complete wall as a sub driver ( planar? ) ... impossible ?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Just a quick thought,
and it may has been discussed already here or elswhere, if so please link me!

What about using a very very light diaphragm of ridiculous dimension ( complete wall section of a room)
for sub bass reproduction ?

( following up on my planar wave project for my next HT , wich came up from IB setup and DBArray thread )

my quick idea would be using mylar as a diaphragm,
on a carbon fiber composite mesh that would probably maintained in correct place by some carbon composite suspension ( probably a sliding teflon lined carbon composite plate with calculated spring action )

and using similar motors as used on any subwoofer drivers, would need to think of suspension, but there would be a need to "seal" the big diaphragm on all the walls, probably some glued light weight suspension

is this possible ?
would it achieve a serious planar wave ?
if once would use a sealed Electrostatic driver of a defenite size XY, in a box that has a surface XY,
throwing on the Z axis, would it achieve a perfect planar wave ?


just to see if there was any work done on that yet

thanks all for your time :)
 
ok
i do posess this book and i don't recall this passage...i'll get intoo it later then

what other technical difficulties does it present?

mylar is usually in 1m rolls, and i can make the composite
structure to match this width without any problems...


what aboout surface irregularities due to tension
and deflection ?
non lineair control of the movement ?

room accoustic complications ?
 
dominate

"using similar motors as used on any subwoofer drivers..."

This wouldn't be a true planar-magnetic, but either like a large NXT driver (bending wave), or a large subwoofer driver (maybe with multiple motors). If bending wave, it really wouldn't work well as a subwoofer (you want some kind of pistonic action to create bass frequency wavelengths). If it is a really large driver, I would skip the mylar and just use a rigid carbon/honeycomb sandwhich.

If you were to do a planar-magnetic, it is totally doable- Maggie 20.1's play to below 25 hz and are dipoles (with the associated dipole cancellation). If you take away the dipole cancellation effect in a simulation and then simulate an infinite baffle, it works fine. Now, you would need to place them at an angle to avoid standing waves in a rectangular room. And you would need to limit excursion in the middle of the diaphragm with either anchoring points or by creating an array of smaller drivers- but it can be done. Carbon fiber would not work well for the magnet structure, as it is non-magnetic, but still conductive (and unnecessarily expensive). I would use cast iron.
 
bass array

If you build an identical wave generator on the opposite side of the room, invert the phase and delay the driving signal just right, you will acoustically "absorb" the plane wave and there will be no standing waves or modes in the room at all...assuming a rectangular room.

Unfortunately, my room isn't built this way, otherwise I would love to try this.

David
 
http://www.school-for-champions.com/science/noise_cancellation.htm

What you are talking about is active noise cancellation, which will not only cancel out the reflection, but the original wave as well. Air moving with no increase in pressure. You would probably do better to use the front and back wall in phase, with delay to have the soundwaves reach the listening position at the same time- compounding the pressure.

I'm sure that there are a number of applications for active noise cancellation in listing rooms however...
 
The method you are talking about will not deal with some of the room modes, but will tend to smooth out bass response simply due to the increase in the number of sources. The plane wave, OTOH, will also not allow floor-to-ceiling or side wall-to-side wall modes to develop...at least in theory... but with rear wall cancellation, the end-to-end modes will not develop either...and the sound in the room will be identical (or close enough) to the anechoic response. The major down side is doubling the cost of bass generators, amps, and crossover/delay units. I'd still like to try it one of these days!

David
 
We need to keep in mind with a subwoofer how long the wavelength is. At 20hz, the wavelength is ~17m (55ft) (halving with every octive). As such, if you had a room smaller than that, with identical drivers at identical volumes time alligned so that the driver at the rear cancels out the one at the front, the pressure wave never fully develops. Not like anechoic output, more like virtually no output. You could adjust the volume, (however) to only eat up what would have hit the wall, but the mechanical losses (no speaker is perfect) would make this less than ideal and very expensive (as well as eliminating boundary reinforcement, which can be a good thing) absorbtion may work better. Also, given the wavelength, you would need an enormous driver to have a directional wave at bass frequencies (probably larger than your room). Due to the wavelength, the sound wave would still be bending and you would have floor/ceiling and sidewall cancellations/reinforcement.

With bass, the more even the pressure you can create throughout the room, the better, and it is really difficult to eliminate all room interactions.
 
I'm not sure about your idea that below a certain frequency bass would tend to decrease faster than anechoic response because pressure wouldn't build in the room...in fact, this is the very concept behind the active wave cancellation theory, and is why no (or vastly weakened) modes are stimulated. Alas, the bass needs to be mono for this to work at all. Please see the link provided.

David

http://www.avsforum.com/avs-vb/showthread.php?t=837744&page=1&pp=30
 
That's pretty interesting stuff- using a mono sub with a delayed inverted signal to cancel out what would have been reflected back as a secondary reflection. I think with DSP, something like this is doable commercially. (program it in for each sub using some kind of bass management program)

At lower frequencies, what I was talking about is related to this quote from the thread you noted:
"At lower frequencies the room works as a compression chamber and the sound will not be reflected by the rear wall anymore."

With the driver at the opposite side of the room, time delayed polarity reversed, same volume, you are counteracting the pressure that the original driver is trying to place into the room.

Another way to think about it is a speaker creates sound by pressing up against a mass of air, that presseses up against another mass of air, etc., creating a wave of pressure. What you are recommending counteracts the mass of the air- if at equal volume, eliminating the sound pressure completely.

Now, you could do some sort of back wall noise cancellation to leave only a simulation of the mass of the air: creating the equivalent of of anechoic (in true planars, the mass of air is often greater than the mass of the diaphragm). However, as the diagphragm is not much more than a sheet of plastic wrap with somethin wires glued to it, it would do little to dampen any sound hitting up against it from behind- you would need to have a large, well damped room behind it. Practically speaking, it would probably be better just to leave it open.
 
"At lower frequencies the room works as a compression chamber and the sound will not be reflected by the rear wall anymore."

I think that this is so for a conventional wave in a conventional closed room, but with an active bass planar absorber, suitably delayed, etc., etc., the rear wall will essentially cease to exist acoustically, and the room will now behave like an infinitely long one with no modes at all...assuming a truly planar bass wave...and therefore, bass response will follow anechoic or can be equalized with great precision and will be basically the same throughout the room. This is one potential downside, in that theory says that the sound level will not decrease at all from one end of the room to the other, so matching levels with conventional point source or line source speakers will be very sensitive to source and listening position(s).

David
 
I got your point... You just need to be sure to compensate for the effect of the wave on the mass of the air in the infinitely long room- which is significant.

The weight of one mole of air is ~28g for ~22.4 liters at STP (for round numbers lets say 25 liters at room temp). 1000 liters are in a square meter= about 1.1kg per square meter.
http://www.engineeringtoolbox.com/molecular-mass-air-d_679.html

Complete cancelation will also cancel the sound in the room.
Compensating for air mass is accomplished primarily by reducing the volume of the cancellation drivers (also mentioned in the thread, although they justified it differently).

The big issue is that, in reality, planar drivers do a very poor job keeping sound coming from one side of the driver through it to the other- about as much as placing plastic wrap over a doorway. As such, you would need an infinite baffel to compensate- that large room behind it would do almost as good of a job transfering the sound to heat as the active system does cancelling it out. Add in mechanical losses, noise, etc., and you start to see the limitations.

However, it is still a great idea to do some sort of active cancellation... I kind of like the single sub solution at the bottom of the first page of the thread, for simplicity's sake.
 
MJL21193 : THIS IS SOO TRUE !!!
mouahhaha ..i will have to give you some credits
for the "base" idea if i ever do the moving wall sub !!!

what kind of response could i get using pneumatic air?
what could be use to drive them from the amplifier or source AC signal ??





lne937s: i will not try to cancel the rear wave...
i will try to absorb it as much as possible ...
this shall be WAYYY cheaper
and since i have no limits for WAF or other stuff like that,
i can do pretty much what i want ( 2-3' tick absorbers)

so you guys agree that having a front driver
that is "pistonic" and that would occupy most of the front wall's area, would create waves that would take care of the side and up/down modes???

adn that the sound would be pretty much even all around the room ?

"This wouldn't be a true planar-magnetic,"
i do not wish to make anything that already exists
i want to make the best subwoofer i can
for my next HT.. i have almost no limit
( only that it has to be fun to build!! !! )
 
I just had one idea...

Planar-magnetic and electrostat drivers are essentially as acoustically transparent as a metal speaker grill and a piece of plastic wrap. The featherweight drivers are about as close as you can get to having an electrical signal act directly on the air itself, with the air weighing more than the diaphragm (which is part of why people like them). Based on some measurements I've seen from people working on Maggies, I would guess that they would absorb no more than 3db of a bass frequency sound wave trying to pass through them....

So, if you took a room and cut it exactly in half with a wall sized planar-magnetic, one half of the sound wave would be projected into the listening room and the inverse would enter it's mirror image. When the waves reflect back to the membrane, they would be cancelled by the wave on the other side within ~3db passively. If the listening room was ~3db more damped than it's mirror image...

Wavelengths longer than the room would be unaffected as the membrane would still be pressurizing the room while the reflection came back...


To the original topic, I do think you would have good coverage. If using multiple dynamic voice coils on one large driver, you would need to be carefull about distribution of forces and construction. And you wouldn't need much excursion to displace massive amounts of air- 2 18" (~3.5 square feet) drivers moving 10 mm excursion would displace as much as a 12' wide wall moving .37mm (slightly larger than the thickness of a human hair). As such, a ~3mm thick foam gasket would probably be good enough as a suspension. Your driver would need to be light and rigid to avoid flexing. I would think about using as many smaller voice coils as possible spread across it- maybe from some cheap closeout drivers (excursion wouldn't need to be that big with so much surface area). You could try a planar magnetic with a heavier/more rigid/damped diaphragm, as efficiency would not have to be that great per unit of area and the forces would be spread out more evenly...
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.