Official Luthiers Forum!

I've almost completed my current build, an OM-sized guitar based on a 1937 Martin 000. Varnishing should start any day now. I did my first extensive job of modal analysis on this one, and it was so much fun that I decided to post some results. I'm very interested to hear your comments/suggestions. The top is Carpathian spruce and the B/S are black walnut.

The tap tones were made using a homemade rubber mallet, striking the soundboard multiple times in the region around the bridge. I took measurements with the bridge off, and with the bridge taped on, although I only show the former results. The back measurements were made tapping over the entire area. I used Audacity for the FFT analysis and plotted the results using Kaleidagraph. The back spectrum has been reduced by 20 dB in the plot below.

The Chladni patterns were generated using a 3" car speaker (driven by an online frequency generator) with poppy seeds on the guitar's top and back. The guitar was supported at the neck and tail blocks using small cubes of foam. I found it easier to get good patterns without the bridge tapped in place, as the bridge always blocked a lot of the seeds' motion at higher frequencies. The final top pattern in the sequence was taken with the bridge taped down, and it shows that the top's monopole dropped about 20 Hz due to the weight of the bridge. I'm a bit worried that this is pretty low in frequency, although I expect it to rise a bit once the bridge has been firmly glued in place. The dipole that appears on the top at 293 Hz is very weak and actually comes from the back. I neglected to include the back's monopole, but it shows up at about 210 Hz.

I thought this was an incredibly fun exercise (I'm a left-brain sort of guy), and I can see that it can become a very useful analytical tool once I get some more guitars under my belt.
Thanks for looking.
Chris

Looking good, Chris. Nice, clear patterns.

Keep an eye on that main air resonance to make sure it doesn't land on G.

_________________
Trevor Gore, Luthier. Australian hand made acoustic guitars, classical guitars; custom guitar design and build; guitar design instruction.

http://www.goreguitars.com.au

To me that is interesting. You seem to have top monopole patterns at several different freqs.

I like my OMs to be 180hz strung up coupled. The pic with the bridge on shows that, but you can expect the glue to stiffen the coupling raising the freq by 5-10 HZ. But you're within through the soundhole planing range.

In the back patterns I'm not seeing anything that looks like a monopole, but the major third interval Trevor suggests works great. If it still sounds tight at 180-225, take a half step off the back, not the top, you'd be surprised what this does for the bass response. It works on smaller guitars but not larger, I would never do that on a dread (unless my spidey senses told me to).

I need to get into some Chladni action...

Thanks for the advice Trevor. (Nice, clear books by the way!)

Ed, I did not include the back monopole photo in my post, but it was easy to locate at about 235 Hz, not 210 as I originally stated. The top monopole pattern at 235 Hz is actually coming from the back. It's easy to tell, because while the top was vibrating rather softly at 235 Hz, the back was going crazy. Ditto for the long dipole at 293 Hz. The pattern at 135 Hz is a weak dipole, where the center of the top vibrates out of phase with the perimeter.

With the exception of the "S" shape at 715 Hz, the top patterns are pretty easy to understand from the bracing pattern. The back patterns simply mimic the back's ladder braces, with each of the four rungs progressively acting as an antinode up to 548 Hz. The patterns get more complex above that.

This turned out not to be as difficult as I had imagined. And as I said, it was a lot of fun. My only advice is to place a tray under the guitar for collecting all the seeds/glitter. It gets pretty messy!