Laser Instead of an Accelerometer? Is it possible ?

[jsa_ind]

Dear Forr,

I believe what you inferred is what Rythmik Audio is offering. I too came away with the similar inference.

Though I would have expected a little more clarity in http://www.rythmikaudio.com/servo_tech.htm

Thanks,

Best regards,

Junia.

 

[jsa_ind]

Hey guys,

What do you think of this http://isp.imm.dtu.dk/nonlincomp/Klippel.pdf ?

So is lazor an option ?

Thanks,

Junia.

 

[soongsc]

Klippel uses the laser in front, if you can find a way to hide the laser, the you got a goood concept. Post#3

 

[BassAwdyO]

my idea

I have pondered this idea for some time.... I have my own idea

It uses an optical laser similar to one found in a CD player and a reflective strip similar to the encoded film on a CD

The laser reads position in the same way a CD laser reads information. Of course the feedback would have to be digital, so its probably out of the realm of DIY

 

[BassAwdyO]

A pic... if the attachment works!

 

[soongsc]

A pic... if the attachment works!

CD reads the data by bit in series. I wonder how it can be done with this setup?

 

[Visor57]

Hi there!

Funny enough, I was shooting the idea around a few days ago and concluded that the better way to use a laser to measure cone movement/acceleration is to employ the laser measuring via doppler effect.

See Brüel & Kjær.

Of course cost is a problem with this approach, but it is possible and very accurate without the need to modify the loudspeaker diaphram.

Now, if only I could get my paws on one of those meters...

 

[SY]

There's an interesting way of doing this that I've been rattling around in my head, but have no time to pursue: interferometry. Place a reflector somewhere on the back of the cone. Pass a laser beam through a beam splitter. One beam reflects off a shiny spot on the back of the cone; it is then recombined with the second beam at a photodetector. As the cone moves, the interference fringes will look like pulses at the photodetector. Counting pulses will allow you to measure displacement versus time; differentiating twice (trivially easy) gets you acceleration.

One small complication is the need for a second reference beam in quadrature to the main one to allow you to know if the cone is moving backward or forward. But that's just an extra photodetector and mirror.

 

[Ron E]

Differentiation is noisy. Any noise in the sensor loop causes big swings in the signal. You can highpass, but then you have more phase bugaboos....

 

[SY]

It should be done digitally. Counting pulses makes the noise irrelevant.

 

[poobah]

I wonder how a doppler system would pencil out. A fixed frequency light emitter on the cone and then some detector on the basket. You would need a high carrier freq to maintain decent resolution through the discriminater... but you would pick up one differentiation for "free".

 

[mrshow4u]

I used to work with a guy named Christopher Struck. He did some work with B&K using a laser interferrometer. Poobah, he described their setup as you did. ...an FM discriminator system. I even think B&K used an off-the-shelf FM IF+discriminator. ..they just had a different front-end/mixer affair. I think there was an AES paper on it. I'll see if I've got it laying around.

 

[poobah]

I did the calcs for cone displacement ages ago. Don't remember anything concrete except for the fact that the displacement can be insanely small for a tweeter at low volume... I mean like yoctometers or something.

That perhaps is not very "reasonable". I think we could argue that the woofers need the most attention.... because we have to "assist" them in so many ways as it is.

Anybody here current on the calcs? Just pick reasonable numbers: 6" driver, 400 Hz, at say 65 dB SPL. What's the peak to peak peak displacement?



And yes Mr.show... that would be an interesting read.

 

[mrshow4u]

.....this isn't what I was looking for, but...
<http://www.bksv.com/pdf/Bo0268.pdf>

I'm gonna try to give Chris Struck a call today and see if he'll give an e version of his paper or give me the pub number.

 

[mrshow4u]

I talked to Cris Struck. He's going to send me a re-print of his AES paper. He cautioned me that this was about 10 years ago and that there has been many improvements since his paper. He mentioned Wolfgang Klippel from Dresden company almost has the market cornered on this type of equipment.
www.klippel.com



.....I'll post Chris's paper when I get it.

 

[poobah]

Cool...

That Kippel link goes somewhere weird...

 

[SY]

Try www.klippel.de to get what you're looking for.

 

[mrshow4u]

.......Oops! my bad www.klippel.de

 

[tschrama]

SY 's idea seems to most practical to me and offers a real accuracy .. but wouldn't the double differentiating lead to a time-lag and eventially instability?

... offcourse you need alot or digital proseccing, but a PC can be programmed to do that.. the 1/4wavelength can had by using 1/4lamba insert lenses.. they sbould be easely availeble ....

I once did a bit of research into accoustical feedback , which seems far more simple to fabricate, but also suffers from time-lag stability issues due to the path-legnth from speaker to mic. But I remmember a certain N. Pass having a patent on that ....?


BTW
ultimately it is not the cone-movement you're interested in, but the sound waves it produces.. due to cone-breakup, they are not the same...

 

[SY]

If it makes any difference to you, my idea has been done (not in audio, but in IR interferometry), and it worked quite well. We had to move a heavy mirror back and forth along a linear bearing and know its position extremely precisely. Typical throw was a centimeter, and we would run at 10-80 Hz, selectable. What we ended up doing was very speaker-like: a magnet and a voice coil at the heart of it, but we used an air bearing. The HeNe laser was set up with a small outrigger mirror to control the main interferometer. In those days (1985), it took several large circuit boards full of parts to pull this off, but that was the Stone Age. Interestingly, the amp I ended up using for the mirror drive was the LM1875, the daddy of modern audio chip amps.

Here's a pic of the instrument, though nothing showing the mirror, I'm afraid:

http://www.gps.caltech.edu/options/geochem/Instruments/Infrared_Spectroscopy/Nicolet60SX.html

If I can dig up one of the old brochures on it, there were some nice photos that make things pretty clear.