I lost the monopole...

[mqbernardo]

Strange thing (I guess) happened to me. After closing the box on a Blanca, I can't seem to find the monopole. Tea leaves start vibrating around 200 hz, in what appears to be a fuzzy monopole but it doesn't quite settle in. The figure gets into focus at about 230 Hz, but then it's a dipole that shows up. If I let the speaker working at 200 hz de monopole never quite comes into shape either. Strange as the monopole is usually pretty easy to spot and forms rather boldly. (Or it did until now )
On the top with sides(no back) and with the back clamped in place (not yet glued) I got what appeared like a monopole.

The top is on the soft side. Out of quarter on the outer parts of the lower bout. Bracing is after gerundino. I can get more data tomorrow, when I get back at the shop.

Can someone please enlighten me? Am I doing something wrong?

Thanks!
Miguel.


Some pics:

[Jim watts]

Miguel,
While others are certainly better at deciphering this than me, I suspect if you were to take a tap spectrum you would see a very broad peak around the 202 mark, or perhaps a really narrow one that you just can't dial into with your frequency generator. I believe the first scenario is most likely however. A tap spectrum should give you more information whatever the cause.

[kiwigeo]

I'm with Jim.....I wouldn't rely just on Chladni patterns.

[mqbernardo]

hey guys, thanks for chiming in.

my normal procedure is both chladni and tap tests. i don´t know what to do from them, but i still do it in the belief that stuff might coalesce in the future. anyway, this present guitar has a 0,39 g/cm3 and 5700 m/s euro spruce top and a 0,557 g/cm3 and 10 Gpa MOE mediterranenan cypress back and sides. neck is spanish cedar.

top spectrum, unmuted back

tap top.JPG (93.72 KiB) Viewed 59348 times

on this tap test, i guess the higher peak (at around 240 Hz) is the dipole, which shows up strongly in the chladni test from 218 til 240 Hz, but has a peak somewhere between 230-237 Hz (depending on where you lace the speaker). its always a dipole showing, albeit more clearly from 230 Hz up.

tap spectrum, back muted

on this one, ive tried to mute the back - you can spot the 240 Hz peak (higher one) and a 218 Hz peak close to it, which i can not visualize in the chladni tests, no matter how hard i try.

The other smaller peaks (at 150 and 167 Hz) don´t even start to shake the leaves.

this top is not very heavy and was left between 2,3 (around soundhole) and 2,0 mm (between bridge and tail). bracing is not too massive either.

sorry, only bracing image i have

thanks again,
Miguel.

[Jim watts]

Miguel, I should of noticed earlier but I see you back is only clamped on with a few clamps. I think every will change/be more defined when it's glued on.

[mqbernardo]

Jim, the clamps were just for a quick n'dirty test before gluing. All seemed nice then. The 1st pic and the tap spectra are from the closed box (Back glued). No bindings yet but I expected a monopole appearing. 1st time this happens to me. I'm just puzzled. The same thing actually happened to a friend of mine - on a very different guitar : a double fan Byers type classical. It baffled us then too.

Best,
Miguel.

[Jim watts]

Miguel,
Why is there not a strong peak for the main air frequency, does the paper rolled in the sound hole suppress it?

[mqbernardo]

Dunno. Good question. I suppose it could lower it (like a tornavoz) but maybe it's not stiff enough. I was asking myself the same question. For a number of reasons I had to leave a tiny bit of the kerfed linings proud of the edges of the sides on the lower bout. It's not a great length, but still... I wonder if it makes a difference. Strangle guitar, this one.

[mqbernardo]

To add to the strange guitar part...I don't know if many around here do the chladni stuff on free tops. I do it just as a check - no fancy voicing. Usually the mode in the pic (the "panty") shows up between 190 and 210 Hz. my lowest was 190, this one only gets to 166. It was not an heavy top, and even if not fully quartered it wasn't particullary floppy. In any case I have worked with worse tops before. Maybe it's related to the completely parallel middle 5 struts? As this was a top I savaged from my earlier days, I wasn't able to do the dynamic tests Trevor suggests in his book and I'm not geared for deflection tests.

Any help?

[Trevor Gore]

For a number of reasons I had to leave a tiny bit of the kerfed linings proud of the edges of the sides on the lower bout. It's not a great length, but still... I wonder if it makes a difference.

Yes, it can make a difference. If you have a lot of holes showing around the edges due to the kerfs, you may not be able to form a good Helmholtz resonance because the vessel "leaks" too much. I've seen this before, reported here. The air, the top, the back and the "sides" all couple, so with your set-up (e.g. clamps on the sides acting as side mass, a leaky box etc.) you may expect some weird results. In the book I talk about the coupling between the back and top via the air and the coupling via the sides being out of phase with each other. If you mess around with the nature of the coupling (leaks, clamps etc.) and the way the box is supported, it may be possible to get the effective cancellation of some modes depending on the relative frequencies of the top and back resonances.

Some woods also "leak". For example EIR can have pores right through the wood that inhibit the Helmholz effect, but everything returns to normal once there is a seal coat on the wood. I suspect everything will return to normal once you have the bindings on and have a normal closed box situation with the box "freely supported".

BTW, I wouldn't be putting too much faith in the Audacity plots unless you have Audacity set up correctly, (including a linear X axis) and have eliminated extraneous noise etc. so that the plots are easier to interpret.

[mqbernardo]

Trevor,

Thanks for chiming in. So i´ll see what happens after binding goes on. I usually do chladni tests after each stage of building the box (bare plate, uncarved and carved braces, sides on, box closed, box bound and bridge glued). This is the 1st time i noticed this behaviour.

Regarding audacity, i mainly use it to dial in the frequencies i have to look at while doing the chladnis. FWIW its set to a sample size of 16K and 22K sampling rate. for some strange reason it sometimes reverts to 512 and 44K. why should i use a linear X axis, if i may ask?

i never got VA to work on my computer.

best
Miguel.

[kiwigeo]

Usually the mode in the pic (the "panty") shows up between 190 and 210 Hz.
Any help?

[Trevor Gore]

Regarding audacity, i mainly use it to dial in the frequencies i have to look at while doing the chladnis. FWIW its set to a sample size of 16K and 22K sampling rate. for some strange reason it sometimes reverts to 512 and 44K. why should i use a linear X axis, if i may ask?

i never got VA to work on my computer.

Audacity needs setting up in a similar way to VA with the FFT sample size set to 16384 and the sampling frequency to 11025; i.e. 16K and 11K. If you set the X axis to linear, you don't get the stretching at the lower frequencies and the compression of the higher frequencies and the plots should look a lot more like the ones in the book, which makes for an easier comparison. VA should work on a PC, but doesn't work on Macs, which is why Mac users tend to use Audacity. I think most of the trouble setting up VA on a PC is getting the input assigned (external mic going to VA) and the signal levels correct.

[bluefuzz]

VA should work on a PC, but doesn't work on Macs, which is why Mac users tend to use Audacity.

Perhaps a bit off topic, but VA actually works fine under Wine on a Mac. Easily installed with Homebrew.

BTW I'm new around here so hi everyone!

Cheers
Ian

[Trevor Gore]

Good to know, Ian, and welcome. VA will also run under Parallels (a Windows emulator for Mac - which costs money) but, of course, not directly on the native Mac OS.

[kiwigeo]

Good to know, Ian, and welcome. VA will also run under Parallels (a Windows emulator for Mac - which costs money) but, of course, not directly on the native Mac OS.

I was using VA running on my Mac via Parallels and it runs fine but moving the data files around is an absolute pain in the proverbial. I ended up buying a $400 dunger HP laptop from Australia Post and I use that to run VA.

[bluefuzz]

moving the data files around is an absolute pain in the proverbial.

Yes, which is why using VA (or other Windows apps) under Wine is rather more pleasant as you can save files directly in the Mac filesystem.

Ian

[kiwigeo]

moving the data files around is an absolute pain in the proverbial.

Yes, which is why using VA (or other Windows apps) under Wine is rather more pleasant as you can save files directly in the Mac filesystem.

Ian

The other P in the A with the Mac portables is the combined mic/headphone socket. The mic input is actually more of a line input and I ended up having to use a pre-amp to boost up the mic signal. Knowing Apple the mic/headphone socket will no longer be a freature of their portables in the near future (if it hasnt already happened on the latest machines).
In the end the $400 HP laptop was a cost effective solution. The HP sits all the time in the workshop covered in dust and ready to go while my MacAir stays clean in the office

[bluefuzz]

I've only ever used an external audio interface or a USB mic. Or simply the built in microphone. New machines are so quiet that I honestly can't see any difference in tap-plots using the internal mic or a $1200 Sennheiser MD441 and pro audio interface ...

But yes, Apple can be a bit infuriating at times - although, on balance, slightly less so than Microsoft.

Ian

[mqbernardo]

Ok, Back at it. Box is bound. The behavior is still the same, i.e., I can only find a dipole near the strongest resonant peak. Didn't get the time to take pics of the chladni tests, but they're pretty much like the ones I've posted earlier. I did a quick FFT spectrum using the phone's mic. Strongest peak 236 Hz, air 98 Hz and there's a peak I can quite focus with chladni at 152 Hz.

[mqbernardo]

So, I was talking to a friend and he said that the monopole should still be there, but its frequency being closer to the dipole whenever I excite the top more on one side than the other it favors the formation of the dipole. I still get a lot of tea leaves movement when I excite the center of the lower bout.



Thanks for the info on macs and VA. I also have an old windows 7 laptop in the shop and a mac at home.


Best,
Miguel.