Official Luthiers Forum!

So I've built my deflection testing jig and everything is tickety-boo as far as that goes. Except of course that I have no real reference numbers to compare to. I know it all depends on build style and how you shape the braces and how stiff they are and a myriad other things but I would like some thoughts on the data I have collected on two tops I am using for my next two guitars. They are numbers 50 & 51, so I am using woods a cut above (and of course more expensive) than I usually do as they are mini-milestones and I would like them to turn out great. I hope the following numbers will mean more to some of you than they do to me and that you'll be willing to share your thoughts...any ideas are welcome and appreciated.
Thanks

The deflection is measured along the grain at an 18" span, the weight is 5lb. I didn't do cross grain 'cause I'm pretty sure 5lb would snap 'em.
Top 1
Bearclaw Sitka
L20.5" x W16.5" x D.115", = 38.89 in^3 if I've done the math right. Weight 11.55oz (327g)
Deflects .172
Will be a Dread, feels too stiff.

Top 2
Adirondack
20" x 16.5" X .115, = 37.95 in^3. Weight 8.8oz (250g)
Deflects .175
Will be an OM, feels too flimsy.
I was surprised that the two tops weighing so dramatically differently for their sizes would deflect so closely.

If your measurements are very accurate, then it seems that Sitka is has a very high density at about 512Kg/m3. I am not that surprised it doesn't come up with much better deflection than a much lighter plate of more reasonable density. My own measurements show that the very dense spruce pieces are inefficient in terms of stiffness to weight, although I do not think they can't be made into great guitars.

It seems the hands can feel the density somehow as well (if not just weight alone), not just stiffness. Maybe that is why you feel the light piece is flimsier.

What I suggest is to join and test several more pieces. They will be fine as long a you have a reasonably good humidity range. I must have some 15 tops (good, not packing) that are cut to shape and tested in all imaginable ways.

_________________
Build log

The way that I see this the numbers you have mean little in an of themselves. I use deflection as a way to measure real effects of a gradual thinning of the top. A .005 reduction in thickness can cause a .020 change in deflection. My set up uses a steel bar weighing about 2 pounds. I place the bar across the grain and shoot for a deflection of .120. (average of both sides) I did this for two tops yesterday. One was at .105 thick and the other .101. I feel that this is a more consistent way to judge stiffness vs. just thinning to a certain number or "feeling" it. It can be a long time between tops and remembering how "wobbly" a top was from time to time is a challenge. I do not get any more out of the exercise than that. My previous number was a deflection of .100. I am gradually making my tops have a higher and higher deflection to try and see where just right is. Or when things blow apart.

_________________
http://www.booneguitars.com
https://www.facebook.com/pages/Stephen-Boone-guitar-builder/488208541257210

I am not sure if this is going to help you, but I am leaning toward the less dense tops. In a perfect world this is around 6 gm/in^3. I am using a 5lb weight for deflection and I am targeting a deflection of .20" over an 18" span.

I rarely pay attention to how thick the plate is until I hit the deflection and then measure it.

_________________
Brock Poling
Columbus, Ohio
http://www.polingguitars.com

Can't help you on finding numbers but I would advise you to somehow take final sanding into account. Going from 80 grit on a drum sander to 220 finished, particularly if you sand both sides, can make a pretty big difference. It makes more of a difference with thinner tops but it is still worth being aware of.

Also, plates freshly warmed by the drum sander will flex more.

Probably the main thing is to check it the same way every time if you want your data to be accurate.

_________________
http://www.chassonguitars.com

Thanks,
These are all good answers. Indeed 'feel' is tricky, exactly the reason to be a bit more scientific about it.
I have top 1 being 8.4 g/in^3, top 2 being 6.56 g in^3. So then, obviously top 2 has a much better stiffness to weight ratio.
Now, let's say I wanted to use Brock's target deflection numbers. How do I use these numbers to calculate how much wood to remove to achieve desired deflection. I know it has to do with the cube rule, but I don't know how to apply it. As I see it, I'm about 175/200th's of the way there, or 87.5%. So I want to make it 12.5% bendier. But I'm sure if I ran it through the sander until it was 12.5% thinner it would be much more than 12.5% bendier. I know there is a formula for this.
I suppose it would be instructive to take it down .001 at a time and measure along the way, but my deflection jig is at home and the very nice TimeSaver is somewhere else, and carting delicate tonewoods back and forth on a crowded city bus is sketchy. And again, I know there is a formula for this...
So, any mathematicians out there?
Thanks

does anybody "go by feel" with these things?

I think the purpose of measuring things in this way is to help you quantify what you feel, not necessarily to remove or replace feel from the process. Let's say you wanted to emulate a guitar that you built a few years ago. Are you really going to remember how that top felt? Wouldn't it be nice to have a little data on hand to see exactly what you did? And feel is obviously quite subjective (which is not necessarily bad). To use my current scenario, one top feels too stiff, one top feels a little flimsy, and yet they both measure out to nearly the same deflection. Ok then, apparently I'm confused about my feelings.

Why would you think it's too stiff for a dread but too flimsy for OM?
Also be careful here because top thickness works in conjunction with bracing so it's reall hard to tell someone how much a raw plate should deflect, it's only part of the design.
Also, deflection set ups can vary widely, does the weight just set in the middle of the plate or does a beam span the width, how big is the area being deflected, and where are you measureing the deflection. Unless you've duplicated the set up from Brock, his number mean nothing to you but mean alot to him. IMO

_________________
Jim Watts
http://jameswattsguitars.com

Jim, I agree with you about the variances in testing jigs and all the other variables as well. I'm just looking for a place to start and how to use the information in a relevant way. And I have two different tops I'm working with.
My weight is an empty BruichLaddich tin (sadly the bottle is also empty) filled with pennies to weigh 5lbs. You can place it either upright or sideways which definitely yields different numbers. For this post I've been using it upright at the center of the panel.

As I say, I am not sure my numbers are much good for anyone else.

I think the real value in deflection is that it helps get you to a standard stiffness from guitar to guitar. I never worry about how much more there is to take off as I work toward my target deflection. I just sand a little, test, sand a little more, test, etc. until I hit my number. So I am not doing cube rule calculations. (I suppose you could, but it isn't necessary).

You're also right that plate deflections can/should vary based on your bracing, but I think the differences would be very minor.

The problem with this approach is it is hard for people to compare results. As you point out, unless we are all using the exact same methodology comparisons are tricky.

But it is a good tool IMO to invest some time in. I just offered my numbers as a place to start investigating. YMMV.

_________________
Brock Poling
Columbus, Ohio
http://www.polingguitars.com

I totally agree with you Brock, I deflect mine to get a handle on guitar to guitar. It's has the same issues Chldni testing has IMO. I think after you,ve testing a bunch of guitars with your method it can help.

_________________
Jim Watts
http://jameswattsguitars.com

I apologize for the partial highjack here - For those of you who thickness to deflection, how do you account for the plates's taper in thickness or do you not taper it? I have wondered about this. I do taper towards the edges and prefer the Chladni method for that reason but I wondered how other people who use deflection factor it in.

_________________
Burton
http://www.legeytinstruments.com
Brookline, MA.

Don't worry Brock, if I'm not happy with the results I'm not going to blame you! As I said I'm just looking for a reasonable place to start and information on how to apply the data. The only way to truly dial it in is of course trial and error. I really wish I'd started on this 30 guitars ago. It's not as though it's difficult to get these measurements. Ah, well.
Anyhow, can anyone tell me how many thousands of an inch to remove to reduce Stiffness by 12.5%? And can you show the formula?
Thanks

If you can put some numbers on things; the size and thickness of the plate, unsupported span, and mass of the weight, you can calculate the Young's modulus of the wood. Then you have a number you can compare with other folks' numbers, even if the rigs are different. You still have to correlate that with the way _you_ build, of course. That takes some time.

Yes, bit it's going to be involved though, since you're not measuring each axis independantly by setting a weight in the middle of the plate due the anisotropic properties of the wood.
Meddlingfool, you might look into David Hurds' book Left Brain Lutherie", I believe he measures his tops this way and calls it average modulus or something like that.

_________________
Jim Watts
http://jameswattsguitars.com

Even if you got every bit of wood involved to deflect precisely the same they would still sound different from guitar to guitar because of all the myriad variables involved. For me, the exercise is not to make them all sound the same but to develop a methodology to help me control how they sound, intentionally.

Yup you read me right. At 50 guitars I am not yet a master luthier! I still think I've got plenty to learn. The line between good, really good, and outstanding is pretty fine.

Fillipo Morelli asked:
"Why would one want all one's guitars the same?"

Who says they all have to be the same just because you measure things? As has been pointed out, there are enough variables in this that even when you do measure stuff they don't turn out 'the same'. I have yet to make a 'matched pair' that sound exactly alike. The point is to be able to make what you want to make, and do it as consistently as possible, given the vagaries of wood. I'm assuming you measure when you cut frets; why not just do that by ear?

If I'm reading your results correctly the along the grain stiffness between the 2 tops is really close.......but one "feels to stiff" and the other "feels too flimsy".


OK, here's my take. You're measuring stiffness along the grain, you're probably feeling the stiffness across the grain, reguardless of how you're holding and flexing it. I measure both, (and attempt to feel both) but with methods that are probably different than anyone elses. I do believe stiffness in both directions matters, for both tone and structural integrity. I certainly don't have a secret forumla or any magic process.
FWIW, you're working on 50 and 51, I'm working on 28 and 29.




Yep, It took me a while to figure that out.

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

Meddling Fool: Think you have misnamed yourself....!!Your no fool at all....!! But rather a smart man who after 50 plus guitars is honest and realizes the real complexity of varying all the factors to be able to get the sound you want. Thanks so much for being candid.
Tom

_________________
A person who has never made a mistake has never made anything!!!

Woody B, yes the cross grain stiffness of the Adi top is much lower. I haven't measured it because I think 5 lbs would snap the tops so I'll reduce the weight and see what happens...
Somehow this thread had tangentalised ( huh? ) er, veered off into the philosophical which isn't what I'm looking for. No offense meant, but I'm looking for something else, namely
1) some general target numbers to go with just as a starting point, from people who use this technique, after which I'll use my own judgement to home in on what sound I'm trying to achieve, and
2)the mathematical equation for finding out how to reduce my plate stiffness in this case 12.5%, which I believe has something to do with the cube rule, but I don't know how to apply it...
Thanks everyone.

If it is any help to you I use the exact same numbers as Brock. Span 18" weight 5lb.weight and .200 deflection. The weight is a round plate with the hole in the centre,which allows me to centre the plate and put the dial indicator shaft in the centre of the hole.Have only done this on the last three guitars so not a lot of feedback yet. Further development would be to do deflection readings on braced plates but don't know of anyone doing that yet. No even sure this would be of any help due to the increased stiffness and low deflection.
Tom

_________________
A person who has never made a mistake has never made anything!!!

Back to the practical (or at least how I do it). I don't worry too much about the cross grain stiffness. Different woods (or even pieces of the same species) are all over the board on this. If I am building and looking for the cross dipole to be strong I either reject the floppy tops or compensate the bracing to add some cross grain stiffness.

To me the biggest things are density and long grain stiffness.

_________________
Brock Poling
Columbus, Ohio
http://www.polingguitars.com

Meddling Fool: I check by feel for cross grain,but because I use the round plate instead of a bar or something else across the full width, I think my readings take into consideration both long and cross grain stiffness. No proof, just my thinking.
Tom

_________________
A person who has never made a mistake has never made anything!!!