Official Luthiers Forum!

I'm gearing up to be able to do some dynamic testing of guitar tops. I have a power supply, amp and a speaker, and I managed to come by a a audio
signal generator cheaply. (hardware unit, not software).

For recording, I'd like a 'scope and a way to digitize signals. I'm assuming that I'll be using an extra Pick-Up-The-World film transducer and preamp that I have sitting around to measure soundboard motion. I've thought about taking my chances on an ebay scope, but if there is a good software package that would serve my purposes, please let me know your preferences. I can run the analysis on either a PC or Mac laptop. I have Matlab loaded on both now, and could possibly work with that alone if I find out that there is a decent oscilloscpe toolbox.

_________________
Jim Kirby
kirby@udel.edu

There are lots of ways of doing this. I have 2 engineering students that are doing some of this testing with me. We plan to compare some of the different methods of measuing MOE both static and Dynamically and see how the results match. One student is an electrical engineering student and the other guy is a mechinical type. The mechnical guy is going to do the static testing. We also will be doing some Q value testing but today we did not get very far with out attempts. I suspect that will get most of our Q values from looking at the "right" peak modes in the spectrum.

Speaking of spectrum, there are some FFT (Fast Fourier Transformation) software prgrams out there. I bought one that I really like called SpectraPlus. It is not cheap, but it is fairly easy to use. I am sure MatLab can do FFTs to but I am not familar with that at all (since DOS versions.)

After we get done with some of this plate testing we hope to look a free top (guitar shaped) and back modes in detail, but that is a long way off (for the students, I have been messing with that stuff for years, but by no means an expert)

There is a lot of information "out there" on this stuff. The students are sifting through a lot of it now to come up with the different models to play with and are doing some of this on thier own. Although it is sort of a directed study, it is actually more of an independent study for them. I am not sure where it will go for sure.

One thing I have found though, is that "I" am not good a keeping great records. If a test set-up is going to take me all day to get some "value" that may or may not matter in the final product of a guitar, I am not going to do it. So "for" me, the test needs to be easy and I suspect that at the end of this, I will be using SpectraPlus and glitter for most of my testing. Hopefully we can figure out what we want to measure and come up with a fairly efficient and consitant way to do this.

So to answer your quesion of hardware or software, for us, we will be using both.

John O

I have an old hardware 'scope, and plug it in once every few years. Not being very experienced with them, I'm not sure how you would be using it. I have a software 'scope, and can't think of much use for that, either!

'Wavesurfer' is a decent freeware recording and analysis program that you can download from:
http://www.speech.kth.se/software/
It works through your sound card, so all you will need is a mic to plug into your computer. It does not have a 'scope' function, but will do FFT, spectrograms and so on.

Chladni patterns are a good way to isolate the lower order modes: without them it's sometimes hard to know what you're hearing.

Thanks John - Hopefully your students will post here.

Al - I've seen the references to Wavesurfer. Since I already have matlab on the computer and runable anywhere in the world as long as I can touch the university site license over the internet, I'm hoping to figure out how to get it to run the soundcard in the same way. The search continues ...

I was thinking of using a soundboard transducer rather than a microphone so that I'd get a direct measure of the motion. When you are using a mic, do you hear mainly the speaker you are using to drive the system, or does the soundboard itself radiate a lot more sound when it hits a resonant frequency, thus controlling the amplitude of the audio signal?

If the answer is that the soundboard is determining the sound level, then, I can probably just use one of the Shure 57's sitting around.

_________________
Jim Kirby
kirby@udel.edu

jtkirby wrote:
"I was thinking of using a soundboard transducer rather than a microphone so that I'd get a direct measure of the motion. When you are using a mic, do you hear mainly the speaker you are using to drive the system, or does the soundboard itself radiate a lot more sound when it hits a resonant frequency, thus controlling the amplitude of the audio signal?"

Driving the guitar with a speaker (air-> air) is a lousy way to do things. Most of the energy just bounces off, and you end up making LOTS OF NOISE for not much result. I like to use a 'stinger' driver for a lot of tests.

Get a small, long-throw mid-range with decent power handling capability. Cut most of the cone out, leaving a few strips along radii to support the coil. Get a wine cork or a piece of balsa, and shape it into a cone. Glue the baseof the cone to the coil, so that the 'stinger' point sticks out past the frame of the speaker.

In use, you can hang the guitar up or mount it in whatever way seems best to you. Set up the speaker so that the tip of the stinger is in contact with, say, the bridge saddle. Hook it up to any convienient signal source. What you've done, of course, is to turn the guitar itself into a speaker. The actual load of the coil is not all that great, and does not seem to purturb the resonant pitches to speak of. I find a small improvement is to mount a small clip, such as a paper clip (the ones made of flat spring steel folded into a pinced 'C' shape) and clip it onto the saddle. With the normal stinger you have to have enough pressure on the tip to keep it from buzzing, but the clip allows you to use no pressure, and can pull as well as pushing. THe added mass is negligible on a guitar.

This type of driver emits almost no sound of it's own, since the amplitude is strictly limited. You can thus back off and record the sound in the room with some confidence. Monitoring the current through the coil allows you to corect to 'constant force' driving conditions, if that's what you want.

The problem with soundboard transducers is the same as that of all pickups on guitars: a limited view. The charm of the guitar is that it's producing some sort of sound from every surface, and there's no way a single transducer can get it all. It is, as I like to say, like watching a ball game through a hole in the fence: there's always some part of the field you can't see.

Mind you, the sound in the room isn't a lot better in many respects. The fact is that there is no simple test that will reliably tell you everything you'd want to know. And, boy, can you waste a lot of time at this stuff!

Thanks Al.

I have more than enough engineering degrees to make wasting time on this stuff inevitable. (People can't believe I haven't gotten to all of this long before this point.)

_________________
Jim Kirby
kirby@udel.edu