Official Luthiers Forum!

well since I have the sweep generator, I may as well play around with it , my question is , how many decibels are needed to drive a top ( or back ) plate to form the Chladni patterns? I get the feeling I will need to amp up a bit .thanks Jody

A 4-5 " speaker at around 15-20 watts should make it.
Wear ear protection!

thanks alain ... Jody

What Alain said, pretty much. It depends in part on the frequency range and the pattern. Often you can get the lower frequency ones to form with only a little power; one or two watts and you can barely hear it. As you get up past 200 Hz or so it can start to take more power, and gets REALLY LOUD up near 500 Hz.

WEAR EAR PROTECTION! (Eh, what'd he say?) The peak to peak pressure change for a pure sine wave is higher than for a more normal sound at a given wattage, and it's the peak pressure that blows ears and speakers. As someone who wears hearing aids (not because of plate tuning: the loss goes back before that) I can tell you that this is something you don't want to mess with.

The Bradley signal generators come with a 15W power amp, IIRC, and that has proved to be adequate for everything I've tried with it.

Thanks Alan, I have a protech sweep signal gernerator , not many watts, it makes enough sound <i just need to figure out how to amplify it. since I am no techno wiz. I am hoping it is as simple as playing it through my home stereo system... we'll see.

so Alan while I have your attention... you are testing square plates, thinned close to your usable thickness? or at the usable thickness? thanks Jody

I test rectangular top and back half blanks, to find the material prooperties. These can be any thickness, but they do have to be uniform. Recently, inspired by Mark Blanchard, I've been looking at the modes of unbraced tops that have been cut to shape. He uses some of these modes to tell whether the top is likely to work out well on his designs, and I'm trying to find the correlation between his mode data and my stiffness ratio information. Then, of course, I use Chladni patterns to tune the top after it's braced, and before gluing it to the sides.

I'll be the first to admit I'm vastly more intuitive than scientific when it comes to approaching voicing, but dang Alan I'd love to spend a few days in the shop with you learning some of this stuff just to expand my horizons.

_________________
John Mayes
http://www.mayesluthier.com

realy thats me too John !But I have the sweeper ,so I may as well try it . I was talking to a high end violin builder for a few minutes this morning , who uses the patterns ,which was very informative.one thing he said was they can duplicate their sound by adjusting and duplicating the various patterns . of course violins are a much simpler system for producing music .ie fewer braces and parts . Jody

Jody:
In some ways the fact that violins look simpler makes them more complicated acoustically. They are more like 'distributed element systems', in which the stiffness and mass are spread out everywhere rather than being in specific places. Any system will have as many modes as it has elements that can move, and since a distributed system has (theoretically) an infinite number of elements it can have an infinite number of modes. Of course, most of those modes are up where even mosquitos can't hear them, but that's another matter. What it means in practice is that, even more than on guitars, changing _anything_ can alter the sound of a violin in ways that will be difficult to predict. Guitars are bad enough!

I will say that, having used Chladni tuning on both violins and guitars, I think it works better on guitars.

Alan,since i have not tuned either .. what the violin maker told me is they are concerned with duplicating 5 patterns, the fundamental, and the next four octaves, what do you think about that ? thanks Jody

There are five or six patterns that most violin plate tuners look at. Since the violin is pretty standardized they come in the same order on pretty much all violins, and we just number them as #1 through #5 or whatever.

There's a heirarchcy of things that I look at when tuning violin plates.
1) Get the mode shapes 'right', if you can. IMO (and it's only my opinion) the shapes of the patterns are probably more important than the actual frequencies, in some ways. The shape tells you something about the distribution of mass and stiffness throughout the plate, while the frequency tells you about the ratio of mass and stiffness. Note that the stiffness distribution will depend on a lot of things: the character of the wood and the arch height and shape, as well as the thickness distribution.

2) Try to match the frequencies of the #2 modes. If the frequencies are much higher or lower than normal you probably have a plate that is too floppy or too stiff, unless it's much heavier or lighter than usual. I'd expect a balsa top to have very high mode frequencies if it's stiff enough to do the job. We're not entirely sure why this matching works, but it may have to do with establishing strong 'B1-'and 'B1+' modes in the completed fiddle.

3) If you can match the #2 modes, try to match the #5 mode frequencies as well.

4) If possible, get the #2 and #5 mode frequencies into an octave relationship.

5) If you can do that, _and_ get the #1 mode on the top (but not on the back) to be an octave below the #2 mode pitch, that might be even better.

These are the rules that Carleen Hutchins derived from her work. I'll note that other makers have come up with different sets of rules. It's possible that the rules that worked for her might not work as well for other makers, since her archings are a bit different from many other folks'.

I will say that the best sounding fiddle I ever worked on was a cheapo that I regraduated for practice. The archings and wood were such that, when I got the graduations 'right' the tuning was 'perfect'.
The top modes were:
#1 - 90 Hz, #2 - 180 Hz, #5 - 360 Hz.
The back modes were:
#1 - 110 Hz, #2 - 180 Hz, #5 - 360 Hz.
It was still a wretched looking French cheapo, but the sound was glorious.

In the end, I feel that any stringed instrument is a system with many co-dependant variables. None of them are really independant, and the 'right' value for any of the variables will depend strongly on what the others are. If you get all of them 'right' you've got a great instrument. If you get a strong majority of them 'right' you end up with a good instrument, and it almost doesn't matter at that point which ones you get right or what the rest are so long as they are not too far out. There is a range of possible values for any of them, and values that simply won't work well, no matter what you do. Plate tuning is almost certainly one of the variables, and it was explored early because it was fairly easy to look at. It's not 'The Secret', but it's part of the secret. I think the secret is knowing what the permissable ranges of all the variables are, and how they each effect the others, so that you can tweak them in such a way as to get the best possible balance without putting any of them out of range. In other words, the secret is in knowing what you're doing. Simple, really.

thanks Alan . jody