Acoustic steel string guitar

[Peter Wilcox]

Some more progress this last week. I made the back with the usual bracing, but flat transversely instead of domed. Since I'll be taking the back off and on, I wanted to make it as simple as possible. It will be screwed on until I settle on a final configuration for the struts. Of course the top is also flat, since it has no bracing to hold it in a domed shape.

[Peter Wilcox]

I shaped and sanded the neck, and fretted it. Still needs a heel cap.

I'm not routing the top and back yet for the binding. The back will have more screws in it. Maybe I'll try purfling for the first time.

Now getting close to the interesting part - installing the struts, and seeing if they keep the top straight, and how/if they can be adjusted to affect the sound of the guitar.

[Peter Wilcox]

My idea for the struts derives from opposing the torque of the strings on the bridge/saddle with an opposing torque from the bottom of the bridge. I realize that much of the sound of a guitar is felt to be generated by the bridge rocking the soundboard, and that this would stifle that action. But then, the Bridge Doctor has this same effect, albeit using compression instead of tension, and at least some folks don't think this has a deleterious effect on the sound. So I thought it would be worth at try, particularly if the soundboard were unbraced or very lightly braced to compensate. Anyway, I am finding I have a huge amount of ignorance regarding bracing and its effect on a guitar's sound, but I'm certainly enjoying climbing up the learning curve (though I'm still very near the bottom).

So here's the first iteration of the struts. I used 3/16" cold rolled steel that I was using for truss rods. They were fastened to the bridge and the heel block using sex bolts so the tension could be adjusted from the outside. Too stiff, too much mass, producing a low volume tinny sound.

So I took them off, and strung up the guitar with no bracing. It tolerated this well because the top was too thick (0.135"), and although the volume increased by several dB, it still sounded very thin. Leaving it strung up overnight, bellying of the lower bout and caving around the sound hole was becoming apparent by the next morning.

[Peter Wilcox]

So my next step was to thin the top. I got out the old belt sander and took it down to 3/32" (0.095) in the lower bout, with a gradual increase to 1/8" from the mid sound hole area to the heel. I then built a new set of struts from 1/8" carbon fiber rod and maple.(The extra holes were in case I needed to double the short CF rod.)

The volume and tone have both improved considerably. The volume is comparable to a cheap Aria dreadnought I have, but the tone on the bass side is lacking in higher frequency harmonics (I think that's what it is.)

Here are a couple of shots of the installed struts.

[Peter Wilcox]

So the next step might be to thin the top some more, maybe to 0.08". I can also thin the elbow pieces of the struts to take a little mass off and use aluminum sex bolts on the bridge instead of steel. For now, I am trying to read through Howard Wright's PhD thesis on the acoustics of the guitar, and maybe see if some judicious use of light braces may improve the tone. I guess I'll have to look into Chladni testing too.

Any suggestions, critiques and comments welcome.

Here is a pic of the guitar as it now stands.

[Hans Bezemer]

I'm watching your experiment with interest. I'm looking forward to the outcome of your next step.

[Dave Sayers]

I've been studying up on phase cancellation, which can occur when the soundboard is producing opposite phase vibrations. I got the info from an extremely clever Australian luthier named Daniel Brauchli. He has a whole section on bracing in his website. I also felt my braceless top 12 string was weak in the bass sounds until I realised that it was producing a much louder and fuller set of mid to high frequencies....and then I understood how clever x bracing is because it helps to emphasise certain tones, except that I've come to really like the full mid and top range sound.
Might I also suggest you try carbon strip rather than tube...it flexes less, much less used sideways on. I am finding this fascinating.

[Dave Sayers]

I also find it interesting that you are basing all of the support on the neck block. Is this to take advantage of the string vibration and sustain?

[Peter Wilcox]

Thanks for the website Dave. I'd seen it already, but going back now I can pick up new ideas. I can see that CF with a rectangular profile might work better than round rod. I've already seen with the guitar strung up for a few days that one vertical rod isn't enough, so I will install the second ones that I've already planned for and drilled the holes. I did a spectrogram of the guitar strummed with open strings, and compared it to the cheap dreadnought that I have. I know these probably aren't very useful for evaluating it, but there seems to be an awful lot of "mud" below -50dB, especially below 70Hz. I'm going to try to record a Martin spectrogram tomorrow to see how it differs.

I'm using the heel block for support as this seems the strongest place to attach struts using tension. I don't know how it affects the sound.

My plan for now is to thin the top a little more and add the other struts. Then I'll have to see what I can do to improve the sound.

[Dave Sayers]

Here's a link to a small video of the 12 string. https://www.youtube.com/watch?v=FDWAVCnNVz4 This link works on my computer

I have to be honest and say that I can only see small differences in the spectograms and I wouldn't have the foggiest idea what they meant anyway. I think I might stick with the earograms for sound testing.

I was referring to carbon strip as used in the 12 string, but triangle and square section would work just as well if hollow I think. Don't forget to allow for tonal difference between actual woods.

[Peter Wilcox]

The single vertical strut version maxed out on tension - turns out they were leaning, not strong enough for the force they were subjected to

So I added another strut in the holes I'd already drilled for this possibility.

I thinned the top toward the edges to about 0.07", and put it back together. To my tin ear it sounds much better, and so far there's no bellying or caving.

[Peter Wilcox]

I did another spectrogram of an open string strum with the same mic I used before - a cardioid condenser.

I did another with an omni measurement mic, and seemed to improve the bass frequencies - maybe due to the loss of proximity effect of the cardioid.

Of course, these spectra are probably meaningless. I tried a tap spectrum, but I don't know how to compare it to the plate spectra discussed elsewhere. This guitar has to be tested strung to pitch (strings muted), and with the back on, as those are major components of the bracing.

A thanks to Trevor Gore for his posts here in MIMF regarding Visual Analyser and how to set it up.

Now I need to get into Chladni patterns - maybe they'll help determine where I can put some extra braces to shape the sound.

[Peter Wilcox]

I made spectrograms of each string, plucked between the bridge and sound hole, and recorded over the sound hole. Here is low E - I don't know how to use these to evaluate the guitar's sound - maybe I'll need to record some other guitars and compare.

[Peter Wilcox]

Then on to Chladni patterns. I put an old 4" speaker into a flower pot, and found a function generator I made about 40 years ago, and hooked it up to an amp and oscilloscope (to determine frequency). My computer is about 300 feet away, so I couldn't use it for this, and if I tried to do the patterns where the computer is, the dog and cats might want to get involved.

So here's the set up.

[Peter Wilcox]

So just messing around, I got these patterns.

200Hz

230Hz (Oops -didn't have any sparkle on the upper bout)

1140Hz

I really don't know what to do with these. I guess placing weights on the soundboard, or temporarily attaching braces to see how that changes things. The crux of the matter is how the guitar sounds.

There is a bluegrass festival in town this week end, and I'll take it there to see if I can get some folks to play it and get some opinions on the sound, since I have a tin ear for that kind of thing.

[Peter Wilcox]

Well I got some folks to look at it and play it at the bluegrass festival - lots of helpful comments and suggestions.

Here's a sound clip at low resolution mp3.

[Dave Sayers]

Sounds good over my low tech speakers. Perhaps a little more experimentation and then I suspect the mark 2 version might start to take shape. I took the dreadnaught with the carbon and light X braced top to an open mic, and didn't get it back for ages. The general verdict was that the light carbon bracing gives a definite increase in volume and the ability to cut through other guitars with a fuller mid range and treble. I'll have to pull the pictures off my phone but it was being played alongside a Gibson L? series of some vintage and definitely held up in the sound stakes. I can never understand how to post pictures to these forums as it seems to involve resizing so I'll add a link. I must confess that I find the earhole assessment to be the best one.

[Peter Wilcox]

I'd like to see/hear the guitar, Dave. You can resize pics with this free download if you use windows: http://www.irfanview.com/ There are instructions in the MIMF faq on how to post the pics.

Regarding my guitar, the consensus of a number of people (players and luthiers) was:
It had pretty good balance but possibly deficient in treble, though good bass response.
It lacks projection or volume, though I think most of them were probably comparing it to the Martin dreadnoughts they all seem to have. But everybody was very interested and encouraging.

Suggestion for improvement:
Decrease the mass associated with bridge (including associated strut supports)
Make vertical struts more flexible to allow the bridge more freedom to move/rock (but still need enough stiffness to keep the top flat)
Brace the treble side of the soundboard around the bridge to increase stiffness there and augment the treble response
Use medium gauge (instead of light) strings to drive the soundboard harder
Try nylon strings for lower tension (with lower opposing torque from the struts) and see if this increases the volume

I have some ideas to work on some of these things - I'll see what happens. Any suggestions appreciated.

[Dave Sayers]

I will endeavour to do the pictures today. I'm planning to take the dreadnaught out to an open mic tonight so if I can get someone to take pics of me paying I will. This one has an x brace that stops 1" short of the kerfing with a line of carbon ribbon around the edge between the kerfing and the ends of the x braces. The carbon ribbon prevents wood splitting and as the glue was high flexural strength it allows fairly free movement of the top. There are 2carbon rods from bottom of neck block to top of tail block to prevent the tail block moving and thus the bridge. One carbon rod beneath bridge to a bottom brace to support bridge. The general concensus is that it is loud.

[Dave Sayers]

Suggestions, carry on Peter. I'm quite sure you are able to build a standard guitar, just as I am, and you will. There's no need to stop experimenting simply because some people only want the accepted standard. After all, we would still be twanging cat gut and banging hollow logs if it wasn't for some clever person trying something different. I must confess i've never heard the name Barry Daniels mentioned in the same breath as, for example, the Australian Greg Smallman. Or Howard Klepper, Michael Thames, Daniel Brauchli (I hope I spelt that right) and many other fine luthiers I do not know personally but whose work I admire. I am often running into experts whose knowledge seems gathered rather than earned. It is your money, your time and your wood. Don't stop now.