Official Luthiers Forum!

Now there's been lots of info an various forums about deflection testing for tops and the most common setup is a board with 2 rails 18" apart, the top/back is then loaded with a 5lb weight and a DTI to measures deflection, now the ball park figure for the top is massive I've read figure from 0.200"-0.350" but most fall in the region of 0.200"-0.250", people seem to like to leave a thicker top and have lighter bracing they say for more complex/richer sound.

Now my questions are
1. Do you always aim for the same along the grain deflection not matter the body size/shape?
2. Is the cross grain stiffness taken into account? eg the along the grain stiffness is still too high from your target deflection, but you feel the top will get too floppy across the grain when you reach it
3. There is not mention at all when I've trawled the forums for information on back deflection are there figures for say a responsive back and a reflective back?

Thanks Guys especially Brock, whose advised in PMs but I thought more people might benefit from my ignorance

John

_________________
Formerly JJH

I learn more from my mistakes than my successes

John,

I am interested in these questions as well. I just did deflection testing for the first time last month. With nothing but dumb luck I ran my rails at 18" for the long grain and at 8" for the cross grain. The final measurements I took, after roughly cutting out the body shape, were .275 for the long grain and .218 for the cross grain. The long grain measurement changed real fast from .215 to .275 when the top felt to be where I would normally thickness it. The cross grain stayed farliy stable, maybe because the rails were too close together, really don't know. Hopefully some of the guys who do this on a regular basis will chime in and shed some light.

Tim

For what it's worth, I ran a batch of tops through the sander recently aiming to get .250 deflection with a 5lb 4 0z weight between a span of 18" and every one came out at an average of .93, significantly thinner than what I was going to before which was .110. I have yet to finish a guitar built with these tops, so I don't know if it's better or worse. Or more accurately, if I'll like it more or less.

I was actually just talking about this the other day with one of the guys I share my shop with and am curious how other people keep track and why. I find the top deflection and then divide by the body length. My reasoning is that the tops will then arch in the same way regardless of length. If you ignore body length and just measure deflection, a shorter body length will have a more severe arch than a longer body. Not really sure what difference it makes, but for my thinking dividing out the length seems to make the most sense.

Mike

_____________________

Indian Hill Guitar Company
indianhillguitars.com

I'm with you Mike. I check mine at 1" shorter than the actual body length, and do cross grain testing at 1" narrower than the actual body width. The desired deflection, and weight it takes to reach that number vaires with the projected string tension.


Added: I check the long grain deflection at the center of the top, as well as the bridge location. I check the cross grain deflection in the center, and the center of the lower bout. I've been doing this for 25 builds, and 3 re tops.

_________________
I didn't mean to say it, but I meant what I said.
http://www.brackettinstruments.com/

Personally a singular measurement in regards to a singular plate really tells you very little without previously gathered data from many plates at a given thickness within a given species. It is mass accumulation of data that paints the picture. Not the singular Brush stroke. Don’t get stuck on a given thickness and a given amount of deflection as being the magic formula. Collect the data use previous history from successful builds and as time goes by you learn to put the puzzle together.

Since we're talking about deflection & I've not yet posted these, thought I'd put up a couple of pics of a deflection jig I made recently for archtops. Obviously, testing an arched instrument is not as straight forward & there is very little to go on, but the concept came from Don MacRostie & some mandolin testing pics he posted on mando cafe. His involved clamping pressure & fishing scales as opposed to freeweights and a gimbling "bridge" on mine. It works on both my carving jig (for free plates and closed boxes) and the binding cradle for completed instruments. Maybe it will help someone.
Hopefully, I'll learn something as I test guitars in progress and completed ones during "checkup".

_________________
Dave
Milton, ON

Here is what I do, but what is important to understand, it is only what I do.

Currently I am measuring back deflection but I am just collecting data. For me the back is like first grade. I have a lot to learn and I am just getting started, so I am accumulating data. Slowly I am changing my approach to the back as a whole, from stiffness to bracing.

As far as top deflection goes, I personally feel that you should NOT aim for the same degree of deflection for all sizes of guitars. My feeling is a guitar top is like a drum head. If you have the same stiffness top on a little drum head (parlor) and a large drum head (a Jumbo) you will have very different results. In very general principles, I thickness my tops to a lower degree of stiffness for parlors that I do for larger bodied guitars. For me, with my style of building, this approach has been very successful.

To confirm this concept, I took a guitar that I built years ago that I didn't like. It was a parlor. (You can probably see where this is going) I used to take all my tops to the same thickness. An approach taken by many builders. Then I took Ervins class and I brought that parlor to class. I quickly learned why I didn't like that guitar. It was massively overbuilt. With a parlor sized body it made it even worse. (Thick top small drum head...yuk) So I routed to top off and put a new top on with my new knowledge. This top was much lighter braced and thicknessed to lesser degree of stiffness. Nothing else was changed in the guitar. AMAZING difference. I love this guitar now. Very well balanced. Amazing bottom end for such a small bodied guitar. One of my favorite guitars.

Michael Payne is right on. A single number is worthless. It is the collection of a lot of data that is worthwhile. I have a database for every guitar I have built over the past 3 years. Everything from body size, back and top deflection data, Brace height and taper, wood choice, photos etc. I also have notes on the end product. Tonal qualities etc. This is priceless. So if the guitar has too much bass response, I might look a my data to see why and make a change the next time I build that size guitar.

_________________
Andy Z.
http://www.lazydogguitars.com

A few thoughts on scaling deflection tests for different body lengths:

The deflection of a beam (or guitar top) is proportional to the 3rd power the beam's length. E.g., for a "simply supported" beam (aka, "3-point bending"), the equation for deflection is d = W * L^3 / (48 * E * I)
(For terms, see http://www.engineersedge.com/beam_bendi ... nding2.htm . A guitar top is supported by flexible end-conditions, so it's somewhere in between the pivots in "simply supported" and perfectly rigid supports, but the L^3 term shows up regardless.)

So, if you want to scale a target deflection between different body lengths, it's a good idea to either:
1) scale the deflection test fixture accordingly, like Woody said -- e.g., if testing at 18" for a 20" body, test at 17" for a 19" body. That will take care of the L^3 correction fairly accurately.
2) Or, if your deflection fixture has fixed 18" supports, the deflection target can be scaled. For scaling, I think the relevant length is the inside distance between the tail block to the UTB. For a 20" body, that's probably around 16.25"; for 19", 15.25"? So let's say you know your target deflection for a 20" body is 0.25" with 18" jig supports. If you want the same top stiffness at the bridge for a 19" body (as 20"), the target deflection in the 18" test jig would be roughly 0.25" * (16.25/15.25)^3, or 0.30".

I don't know if it's good to have the same top stiffness at the bridge for 19" and 20" bodies, but hopefully this will help people calculate and get what they do want.

_________________
David Malicky