Frequency response curve, advise please.

You can ask questions here about Trevor and Gerard's exciting new book on Luthiery.

Moderators: kiwigeo, Jeremy D

Post Reply
Jorge Bertholdo
Wandoo
Posts: 13
Joined: Thu Aug 22, 2013 9:41 am
Location: Porto - Portugal
Contact:

Frequency response curve, advise please.

Post by Jorge Bertholdo » Tue Aug 26, 2014 6:48 am

Fleta1408_coup.jpg
My first tap test just after closing the box is as in image Fleta1408_coup.jpg as the blue line:

What I can see so far is:

1. T(1,1)2 at 215Hz is too hight;
2. The peak at 257Hz I assume is T(1,1)3

After a few sand job on the back (red line):

3. The T(1,1)2 drop a little to 212Hz
4. The assumed T(1,1)3 "disappear"

At first I though the peak at 249Hz was the T(1,1)3. But it have no change after the sanding, so I conclude that it isn't. To make sure that is the case, I make another tap test on the uncoupled top and back:
Fleta1408_back_top_main.jpg
5. The B(1,1) is at 238Hz
6. The T(1,1) is at 205Hz

So, I ask for some advise on what to do next:

7. Should I sand the borders of the top, to make the frequency drop?
8. Why T(1,1)3 is not so prominent like seen on the book and the others posted here? Is that because it is only 3 semitones higher from the top? B(1,1) at 238Hz doesn't seems it is too stiff.

I don't if that is useful at the moment but, if I did the math right, the "monopole mobility" is 40.37 (s/Kgx10-3) and the "flexural Rigidity" (EI) is 26.32 Nm2
Guitar, Lut and Violin maker

http://www.jorgebertholdo.com

jeffhigh
Blackwood
Posts: 1536
Joined: Thu Feb 07, 2008 5:50 am
Location: Caves Beach, NSW
Contact:

Re: Frequency response curve, advise please.

Post by jeffhigh » Tue Aug 26, 2014 7:25 am

Hi Jorge,
At this stage, I'd be installing the binding, applying the finish and installing the bridge, saddle and strings (get the guitar complete and strung)
All these things will have an effect on the finished resonances.
Tapping the top once you have done that will give you a guide to what you need to do
Making modifications now is too risky
The Fleta has a 4 brace back which may be too rigid to give a T(1,1)3 regardless

Jorge Bertholdo
Wandoo
Posts: 13
Joined: Thu Aug 22, 2013 9:41 am
Location: Porto - Portugal
Contact:

Re: Frequency response curve, advise please.

Post by Jorge Bertholdo » Tue Aug 26, 2014 7:40 am

Thanks jeff for your reply,

I was think about doing exactly what you say. By the way, I choose to do the back active, with the radial brace system. I don't know if it work well with fleta, but we'll see.
Guitar, Lut and Violin maker

http://www.jorgebertholdo.com

User avatar
woodrat
Blackwood
Posts: 1155
Joined: Tue Nov 25, 2008 6:31 am
Location: Hastings River, NSW.
Contact:

Re: Frequency response curve, advise please.

Post by woodrat » Tue Aug 26, 2014 12:19 pm

Hello Jorge...Welcome to the forum....I couldn't add anything more to what Jeff said as it is good advice....I just wanted to say Welcome to the Forum. I checked out your website....Very nice work...:)

Also I wanted to ask you what program you used to develop the frequency response curve? It looks different to Visual Analyzer.

Thanks

John
"It's never too late to be what you might have been " - George Eliot

Jorge Bertholdo
Wandoo
Posts: 13
Joined: Thu Aug 22, 2013 9:41 am
Location: Porto - Portugal
Contact:

Re: Frequency response curve, advise please.

Post by Jorge Bertholdo » Tue Aug 26, 2014 5:11 pm

woodrat wrote:Hello Jorge...Welcome to the forum....I couldn't add anything more to what Jeff said as it is good advice....I just wanted to say Welcome to the Forum. I checked out your website....Very nice work...:)

Also I wanted to ask you what program you used to develop the frequency response curve? It looks different to Visual Analyzer.

Thanks

John
Thanks john,

The web site is still a working in progress, I just did so far the violin side, I'm organizing the photos of the guitar works to post on the site. Stay tuned.

The software actually it is Visual Analyzer, exported spectra as text and plotted in a chart in LibreOffice (or Exxel). Thats how Trevor did in the book several times when needed to show 2 curves or more for comparison. What is different is those "check box", that take me a lot of work and headache.

* I discover that when I hide the column where the data is stored the curve disappear. I don't know if this is a bug or it is supposed to happen this way.
* The check box and the chart is 2 independent entities
* a cell can be programed to show the output of the check box (on/off, 0/1, show/hide, etc whatever you want)
* I record a macro to hide the data column, one for each curve
* make each macro to be used as Function
* in another cell I use the "IF" function to compare the check box output cell and call the appropriate macro. "=IF(Data.B2="hide",COUP2HIDE(),COUP2SHOW())". here also one "IF" cell for each curve of the chart.

I was looking for a solution to have several curves in a chart, this could be very confusing if one have 15 curves, for example, but be able to choose only 2 to visualize, like SpectraPlus do, witch is a payed software (very well payed).

cheers

Jorge
Guitar, Lut and Violin maker

http://www.jorgebertholdo.com

User avatar
woodrat
Blackwood
Posts: 1155
Joined: Tue Nov 25, 2008 6:31 am
Location: Hastings River, NSW.
Contact:

Re: Frequency response curve, advise please.

Post by woodrat » Tue Aug 26, 2014 10:50 pm

Thanks Jorge...:)
"It's never too late to be what you might have been " - George Eliot

User avatar
Trevor Gore
Blackwood
Posts: 1605
Joined: Mon Jun 20, 2011 8:11 pm

Re: Frequency response curve, advise please.

Post by Trevor Gore » Wed Aug 27, 2014 9:47 am

Jorge, there will be significant changes to modal resonances as you proceed through the build. The most significant change will be the addition of the bridge so it is worth spending a bit of time attempting to predict what will happen when you glue it on. Simulating the effect of mass is quite easy, but simulating the effect of the added stiffness that the bridge gives is not. Section 4.6.1.3 covers this and has a chart that you may be able to make some predictions from. If you add some mass to the top in the bridge postion (say 15grm or whatever your bridge mass is) and then measure the T(1,1)2 you will be able to see how much the T(1,1)2 drops. If it doesn't drop to your target (say 190Hz) you can be pretty sure you'll need to make the top less stiff, because the bridge itself will add stiffness (as well as mass) and so the T(1,1)2 will not drop as far. At this stage, I wouldn't do any more to the back as that can be easily tuned if necessary when the guitar is finished and it is much easier to make it less stiff than to make it stiffer.

Jorge Bertholdo
Wandoo
Posts: 13
Joined: Thu Aug 22, 2013 9:41 am
Location: Porto - Portugal
Contact:

Re: Frequency response curve, advise please.

Post by Jorge Bertholdo » Fri Aug 29, 2014 7:12 am

Thanks jeffhigh, woodrat and Trevor,

Besides the information in the Section 4.6.1.3, the fig. 2.4-5 also give some prediction.

I'm done with the box for the moment, I'll moving with the bindings and I'll be back to tunning at late stage.

cheers
Jorge
Guitar, Lut and Violin maker

http://www.jorgebertholdo.com

Post Reply

Who is online

Users browsing this forum: No registered users and 11 guests