Official Luthiers Forum!

Merry Christmas!
For those who test to deflection on an unbraced top am I correct in the following process: provided the given deflection is what I’m shooting for.

1: roughly cut out top
2: place a parallel under the neck and end block area of the top on question
3: add a predetermined weight to the center of where the x brace will be located (say 5lbs)
4: measure the (sag / deflection) : I think the video I saw a while back shot for 1/4”

I ask the following because I’m considering a top material other than Spruce, and I’m not familiar with how it should “feel”

Thanks for insight
B


Sent from my iPhone using Tapatalk Pro

Merry Christmas B!

I always thickness my top wood to .140 prior to testing so as to level the playing field on material thickness.

Other than what you already covered, specific and constant spacing between supports, specific weight, etc. that's it.

Now the difficult part is building the database so that what your results are is meaningful. Or, in other words data from several tops won't mean much but data from say 50 tops becomes more useful.

Lots of us have done deflection testing differently and there is no right or wrong here just follow the scientific method, be consistent in your set-up, weights, etc. and understand in advance that you are setting out on an effort that will take some time and many tops before it's meaningful.

Lastly and something that was very important to me. When I started building my life was full of specific metrics for this and that and the other thing. I always wanted to be the sort of Luthier, the romantic version who can flex a top and know from experience at least to some degree what to expect. With this said be sure to take the time to smell the roses and do some subjective flexing too along the way.

One of our friends here and a highly contributing member in the past, Mario P. once talked about carrying around a small block of wood in his pocket to over time understand the feel of the wood. No jokes please.... But I did this too and still have the two blocks of spruce and BRW that I carried around in my pocket for about three years. Every day these little sugar cubes of wood went with me. The TSA didn't appreciate it either but this is what I did. Don't know it it helped at all but I suspect that there was a calming influence not unlike worry beads....

My approach is a bit different.

1. I do not cut out the basic guitar shape. I prepare the joint for gluing and rip the two halves down to 8-1/2" before gluing up. I'm not likely to ever build a guitar wider that 17". I prefer to leave the top square as what I'm trying to do is compare the quality of the top material. In my opinion cutting the general shape wouldn't provide me with a useful comparison.

2. After sanding the top face to 150 I take the plate to my thickness sander. My test rig has supports that are 18" apart. I use a piece of brass hex bar (doesn't tend to roll) weighing about 4.5 pounds. I then thickness the top to my target deflection. I find It's a good idea to flip the top and test both sides and to also test once the top is at room temp. Also leave room for final sanding. I usually stop at about 80% of my target deflection and finish up using my RO sander. It's so easy to overshoot!!

3. Once that's done I cut out the basic shape and move on to the rosette and sound hole.

Happy holiday!

M

What Michael said, plus one more point.

BE CONSISTENT!

The gauge and weight in the same place every time. The supports in the same place every time (lock them down on a piece of ply and make it a permanent jig).

Otherwise your data is useless.

Good luck

Steve

I do essentially the same as Mike too. And it bears repeating, it's essential that your deflection test, however you decided to set it up, is done exactly the same every time thereafter so that numbers are comparable across tops. Same span between supports, same weight (meaning the same object used as a weight as well as the same mass), same point of measurement on the top, same position of the weight on the top, etc. You want the test to be as standardized as possible.

I started doing this only four guitars ago so I'm just starting to build the data set. I was taught to thin the top, as a first approximation, to the point that it starts "oil canning" when you shake it. I've been paying particular attention to what the deflection values are when that point is reached as a potential predictor for future guitar tops.

In case it might be of help, I've been doing an initial measurement when the top is around 0.140 - 1.150" thick. I then thin to about 0.130 and take another measurement. Repeat at 0.120+. All the while checking the top by tapping and shaking it between measurements. Then refine the thickness until I'm satisfied with how the top oil cans and how it feels when bending it along the grain and take a final measurement. At this point, the deflection measurements are not driving what the final thickness ends up being. I'm just recording them for now until I have enough data points to see how deflection values correlate with with how the top behaves. I'm sure there are several other approaches being used and you may come up with an approach that works better for you.

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

Here's a couple pics of my test rig. I can't remember but I think I saw something like it in the Gore books and came up with something similar.

I read about how Brian Howard does it and pretty much copied that method with a few minor differences as follows.

1. Use a brick as a weight and it weighs approx 2.5kg's iirc.
2. Joint the top and keep it as square/rectangular as possible and install the rosette (do not cut out sound hole).
3. Use a hand plane to thickness the top on the inside of the guitar to about .15 inch or so then draft out the top bracing pattern.
4. Draw two lines at right angle to the center line tangent to the head and tail block curves and place two triangle shaped slats under the top right along those lines.
5. Place the brick on the bridge patch with the dial caliper on the center of the X. With the weight on a zero out the dial caliper then when I take the weight off I measure the deflection.
6. Use a thickness sander on the outside of the top, the side with the rosette installed, to get to the deflection I want and keep the inside of the top freshly planed smooth.


The last guitar I built I was experimenting with a very light top and thick sides and a live back. My deflection was .5in which is quite a bit. My current guitar, an 000, is at .25in.

With all this information I should be able to calculate Young's Modulous but I just have not gotten around to doing that yet. I think YM is a good value to have when comparing deflection testing with others. For me to say I put a brick on and it deflects .4in is not very useful or not AS useful anyway.

I'm only about ten guitars in doing deflection testing but it's already becoming a valuable tool.

BTW once the top is braced I will do another deflection test. This time with little blocks placed at the end of each X-Brace.

YM!
Something for me to research today!


Sent from my iPhone using Tapatalk Pro

Really useful information here. Thanks for making the ask, Brandon.

_________________
Insta - https://www.instagram.com/cbcguitars/
Facebook - https://www.facebook.com/cbcguitars

This is what I have been doing the last 6-7 years and have found it extremely helpful for consistency. At first I noted the deflection when I thought the tap tone was right, later use the critiques of the finished guitars to hone in on what deflection seemed to give the best sounding instruments based on the response of players I respected and my own ear.

Of course top thickness is only one part of the total equation but it's an important one.

I have the top cut out, rosette installed, and sound hole cut out. Thickness is usually around .135 range. I put it on the frame in a standard fashion and picked a spot 3" below the sound hole in the centerline for measurement. Pre-weight with 1 lb, zero the dial gauge and add an additional lb. ( I use old orthopedic traction weight from the hospital) Repeat several times and average.

Then over to the thickness sander and start taking wood off the back side of the top as the rosette has already been leveled. With this technique I like to see around 0.043-0.050 deflection. On an OM that usually works out to around 0.114-0.118 thickness.

My two cents.

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

With all this information I can’t wait to get time to make a dedicated jig!!


Sent from my iPhone using Tapatalk Pro

I suppose, provided you knew you were looking for a specific deflection, you could simply make a block the same height as the end supports and subtract what you want to deflect from the height. Then, just thin it until it just touches the block.


Sent from my iPhone using Tapatalk Pro

Like so...


Only horizontal haha


Sent from my iPhone using Tapatalk Pro

Now .... another question.... how does one know how heavy a weight to use? Or is it just a trial and error, ... although one must know where to start .....


Sent from my iPhone using Tapatalk Pro

Is this a magnetic base??


Sent from my iPhone using Tapatalk Pro

Yes it is, if it was right side up you would see a switch on it. Works like a charm. I like this set up because it's simple and stores away easily.

I’ll say that I found the pressure from a downward facing probe deflected my tops when I ran a bunch through just to develop some baselines. If you could arrange for the top to just kiss the top of an upward facing needle, results would be more definite.


Sent from my iPhone using Tapatalk

The downward facing probe on my setup does that too, but that little bit of deflection doesn't really matter since it's the same same probe/gauge each time and it's zeroed out the same way and everything else used to make the measurement is constant too. With the deflection gauge zeroed, all the deflection measured thereafter to is due only to the added standard test weight. It's the same as taring a scale.

On the other hand, if the gauge couldn't be consistently zeroed because of a sticky probe, that would be a problem.

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

On the other hand the upward facing probe will be influenced by the top sagging a bit too. I just lift the upward facing probe and then bring it down to touch the top and get the measurement. Like Rocher said, as long as you are consistent and do it the same way every time then it will start to make sense to you. You will find it actually takes a few seconds for the top to come to rest and in fact you will find that different times of the day or different days of the week (for what ever reasons) will make small changes. It's full of systematic errors but as long as you own those and take care of them the same way it will still be meaningful. Where is it not so much meaningful is in a discussion like this because of all the variability. That's why I think obtaining Youngs Modulus is a real point of discussion, but then again one has to obtain that accurately too.

With the upward facing probe I found pre-weighting the top with 1lb to stabilize the interface to be essential for repeatable readings.

Zero the caliper after the pre weight and add additional weight for the actual reading.

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

The only thing that matters, is to use the same weight every time on your system.

_________________
Pat