S Glass Satin Weave, Thickness: 0.0097"
4. Two plies of 1/32" basswood, lengthwise grain

This picture shows the layers as they will be assembled:

I assembled them painstakingly between two pieces of super-flat polished granite from a big box home improvement store, each covered with sections of black garbage bag to protect them for another project.  Weight added to mash everything together.

I don't have an accurate scale but it feels kind of heavy so I'll skip the oak handles and use more basswood instead:

Use one of these when you sand.  It cost about $40.  The filter packs are replaceable.  It handles particulates, some VOCs, some acid vapors, and other things I cannot pronounce.  It works infinitely better than those silly fabric masks you buy by the dozen because it is comfortable, seals perfectly, doesn't get all soaked with moisture, and doesn't fog your glasses if you have them.   That last point is worth at least $40 to me.  

The handle halves were laminated separately on the same granite slabs using regular glue, then bonded to the blade handle with regular glue.  I'm tired of epoxy at this point - shit sticks to everything and won't ever come off, forces you to throw tons of gloves and plastic and brushes straight into the landfill.

I shaped the handle using a rat tail file and 100 grit sandpaper.   I finished it using 220 grit sandpaper. Careful when you use a coarse file, or even course paper - you can shred the grain.  The laminations help with that, though.

What did I learn?  

1. Use a half-template:  one half of a paddle shape, flipped on its axis.  That way your thing is symmetrical
2. Just buy one next time
3. I'm tired of chemicals.  Every time I do a project I'm up to my ass in chemicals.

Building a Table Tennis Racket

We got a table tennis table at work and it revived my old enthusiasm for the sport.  I played in the 90's and I had a semi-custom paddle and everything.  I wasn't very good but I enjoyed the hell outta it.

So naturally, given that I have a bunch of raw materials left over from building my composite guitar neck, I decided to embark on a journey:  making my own wood/composite table tennis blade.  Naturally this is going to turn out more expensive than simply having bought one from a store.  But I guess that's not why I do this stuff.

First, a picture of a sandwich discovered on my little blue camera chip thing.  I got a message from my HP 735 3.2MP 15x Photosmart camera that I'm running low on solid state storage.

I don't want to lose this.  It's a metaphor for composite laminates:

Next, I assembled materials.  The black plastic sheeting is part of a garbage bag, laid on top of an old mirror which I hope will provide flatness.  The basswood sheets I have are only 4" wide and so I'm going to have to butt them against each other in a sort of alternating layer thing.  The book is to push down on top.  Good book, BTW, if you like solving "Mate in X" puzzles...lots and lots of them:

Next I put some wood down on the plastic, then a layer of fiberglass fabric on the wood:

Then I had a sandwich containing three layers of wood plus two of fiberglass, soaked in resin.  Yum:

I added weight:

I let it cure overnight and prepared to cut the shape of the blade:

And I ended up with a slightly-too-flexible paddle that has epoxy puddles in the gaps where the butt-joined layers meet:

Not good enough.  It's got enough rigidity in the long direction, I believe, but it flexes a bit too much across.  I think that's because there are two long joints in the middle.  I should have done two layers joined at the edges and only one joined in the middle, but I did it other way around.

I've ordered more wood and when it comes we'll use this one as a template.

The Meaning of Phrygian #6 - What I Missed

In my last post I went through convulsions to derive the so-called Phrygian #6 from the major scale by means of an involved process.  I left for later the question of what Phrygian #6 actually means.  


I turns out you have to do is get the Phrygian mode and do what it says - sharp the 6th degree.


1. C Major scale with # of semitones between each degree:

       C   D   E   F   G   A   B   C
         2   2   1   2   2   2   1

2.  The 3rd mode of the major scale is the Phrygian:

       E   F   G   A   B   C   D   E 
         1   2   2   2   1   2   2


3. Sharp the 6th degree to "slide" the one-semitone interval over one degree:


       E   F   G   A   B   C#  D   E 
         1   2   2   2   2   1   2


4. Transpose to D:


       D   Eb  F   G   A   B   C   D
         1   2   2   2   2   1   2


Having said that, I don't really like Phrygian #6.  I wanted the regular Phrygian after all, because it sounds like it goes with my riffs better:


       D   Eb  F   G   A   Bb  C   D
         1   2   2   2   1   2   2

Building a D Phrygian #6 Scale

Yesterday I played around with a couple of power chord riffs in standard drop D that have root notes at frets 0, 1, 3, 5 and 7 (D, Eb, F, G and A).  I got to thinking, what is that scale?  So I went to JGuitar.com, which totally rocks, and set up my tuning to Drop D, then set the root note of the Scale Calculator to D and started calculating different scales until I found one that fit:  D Phrygian #6!

Of course, how could I have forgotten?  Good old Phrygian Sharp Six.  I wanted to understand where that name comes from, so I searched for it and found this at Marc Sabatella's site.  Mr. Sabatella writes:

There is no common term for the second mode of the melodic minor scale. 

It doesn't tell me where the name comes from, but it is what I need to derive this thing from first principles without twisting in the wind for hours:

1.  C Major scale with # of semitones between each degree:

       C   D   E   F   G   A   B   C
         2   2   1   2   2   2   1

2.  The 6th mode (Aeolian mode) of the major scale is the natural minor:

       A   B   C   D   E   F   G   A
         2   1   2   2   1   2   2

3.  To make the harmonic minor out of the natural minor, raise the 7th scale degree one semitone.  The classical composers did this to be difficult make the 7th degree of the scale a proper lead tone, half a step away from the root.  So much for the mathematical precision of the "music of the spheres".  They were like, "Yeah that's a pretty good mode but you know what?  Doesn't sound right for my song here. Let's just change it."

       A   B   C   D   E   F   G#  A
         2   1   2   2   1   3   1

4.  OMG!  See that 3 semitone interval in there?  Ugh!  It sounds "unnatural"!  To remedy that, they make the harmonic minor into the melodic minor by means of another little tweak:  raising the 6th a semitone to "take up" the allegedly awkward-sounding 3 semitone leap between the 6th and 7th degrees.

        A   B   C   D   E   F#  G#  A
          2   1   2   2   2   2   1


Another thing:  that's the birth of a scale that contains two 1 semitone intervals that are only a single 2 semitone interval apart.  Scales built from natural or harmonic minor scales don't have that pattern.  Scales built from modes of the melodic minor scale do have that pattern.  Useful side effect of all that goofing off with the minor scale.  So next time you have a partial scale that goes "half-whole-half" and you start searching for a name, and you find yourself in a sea of Altered Locrians and Phrygian #6 goofiness, it might help to remember that you're simply perhaps in some mode of the melodic minor.
And finally:  in the classical approach, you only play that when you ascend.  You play the natural minor when you descend.  Dicks.
Oh wait one more thing:  When you read about minor scales to find chord progressions that go with them, you are given a whole pile of options.  Where do they come from?  It's all the chords that are diatonic to both the natural and melodic (and sometimes the harmonic) minors, chucked into a bucket.

5.  The 2nd mode of this one-way scale is sometimes called the Phrygian #6 for some reason - that's for another time:

       B   C   D   E   F#  G#  A   B
         1   2   2   2   2   1   2 

6.  Transposed to my chosen root of D, because that's the lowest, coolest note of a guitar in drop D:

       D   Eb  F   G   A   B   C   D
         1   2   2   2   2   1   2 


Note: It was helpful to bring in the flats instead of the sharps, to keep the accidental count as low as possible.  Since the riffs I'm playing only contain the first 5 notes, the question of whether I want to have A go to B or A go to Bb is up to me, and I haven't decided on that yet.  That will be a totally different scale!

Straight Pull Headstock Conversion, "Finished" but Unfinished

I still have to adjust the nut slots and tweak the bridge and saddle, but this guitar is now playable in its new 3+3 straight pull configuration.  We'll see about stability and durability.  Here's a detail of the headstock with its battle scars.  The new bone nut is in place, and the Sperzel tuners.

Here's the whole guitar:

Acoustic Guitar Straight Pull Conversion

Straight pull means that the strings, when viewed from directly above, come over the guitar nut and drop to the tuner shafts without an divergence to the left or right.  Fender electrics often have this arrangement, but it's far less common on 3+3 style headstocks.  And I'll show you why.

For what purpose do I wish to do this?  I've seen differing opinions, some of them strong.  And the person in that link has a point.  In any case, when the guitar nut is viewed as a dimensionless line, the string is only going in one direction:  either it goes over the nut and straight "down" to the tuner, or it goes over the nut and "left/down" or "right/down" to the tuner.  That's a fact.  But to my mind, that misses a couple of points:

1. Straight pull just seems like it ought to be better, facts be damned
2. I like how it looks
3. The argument fails to take into account the actual geometry of the guitar nut, which is a vertical slot cut into the nut material.  A string that goes straight "down" lays naturally in this slot.  A string that goes left or right as well angles down into the slot from the front edge of the nut, then makes another angle from the back of the nut slot to the tuner.

Anyhow I'm doing it.  If I screw this all up I'll simply saw off the headstock at the scarf joint and glue on another piece, or something.

First job is to determine the natural lie of the strings.  I haven't been able to figure out a good way of doing this, really.  I figure close is good enough for this cheap acoustic.

No, wait.  The first job is to plug up the four holes that will be moved.  I used 3/8" oak dowel glued in with that same hardcore two part epoxy I used to do the fiberglass layers on my electric.

Then I put tape across the last inch of the end of the headstock, put the bridge pins in place, made a loop in a piece of string, hung it around each of the bridge pins in turn and ran it through the correct slot in the nut, and eyeballed a line on the tip of the headstock.  When the lines were pretty close to evenly spaced, I figured they were pretty close to correct (proportional string spacing or not).

Then I used a little piece of brass tubing from the torch tip cleaner kit that was almost the same diameter as the wrapping portion of the tuner shafts to mark the desired locations of the new holes.  This is hard because each of these 3/8" holes will interrupt valuable grain in the headstock material, inevitably weakening it.  So I staggered the hole locations.  

Drilling simple holes when you don't have a drill press can be an adventure as well.  Drill bits wander from starter holes; verticality becomes an issue.  I solved the wandering problem by making a template from a piece of steel stock:

Yes, really.  Then I measured from the holes to the desired edge of the headstock based on providing enough clearance for the tuner buttons to rotate.  I then made a silly flowing profile and attacked it with a coping saw, because a tilt-back guitar headstock is not easily amenable to cutting on a bandsaw.  Nice filled holes, huh?  This guitar is looking better and better every second.  

There are a million little holes on the back of the headstock where the old tuner mounting screws were removed, and where I misplaced the new tuner locator pin holes a million times.  Since strength is an issue here, I'm going to fill all the small holes with raw epoxy puddles.  I guess that should be strong, unless it's not.  I have the option of finishing the edge and face of the headstock with veneer to cover  up all the mess, but in order to do that I'd have to take away native grain and that concerns me.  I think this is just going to be an ugly guitar.

Proportionately Cutting a Guitar Nut, Cheap and Naive

I ordered a 1/4" blank bone nut from StewMac and decided to make my own.  I'd never done that before.  I cut it to length with a hacksaw and to shape with 100 grit laid on 1/4" plate glass.  I've read that 3/16" of bearing area is sufficient.

I cut it roughly to height with the hacksaw as well.  Since I knew that the original nut was a bit high, I used the one to draw a line on the other, placed the new nut in the jaws of the vise with just the discard area sticking up, and chopped it flush.

I used a tiny hobby triangle file to start the slots for strings 1 and 6.  Then I was faced with the problem of spacing the rest of the strings.  This was fun.  I wanted to space the strings proportionately, which means that instead of being spaced evenly on center, they will be spaced evenly based on the distance between their edges.  First, I needed a jig:

I obviously made that, and it was obviously cheap.  I put six nuts on a small threaded rod and used them to enforce my desired string spacing.  This works best, I think, on a traditional 3+3 setup lacking straight string pull. - the horizontal break angles pull the string into the adjustment nuts.

Then I calculated thusly:

1. 1 and 29/64" between the high and the low strings =  1.453125
2. The total diameter of strings 2 through 4 = 0.114
3. The total space to be divided among the remaining five string gaps = 1.339125
4. The amount of space between each string = 0.267825
5. Multiplied by 64 to make it 64ths = 17.1408

I thus attempted to put 17/64" between every string - that means diameter to diameter or what have you.  That way thicker strings get their share and fingers get the same amount of space around each string.  Is the better?  Who can tell?  Measuring was hard with my eyes.  Around a year ago I noticed that I have trouble at close range with tiny little 64ths of an inch.  Luckily, I found this weird lens in the junk drawer of my tool chest:

It's badly scratched but it actually worked well for this task.  Once I had the spacing, and believe me that (for me at least) committing to a measurement is a real act of faith, I marked each side of each string with a pencil.  Not an ordinary pencil, but a pencil sharpened on a flat sheet of sandpaper to make a knife edge.  

Then, as I said, I started each slot with a hobby file of small size and triangular cross section.  This is nerve wracking in itself:  I couldn't commit a deep cut until I was sure the groove had landed exactly between the two bracket lines.  When I'd cut deep enough to just touch both lines, and that contact occurred at the same time, I had confirmation that I'd hit the center properly.

Then I finished the slotting with a cheap set of welder's torch tip cleaners, a trick I'd found in several places on the internet.  I am unable to make the exact citation, and I do not know who originally proposed this solution.

They look like that.  I just picked the one that looked next-size-up from my string using my untrusty eyeballs and my fingers (important).  The large ones were able to hold up to the filing pressure, while the small ones need to be supported by your fingers wherever you can hold them.  I settled on one finger from my free hand on the edge of the nut, and one free finger from my filing hand pushing down on the file right where it crossed the nut.  It worked.

My next task will be more outlandish:  straight pull conversion.

Fred Edge Dressing, Cheap

This fretboard is bound by a strip of black plastic.  I could have taken this opportunity to trim the frets to size and snip their tangs to preserve the binding, like you might for a good guitar.  But I chose to extend the fret slots through the binding and do it the easy way.

You can get special tools to allow you to hammer frets over the body of the guitar, but why spend the money on that when you have a large hammer that can also be used for other things?  That big one at the bottom:  you can reach through the soundhole and support the fretboard with it.  It helped that the fret slots were generous in width for the StewMac fret tangs.

I trim the fret ends with a regular end cutter, not a fancy luthier's model.  My whole goal is to avoid buying specialized tools.  When I cut, I push the cutter against the fretboard and then pull it back just a hair, to prevent the fret from getting levered out of its slot by the unsteadiness of my hand as I apply pressure to the cutter handles.  

Then I was left with a row of spikes on both sides of my fretboard.  When I built the neck for the electric, I simply ran a mill file freehand down the sides.  The worst that happened there were a hand cramp, a bloody nick, and the occasional ding on the headstock when I ran the file too far that way.  The acoustic presents one new problem, however:  freehanding it is impossible because the file would repeatedly hit the guitar body where the fretboard goes over it.  

So I made a tool that consists of two clamps holding a mill file to the edge of a board.  The clamps are heavy, so I arranged them in opposite directions.  The file descends only 1/8" below the face of the board, which is run along the fingerboard.  It was heavy as hell but worked perfectly.

The black plastic binding took some damage so I sanded it down with progressively finer grits of paper until I'd taken off the damaged finish material and plastic fuzz.  It's visible but doesn't interfere with function.

Fine Radius Adjustments on a Fender LSR Nut

The action on the low-E string of my LSR nut was a little high.  I think it's because the nut's supposed 9.5" radius doesn't match the fretboard's supposed 10" radius at the nut, or my alleged fret leveling job, which might very well have changed the effective radius of the neck.  Either way, the low string was way high.  I decided to explore a theory.  No pictures because my crappy old camera won't focus on anything so small.

No warranty, expressed or implied.  Proceed at your own risk.  Good luck.

Here's what I did to lower the action on a string.  Seems to have worked for me, except that this operation seems to have wrecked some of the springy retention ability of the retainer prongs.  I will have to be careful when changing strings forever more.  I think if I had removed the nut completely and taken the retainer finger assembly off in one piece, that might not have happened, but imagine what a pain in the ass that would have been!

1. Place the guitar on a nice smooth surface covered neatly in cloth.  The little ball bearings, once lost, will never be found again if they bounce on something hard.
2. Loosen the string and pull it aside into a neighboring string slot.
3. Pry the retainer prongs back toward the headstock a little.
4. Use a slightly-magnetic awl tip to draw out the two ball bearings.  Grab them between your pinched fingers before you lift them very high, and put them in a container.
5. Use a drill bit slightly smaller than the ball bearing cups to deepen the holes that hold the bearings.  Measure forty five times and cut twenty times.  Go a little at a time.  You can't go back.  How do you know where to put the holes?  I dunno.  I just leaned the bit into the corners ever so slightly and drilled.
6. Replace the bearings, replace the string, and tune to pitch.  If you are convinced you need to go deeper, do so.
7. Lather, rinse, repeat.

Here's what I did to widen the string slots on the LSR nut.  I'm using a set of very heavy strings (12's) and I've been worried that the low string was hitting the edges of the little exit slot.

1. Steps 1-4 above
2. Use the awl to pop out the little foam dampener on the headstock side of the nut, and place it with the ball bearings.
3. Sand away with a tiny rectangular file.  Don't wreck the retainer prongs.

Your mileage may vary, especially considering that I'm just making all of this up.

Acoustic Refret

First I had to get the fingerboard all flat. I've struggled with flatness before.  This time I think I did better. Below you see my cheap aluminum level with a layer of fabric carpet tape plus a layer of 220 grit paper.

1. I use a solvent to clean the old tape scum off the level
2. I put on the carpet tape.  It sticks like a beast so hold it at one end, line it up an inch off the opposite end, and lay it down easy.
3. I use a razor blade along the edge to cut the tape flush, then peel the backing.
4. I put on the paper in 11.5" strips, then use the razor the same way to cut that flush

I made sure the truss rod was loose all the way, then I sanded.  There was a high spot at the soundhole end of the fingerboard, and another longer one from fret 1.5 to about fret 11.  By the time I was done, all the fake black ebony stain had come off, and I'd taken enough wood off that the whole idea of putting on high frets to lower the action was nullified.  Way to think it through, kiddo.

Also excellent:  the nut, which is supposed to pop off easy, did in fact pop off easy, but it took some grain with it.  I'll have to flatten that surface with a file.

Here is the first several frets.  I didn't order enough fret wire because I messed up, and the next batch didn't arrive in time for the hurricane.

Acoustic Refret, Nut and Bridge Shave

I've got an 80's Harmony acoustic, given to me by my brother.  It has a terrible high action.  The neck doesn't seem messed up, and the soundboard isn't too bellied out.  I don't want to reset the neck because it's a bit of a big step that requires tools that would cost real money:  steam and all kinds of things.  I know I could slap something together from junk, but in my experience junk (and connecting junk to junk) costs real money.

So for my next project, I'm going to try to lower the action on this guitar - backwards:

• Put gigantor frets on it
• Redo the nut to compensate
• Shave the bridge
• Perhaps convert the bridge from an adjustable-height bridge, which seems to take up precious verticality, to a fixed height
• Put in a brass plate below the bridge plate to prevent the string ball ends from tearing up the soft wood down there.  StewMac wants $12 for this but I can make one for more like 20 cents.

Here's the old adjustable bridge:

 It won't go low enough.  I think I'll have to shave the bridge down too.  That will come after the refret, once I see the action I'm getting.

Here are the strips of feeler gauge I'm using to protect the very soft fretboard against the heel of my fret-removal chisels:

I don't have a fancy fret-removing end cutter that's ground down so the cutting edges are flush with the flat face. These chisels work beautifully.  I do guess that there's a slight outside chance of a truly horrifying accident, if one of the chisels comes loose, rides up the other chisel and unzips the flesh of my forearm.  That would suck.  But I didn't get the feeling that was likely to happen.  There's not a lot of force involved in this; the trick is getting a sharp edge under the metal/wood interface.

I'll be ordering fret wire and replacement blank nut and and bridge this week.

The Armrest is Finished

Here's the completed guitar.  It was done a while ago but I didn't post for a long time.  This photo was taken at work by a coworker.

You can see how the upper rest has been bent almost parallel to the body.  I did that with a bench vise and a bad attidude. 

What else did I learn?

1. More composites in the neck, or none at all.  The neck is strong but I'm not sure it's much better/different than a wood neck.
2. Keep knobs farther away from the bridge.  Sometimes during extended palm-muted riffing sessions, I find I've nudged the bridge volume down with the back edge of my hand.
3. I've always hated the G string - it sounds nasty on every guitar I've ever owned.  But on this guitar it doesn't, because somehow the composite neck is eating those frequencies and damping them down.  It's so pronounced that I've noticed the G string has less sustain than the others.  This is kind of cool but I'm not sure it's all good.  Some fret positions on the G string die out quicker than I'd like them to.
4. The balance of the guitar is excellent with respect to the leg rest.  I record with this guitar exclusively and whether I use the strap or not it stays where I put it when I'm sitting down.
5. The arm rest is still a lot taller with respect to the bridge than most other guitars.  But it works well.
6. The EMG 81 TWX and 89 XR pickups sound very nice.  It's a pretty non-resonant, clinical sounding guitar and the pickups convey that accurately.  I've never played any other EMGs so I can't tell you about the X thing.  The guitar sound itself sounds sweet enough in neck single coil, fat and nasty in neck humbucker mode, brutal and cutting in bridge humbucker mode, and thin as all hell in bridge single coil mode.  The neck single coil mode evokes enough of that strat neck thing that I'm very satisfied with it.  The middle position doesn't do much for me; mixing the two pickups in single coil mode doesn't evoke any of the sound of mixed neck/middle on my G&L S-500.  So...
7. Based on all that, next time I'd skip the dual mode pickups.  The dual mode pickup in the neck position is probably more than I need, as well. I should have done single 80X/SAX/SAX, I think. That's just me.  I love the strat single coil sound in the neck/middle positions, and I love humbuckers at the bridge, and that's all there is to it.
8. Battery life is nuts.  I've had two batteries in there in the past 6-7 months, and I only changed them to see if a new one sounded different - it didn't. 
9. I'm glad I made it funny looking.  There are quite enough strats, teles, pauls and prs's in the world IMHO.
   

Even Dunner (or...How I Bent My Rod with a Hickey)

So my friend Tom lent me his 1/2" Hickey EMT bender.  I attacked the 1/2" aluminum rod with it.  It was not pretty but here's what I ended up with:

That's one single piece of aluminum.  If I can, I'll bend the armrest in toward the body more.  But it's not happening tonight.  How's it work?  The leg rest is awesome.  The armrest is a bit tall and it doesn't seem aesthetically pleasing, exactly.  But hey.  What're you gonna do?  That thing's permanent.

Intonation - the 12th Fret Harmonic vs. the Open String

The question I tried to answer today is, "Which is more better?"

• Comparing the note at the 12th fret with the harmonic at the 12th fret, or
• Comparing the note at the 12th fret with the open string

Today I set the intonation on my new guitar for the Xth time.  I have a Peterson Strobostomp, which is good because neither my ear, nor a needle tuner, has ever helped me intonate a guitar.  And I couldn't tell see much difference between the two methods.  I think the reason is that if the intonation is already close, the harmonic and the open string are converging somewhat.  Who knows?  It's such an inexact thing.  My favorite E-form barre chord sometimes sounds good to me, and sometimes sounds horrible, when played up the neck.  That can change from day to day.

I tried hacking the G string compensation a little by increasing it arbitrarily, because it seems like the G is the primary culprit in spoiling my perfect chord.  I've heard of studio musicians doing that when they know exactly what intervals, and where on the neck, they're planning to be.  I think experienced luthiers, guitar techs and musicians consciously or unconsciously "sweeten" their intonation, and they're not telling me how it's done.  I've read that experience musicians of the fretless persuasion (eg violin) adjust their intonation dynamically, like magicians.  

I'm trying to work up a system for hacking intonation that works reliably, and when I do, I'll post it.  It'll be something along the lines of:

1. Play your chord or group of notes where it needs to be
2. Strum repeatedly and silence different strings in turn, looking for the one that's ruining it most for everybody
3. Use the regular guitar tuners to nudge that string up or down a little at a time (basically make it out of tune) until it plays better with others
4. If you sharpened it, decrease compensation for that string beyond what is normal, and if you flatted it increase compensation beyond what is normal

Also I just realized that I might be able to do something using just the tuner:  hold the chord up where I need it and, in effect, see how far from equal temperament perfection each note is.  That's basically simple intonation, but maybe it changes the equation when all the strings are fingered?

I am a monkey.

Strange High E String Failure Mode

I've heard it said of tremolo use and Kahlers in particular, that guitar string windings at the ball end should be soldered.  I always wondered what the heck that was about.  From what I have discerned, Kahlers have fixed saddles and wrap the string ball ends around a semi-circular rotating stop, similar to Bigsby tremolos and, in a way, to some wrap-around TOM (Tune-O-Matic) fixed bridges.  And there, I thought, was the crux:  something about turning a corner with the wrapped end perhaps caused problems.

Now maybe I understand.

I lost the ability to maintain tuning on my high E string on my new homemade guitar.  Every time I did a bend, it came back low - and it's a fixed bridge.  It happened over and over and over.  I even loosed the Steinberger gearless tuner at the headstock, unclamped the string, grabbed the free tail and reclamped it with the slack taken up.  It kept happening.  Here were my theories, in descending order of probability:

1. string clamp mechanism slipping - but I didn't see any less tag end than I remembered having after the initial string installation
2. string stretching like all getout - but I couldn't imagine a string stretching that much without breaking
3. tuner shifting in the soft basswood of the tuner hole - but i saw no visual indication of it
4. adjustable saddle moving of its own accord - but I checked the compensation screw head and it was against the rear of the bridge
5. bridge moving - but the other strings were great

Then I measure the wrapping at the ball end where it comes through the rear of the bridge.  I measured 19/64ths, worked the string by bouncing the guitar on it repeatedly.  I measured 16/64ths, but I didn't believe it.  I stretched the string again and it became clear:  the high E string was getting pulled around the ball end and  smooshing the wrap toward the ball end.  In this picture you can see it:

Imagine for a moment that I had a camera made after about 2003, that could take a picture worth a damn.  You would see that the green arrow is approximately where the final wrap used to be, and the yellow arrow is where it is now.  Between the two is kinked-up string, unwrapped wraps that have wriggled out of the wrap zone.

From now on I will solder my strings, I guess.

The Guitar with the Finish (but not the Finished Guitar)

After weeks of applying thin coats of wipe-on poly I got enough layers on there.  The poly is not mirror smooth - that happens later, after the poly has cured (2-4 weeks, and/or until the smell disappears), when it's wet-sanded and buffed.  Who knows if I'll even bother doing that?  In the photos that follow, you can see the ripples in the poly.

The headstock with the Steinberger gearless tuners and Fender LSR nut. I've been slacking the strings, adjusting the neck relief with the truss rod, and re-shimming the nut each time.  What a pain in the sack.

The battery is wrapped in blue electrical tape, both to prevent it from shorting any contacts, and because CVS batteries are Not Very Metal.

Six-bolt neck attachment:

I still haven't figured out how to bend the 1/2" aluminum rod. I'm working on that.  It may involved annealing the metal with a propane torch (until the carbon in a Sharpie line fades out).  I've located a machine shop very near my house, but I haven't given up yet; I've managed to do just about everything by myself so far.