John Curl's Blowtorch preamplifier part II

[gpapag]

George,
I was wondering myself why you couldn't have something on the former that would give absolute position, something that was outside the gap above it where you could measure absolute position?

-LVDT .
-Eddy current position sensor
-Capacitive position sensor
-IR transeiver translating to distance.

I mentioned accelerometer because
Displacement can be derived from acceleration and time data.
True, displacement is not always the same as distance travelled but the three- axis acceleration data can be used to extract true distance travelled along a desired axis.
Relative position data then translates to absolute position if a position reference is provided.
This reference is coil’s rest position which can be sensed and stored during system initialisation or at zero-crossing of a cal sinusoidal excitation signal. (*)

Brad is mastering all these. I guess that his objection is on the routing of the MEMS wiring (sensor on the cone).
I agree, this is an issue with production units.
An alternative would be to place the MEMS on the driver’s chassis. Calibrating for the reaction forces there and linearizing for the parasitic vibrations is a major SW undertaking but the bulk of programming will be done once, adjustments then performed for each new speaker model.

George

(*)Using the later and increasing the amplitude, zero crossing will register the dynamic shift of the coil’s rest position by tracking any asymmetry of the cone fore-aft excursion caused due to either driver’s suspension or the air-load’s asymmetry at the two sides of the cone (enclosure’s air loading)

 

[scott wurcer]

-LVDT .
-Eddy current position sensor
-Capacitive position sensor
-IR transeiver translating to distance.

Why not just 3D doppler laser and a few G-flops of DSP? You should see how much (tech and $$$) the biggies are throwing at the enhanced gaming experience.

 

[snup]

 

[bcarso]

-LVDT .
-Eddy current position sensor
-Capacitive position sensor
-IR transeiver translating to distance.

I mentioned accelerometer because
Displacement can be derived from acceleration and time data.
True, displacement is not always the same as distance travelled but the three- axis acceleration data can be used to extract true distance travelled along a desired axis.
Relative position data then translates to absolute position if a position reference is provided.
This reference is coil’s rest position which can be sensed and stored during system initialisation or at zero-crossing of a cal sinusoidal excitation signal. (*)

Brad is mastering all these. I guess that his objection is on the routing of the MEMS wiring (sensor on the cone).
I agree, this is an issue with production units.
An alternative would be to place the MEMS on the driver’s chassis. Calibrating for the reaction forces there and linearizing for the parasitic vibrations is a major SW undertaking but the bulk of programming will be done once, adjustments then performed for each new speaker model.

George

(*)Using the later and increasing the amplitude, zero crossing will register the dynamic shift of the coil’s rest position by tracking any asymmetry of the cone fore-aft excursion caused due to either driver’s suspension or the air-load’s asymmetry at the two sides of the cone (enclosure’s air loading)

Many are the ways. What I have in mind adds no mass to the voice coil/cone and would have a miniscule mechanical interaction with them, although it is not optical. The drawback is that for many existing transducers it wouldn't work well at all. So the driver must be of a certain sort, or constructed with certain details incorporated into the motor.

 

[snup]

or a mirror on the dustcap

 

[snup]

and a laserdiode pointed at it

 

[gpapag]

Why not just 3D doppler laser and a few G-flops of DSP? You should see how much (tech and $$$) the biggies are throwing at the enhanced gaming experience.

I don’t think they have to invest heavily in audio but they surely will invest for to enhance the experience of the other senses.
3D visualization is one of the target technologies.
Yes, G-flops of DSP and a lot of complex coding is required today, but if you’ve followed the progress of image holographic reconstruction from the beginnings (remember the NU Museum of Holography? See what is going on at med advanced remote operation rooms today or some secure tech locations), you know how far we have traveled and how close we are now to a realistic outcome for the masses, without the hassle, inconvenience, expertise ($$$$) and time for setting-up and maintaining the older technology.


George

>Edit

Many are the ways. What I have in mind

I have faith in you.

 

[simon7000]

I would limit the thermal mass but at least using a finned heat sink on the cold side. The thermal mass of aluminum is half of copper, so aluminum is good there. would use a fan inside too. A good heat sink on the hot side is desirable too. As you lower the dt both hot and cold, you get more -Q for the buck. Yes they are not linear. http://www.marlow.com/media/marlow/product/downloads/pl054-6-40-01ls/PL054-6-40%20Pre-production.pdf

The link shows this in a similar size TEC by Marlow, you can see the run away region. If you load down the TEC too much, thermal mass, you will be prone to run away.

Run this program, you can find a similar TEC in the database. pretty accurate, well help with transient response. for simple shapes. as you get closer I can run an FEA for you.

AZTEC Software Download | LairdTech

Thanks. From my tests I was under the impression that if I limit the voltage and have close to an infinite heat sink I can avoid runaway. Very easy to have a massive fan cooled heat sink in my design. Have about 25' of 5" wide stuff. Used to be a motor controller manufacturer used the same aluminum distributor I use. They paid for the custom dies and then went out of business. So I got some of the leftovers at $.85 per pound.

 

[snup]

if the mirror is on the inside of the phaseplug, one can shine a light at it, through the whole motorassembly from the back of the magnet. How big a mod can that be ? And you don't touch anything, while measuring excurtion.

 

[Kindhornman]

snup,
Do you mean to shine the light on the back of the dust cap, the phase plug wouldn't be moving, what would you be measuring?

 

[snup]

That vent can be used for measuring all sorts of things going on in the center of the motor still without touching. (pressure, temperature, volumeflux....etc.) while venting and cooling at the same time. Why touch it ?

 

[snup]

You will get inacurate measurements and raising weight of voicecoil if you try to put a measuringcoil on it to drag around. You can just have a coil around the polepiece for that.

 

[canyoncruz]

Check out this paper. For a woofer I would use metallic pain or gold leaf if something lighter were needed.

 

[snup]

That file I am not cleared to see.
Not qualified, sorry.

 

[canyoncruz]

If you guys are serious I could borrow an optical phase demodulator. This paper was done a while ago. I would use a C-lens which is a fiber fused to a collimating lens. I actually have a few on hand. The collimated beam size is about .8 mm

 

[snup]

A hero turns up !
Out of the blue, how convenient.
How can one get ones grubby mits on one of those ?

 

[canyoncruz]

The erbium source and filter can be replaced by a super-luminescent source. I have some from my manufacturing days with a coherence length of 5 mm at 1300 nm. the detector is straight forward, but you would probably have to build the transamp for it. couples are now crazy cheap.

 

[canyoncruz]

I have worked with opticphase. We all worked at Litton G&CS ( now Northrup Grumman) in the fiber optic gyro and acoustic sensor group. I was the 5th guy hired. I would have to ask to borrow.

 

[snup]

A genuine expert, then !
borrow one for me, please.

 

[31697B]

.
.
.
You are never too old to
try something stupid.