Acoustical Impedance Tube

It sounds like the usual story of the Prof and phd student know SFA about the subject they are working with. I've met prof's with egos bigger than their common sense

I think you need a clear definition of who is driving this project.
I think you have 3 choices.

1 Get on and do what you know works. (assuming you are driving it).
2 Play along build whatever they ask you to and let them dig themselves a big hole.
3 Walk away from the whole sorry mess

I dare say it may be possible to measure the electrical impedance of the loudspeaker as it's driving the different loads, same as it may be possible to drive your car whilst sitting in the back seat, using a couple of sticks, buy why would you want to?
It seems nonsensical to exclude the use of micophones from audio measurements.
You have my sympathy, as it sounds like you know your stuff.

Del

...'tis not wise to argue with the person(s) who "sign-off" on your thesis or dissertation!

Well put! Great paper too.

I've been out of the field for over 10 years, so can't add anything.

Great post, Redfred.

Reverse the polarity of the speaker

What would this do?

converts the speaker to a microphone

But doing that is like switching the polarity on an incandescent light bulb. It doesn't do anything

Doesn't do anything? Doesn't DO anything?

Reversing the polarity on an incandescent light bulb causes it to absorb dark!

^{_{There, I feel much better now.}}

_Thanks!

Oh, that. Use a working bulb.

"Well their first idea was to use the speaker somehow to find the resonance of the tube (which I showed you can't do)"

Hi,

A.:

if you want to use a speaker as a detector (converting small vibratory velocities into induced voltages) then you need a coil with very many windings.

This is derived from the relation that the ratio of force/current equals the ratio of induced Voltage/velocity.

You get a very small velocity to detect and the speakers have a low force/current ratio as designed for 4 Ohm resistance.

So if you can a second speaker to be modified as a detector by changing the number of turns of the coil to between 1000 to 10,000 then this may work depending on the velocity you want to detect.

This is the sensor type that is usually used in balancing machines. To be operated with a high impedance - low noise amplifier, if high precision is needed at lock-in

B.:

Measuring damping by frequency change is ridiculous. Look to the slope of the phase curve near resonance. This is the one and only to evaluate. Phase-steepness (Phasen-Steilheit) equals the Quality factor of an oscillator (may be a factor of 2 to be introduced, I have to look if important.) We measured very small damping coefficient of clamped high quality flexures by this in the low damping region (Q< 0.001) and damping of gas-bearing gaps (Q> 10) in the high damping region.

C.: (but even then only by a few Hz)

If you resolve to µHz this may be a pretty big signal!

D.: (The damping material inside a well designed enclosure makes very little difference in the electrical properties)

It makes a big difference if you measure damping only! But make sure there is no serious nonlinearity where the damping may occur in excited higher harmonics and not seen in the phase curve at the fundamental resonance.

E.: Recommend to the PHD student a visit to his local library (and the Prof. too) to read more about oscillators. Old books of Grammel and Magnus to start with. If necessary I can give them a tutorial. (Only 400Km from Munich).

RHABE

If you want to use a speaker as a detector ... then you need a coil with very many windings.
So if you can a second speaker to be modified as a detector .

Yeh, but this is the exact bonkersness of the whole thing...

Such a device is called a MICROPHONE.
Del _{<Arggghhh Yowl yrawlllll mrawwwww FTZZZZZZTTTT>}

I agree,

some advantage: you can modify the sensitivity - much space available for copper (or Al) windings,

some disadvantage: waste your time and money by gaining experience, bulky, needs suited amplifier ...

RHABE

My cats are faster than yours!

Cool thanks everyone for the advice. I'm really glad that Del agrees with me, I remember when you first joined, I know how smart you are lol. I've been on a bit longer but haven't done much posting, just reading.

I'm actually a graduate engineer. I finished last spring. I seem to have a really good practical understanding of the engineering world (for example I designed a full scale continuously variable transmission for a Porsche track car from design to prototype for my senior design project, all in 10 weeks lol).

I'm just here "helping out". My school asked me "hey, you want to study in germany?" So I was like, "Yeah, Sure I mean I'll already be there for my internship at BMW in Munich." I'm just here for fun, I don't get credit for it.

So enough bragging... down to business

So today I continued to argue with the PhD student and the prof but by some miracle I finally convinced them they were wrong!!! lol. It only took 8 weeks (and I've only been here 8 weeks )

Redfred, thanks for your advice. That is very good info. So it definitely is theoretically possible. And I did use the exact same method by Elliot Sound Products. I remember trying to use that method when I first started college with not very good equipment and got really odd numbers

But this time I have extremely good equipment, even besides the $75k thingamajig. So, to think things through, I need to move the length of the tube to the resonance of the speaker. Unfortunately, since the resonance frequency is ~75 hz (it's a small 4" driver) the tube would have to be 2.28 meters long (since it's effectively a double closed tube, it hast to be half wave instead of 1/4 wave). So the closest I can get is about 115 hz if I capped the end.

And I like your description of the idealized model using just RLC components. I've seen it before but somehow never thought about how the parallel higher value resistor represents the effective resistance of moving the air. It shows the scale between the speaker impedance and the air impedance.

Sorry for taking so long to reply back. I But I will read that other article you sent as well and maybe try something to see what happens.

Also, RHABE, about "A" although I think your idea would work it seems like it would be rather difficult. We have lots of microphones already. And "B" I'm really glad you agree as well. I couldn't believe they wanted to measure damping by frequency shift, and that's just from my basic knowledge from my circuits and mechanical vibrations classes. "C", I suppose that would be correct lol. "D" That also makes me feel good that my intuition about the setup was correct. Now I know I wasn't doing anything too technically wrong for trying to measure any electrical differences because they would be hard to detect. "E" well I think it's unlikely that I could get them to read any books lol. I found several doctoral thesis publications on the subject, read them all yet wasn't able to get them to read them. I even highlighted key sections lol. But I just realized that I haven't updated my location recently. I'm unfortunately not in Munich anymore, I'm in the city of Jena (Thüringen). Where are you with respect to Jena?

That project sounds simuliar to the Bose Wave system.

I have an electro-magnet speaker (post WWII) sitting here, about 10" across, so it may be your speaker-microphone combination ?

Pulse it and then listen, like a sonar or radar ?

I also have a Baldwin speaker (WW I vintage).

A bigger speaker would be more sensitive, but then I'd need a tube the same size diameter. I was actually going to try a pulse then listen technique. Basically play white noise for a short time and then suddenly turn it off. Then read the signal from the speaker. But when you use the speaker as a microphone, you might as well use an actual microphone anyways.

Hi Nick,

I've found this thread by chance: I was looking for some people trying to do the same measurements I'm planning to do (with a much much cheaper instrumentation).

As far as I know somone else did (quite) the same experiment: Martin J. King and Robert A. Robinson.

Me, I had some idea to extend their work and I was planning an appropriate impedance tube to do the job.

I would like to better understand your difficulties (they could become mine), help you in some way and get some work done for me.

Can we continue this discussion offline ? I could share some calculations and simulations that are difficult to share on the pages of a forum.

Thanks

Teodoro (teodorom@hotmail.com)