Official Luthiers Forum!

At my rate of maybe 1 guitar per year I've realized that I'll never flex enough tops to get a real good feel for the appropriate thickness. Deflection testing seems like a reasonable way to get some objective info. I'm sure it can get very complicated but I'm looking for a very simple setup - Top suspended at either end or side, weight in the middle - just the 2 measurements. Can anyone suggest what weight to use and some ballpark deflection values? Thanks!!

The numbers I’ve most heard is 5lb over an 18” span. That’s what I use...

Also, my weight is 5lb of pennies in the longest whisky tube I could find, which distributes the weight across the top rather than centralizing it. I only test long grain, I don’t care about cross grain deflection.

5 lb weight and and a span of 18" is what I've seen from a number of builders and it's what I use too. There's a benefit of going with something close to that semi-standard in that you can approximately compare your numbers with those of other builders using the same or similar set up.

I originally learned to assess how much to thin the top using the "oil canning" method (I think people have other names for it too) where you shake the top back and forth while holding by the edges. When the top has been thinned enough that it starts making a wacka wacka sound, the thickness is close to what you want. With my setup, when that point is reached, the deflection has typically turned out to be around 0.180".

Anther thread has a discussion about deflection testing: https://www.luthiersforum.com/forum/viewtopic.php?f=10101&t=50083&p=659891&hilit=oil+canning#p659864

This is the set up I use:

_________________
Once in a while you get shown the light in the strangest of places if you look at it right - Robert Hunter

are you holding the edges along the sides (long grain) or ends (end grain) of the top?

_________________
Bryan Bear PMoMC

Take care of your feet, and your feet will take care of you.

Along the long edges (long grain) is how I was taught to do it.

_________________
Once in a while you get shown the light in the strangest of places if you look at it right - Robert Hunter

I have been doing deflection testing for about 8 years and basically have correlated it with the final outcome as far as how the instrument was rated by a small stable of great players who worked or hung out at the store where I sold and of course what I thought of it.

I built a frame, selected a spot about for measurement about 3.3” below the soundhole in the midline and pre-weighted with one lb. zeroed the caliper, and added an additional lb. Took several readings until things stabilized.

I think pre weighting is important. I do it after the rosette is installed and with the soundhole open and the top cut to approximate final size.

I think it has been quite helpful for consistency. With my technique a deflection around 0.045-0.050 is a good ballpark.

Just one cog in a million little things though.

Untitled by Terence Kennedy, on Flickr

_________________
It's not what you don't know that hurts you, it's what you do know that's wrong.

This is my very simple set up. For any given model guitar I extend the top sheet 1cm past the outline and use a dedicated brick placed right on the bridge patch location and measure dead center of the X-Brace.

Here’s mine...


Sent from my iPhone using Tapatalk

Do any of you use different amounts of deflection on different size guitars ? Or different woods ie. spruce vs cedar vs redwood ?
Hope this isn't considered a hijack-still all about deflection.

Brent

Top stiffness is there to withstand bridge torque; and top that's stiff enough would be more than strong enough to take the string tension as a simple load. The problem is that it's not simple tension or compression because the strings are above the plane of the top.

The torque load will be a product of how much tension the strings pull and the height of the strings above the top (yeah, technically it's the height above the center of moment of the top/bracing system, but it's a small quibble in this case). You can reduce the torque by using lighter strings or by making the bridge lower. This allows for a thinner top. They use low bridges on Flamenco guitars, and it contributes to the sound they're after by allowing for a thinner and more responsive top. So, yes, you probably would want to use different deflection measures for different sizes and flavors of guitars.

In this respect what's important is that you establish some sort of measurement routine, and keep records of what you did. I use a vibration test to find wood properties myself; I already have the signal generator, so what the heck. What's important is that you set up a test that you'll use and that's reasonably reliable. Then keep track of what you did and how it worked out. You don't want to be in the position of making a really fine sounding instrument and not remembering what you did.

Definitely different numbers for different size guitars.

Same numbers for different woods.

Do you find that the tops tend to sag a bit before adding the weight? My deflection jig has the dial indicator above the wood, but I’m wondering if I should re-design to measure from below like Ed.

This is particularly timely for me, as I have five Sitka tops ready to go to repeat a test I did comparing deflection results to the results of the Gore equation. I want to make sure my results are as valid as can be.

Sent from my iPhone using Tapatalk

I definitely think that adding a preweight before zeroing is a good idea.

I don’t though, lol.

It’s actually pretty rare that the wood I use is dead flat, which is also why the tap method is something I don’t do.

I just measure one sides deflection, flip it, and use the average.

Bricks, tubes of pennies, dumb bells, quite a variety - weights, methods, lengths, which is fine if you're just developing your own set of data.
I try to measure E(long) - modulus of elasticity, long grain, in ABSOLUTE terms, measured in GP (gigapascals)
Otherwise I can't make comparisons with anything, or anyone else, it's like comparing apples to bananas.
450mm spacing and pre-weighed magnetic tool bars (8" and 12") for weights (stick together nicely), measure in the middle, underneath the weight position.
PRE-LOAD each half of the plate (otherwise quite large errors are induced) with an 8" weight/toolbar, then measure the extra deflection of added (12") weight(s) - mine are about 1.7kg..
This extra deflection is used for the calculation.
Turn the plates over and measure again - I use the average of 4 readings in each direction.
Or of course you can use Gore's frequency method.
At the same time I flex plates to try to get a feel for them, tap them, rub them, listen to them, take notes.

_________________
The name catgut is confusing. There are two explanations for the mix up.

Catgut is an abbreviation of the word cattle gut. Gut strings are made from sheep or goat intestines, in the past even from horse, mule or donkey intestines.

Otherwise it could be from the word kitgut or kitstring. Kit meant fiddle, not kitten.

I don't remember where i got this, but it works for me. I fixed two 1/2" dowels 18 " apart on plywood. Centered between them is a 1/4" dowel. Send the top through the drum sander until the top touches the 1/4 " dowel with a 5 lb. weight added to the top. So, .250 deflection over an 18" span with 5 lb. Gets me in the ballpark.

Exactly. You have to measure for modulus in order to make an apples-to-apples comparisons. For several years I measured plates using both methods, frequency and deflection. The deflection test was used as a backup to verify the accuracy of the frequency measurements. I found that they were always so close that I eventually stopped doing the deflection tests.

_________________
It will probably be alright.

I mostly use the tap frequency method as taught in the Gore/Gilet books. I have compared the target thickness to both defection and for steel string the shacking method Jay described above. All three of the methods compare for a steel string. With the deflection testing I solve for both long and cross grain Young's modulus and plug those numbers back it to the Gore equations. I use the deflection method when I get a non recectangular plate I need to test.

How many do deflection testing on a braced top and what sort of deflection is looked for for a given mass?

_________________
http://www.Harvestmoonguitars.com

Exactly what and how I do... i will not share my prefered deflections but will tell everything else.
https://howardguitars.blogspot.com/2016/01/deflection-testing-of-acoustic-guitar.html

_________________
Brian

You never know what you are capable of until you actually try.

https://www.howardguitarsdelaware.com/

And a big thank you to you Howard becasue that article got me started.



That's my take on it too. If I am going for the same quality of tone then regardless of top wood species I go for the same deflection. And it absolutely makes sense that a little parlor guitar will have a different deflection then a Dred.

---

When possible I use the Gore/Gillet method as well and I was very happy to find out that once I started using that method it is pretty much spot on to my deflection testing. In fact it's close to the traditional stiffness testing by hand and feel that I do too.

The first thing I do is thin the top out till it has that sheet metal sound when shaking it back and forth. That's when it's pretty much good to go and that simple test is very close to the G/G acoustic testing and deflection.

In the end it's good to have a couple of methods for getting it right

I have to say all these bricks around the fragile tops rather worry me! Couldn't we find some less unfriendly weights to use?

The principle of measuring the Youngs modulus has to be right though if we want to be consistent, given that we know the properties of any particular piece of wood is not defined by any average value for the species.


Dave

Is the mass you're asking about the total mass of the top + braces?

I started measuring deflection on braced tops three guitars ago just to see if anything interesting would come out of it. All I'm doing at this point is collecting numbers and not trying to use them to guide any decisions, yet.

The set up I'm using to do take the measurements is shown below. The numbers from the braced top can't be compared directly to the deflection numbers for the original joined rectangular plates for several reasons, so they are stand alone numbers.

Deflection numbers for the three tops were 0.044", 0.032", and 0.049" with fully carved braces. The guitar with the smallest deflection value is the least responsive of these three guitars to my ears, although there could be a number of reasons for that difference other than the deflection. A sample size of three is too small to draw hard conclusions, but it's something I'm going to pay attention to in the future.

For the top in the photo, I took deflection measurements before and after carving the braces:

Uncarved braces = 0.040"
With cap added to x-brace joint = 0.027"
Fully carved braces = 0.049"

I was impressed by how big a difference there was between the uncapped and capped x-brace joint numbers.

_________________
Once in a while you get shown the light in the strangest of places if you look at it right - Robert Hunter

Wow that is impressive with the X-Brace cpa, I never would have guessed it made that much of a difference. I too measure deflection on braced tops. I'll have to give that a go next time and compare before and after.



If you look closely at my pic you will see the brick is laminated with cork...

Besides a chisel would probably do more damage and we use those all the time

I was looking for how much weight you use in your test. Was that 5 lb?

Your numbers are interesting. I think I will just put a cloth cap on my x brace.

_________________
http://www.Harvestmoonguitars.com

Yes, 5 lb.

_________________
Once in a while you get shown the light in the strangest of places if you look at it right - Robert Hunter