ah cool
some aspects remind me of a "air motion transformer" Air Motion Transformer - Wikipedia, the free encyclopedia
some aspects remind me of a "air motion transformer" Air Motion Transformer - Wikipedia, the free encyclopedia
Problems: air leakage, pump noise, exact matching of amp and speaker, high impedance voice coil, voice coil heating. If you try to make a sealed loudspeaker unit then it would have to be quite large to get enough bass.
This idea is daft enough that there is bound to be a US patent on it, probably from the 1950's or 60's.
This idea is daft enough that there is bound to be a US patent on it, probably from the 1950's or 60's.
Reluctant as I am to sound a note of sanity in this thread, I feel I ought to point out that the proposed scheme does not magically remove the requirement for the driver to be mounted in an enclosure that suits its parameters. In other words, "Hoffman's Iron Law" still applies.
The rest of the scheme is quite valid, if impractical.
However, I would suggest that the system be designed so that the pressure in the enclosure is lower than the surrounding air, to take advantage of the change in effective volume. In other words, the lower the pressure inside the enclosure, the smaller the enclosure can be. It won't make a huge difference at the pressure differentials likely in a "practical" system, but every little bit helps. The "Holy Grail" of this design would be the ability to have a full vacuum in the enclosure, in which case the enclosure need only be large enough to surround the rear of the driver. Just need a way to stop the minor problem of the cone imploding... say a 12 inch driver with Sd of about 500 cm2, that's something like 500 kg (over 1100 lbs) of force on the cone...
The rest of the scheme is quite valid, if impractical.
However, I would suggest that the system be designed so that the pressure in the enclosure is lower than the surrounding air, to take advantage of the change in effective volume. In other words, the lower the pressure inside the enclosure, the smaller the enclosure can be. It won't make a huge difference at the pressure differentials likely in a "practical" system, but every little bit helps. The "Holy Grail" of this design would be the ability to have a full vacuum in the enclosure, in which case the enclosure need only be large enough to surround the rear of the driver. Just need a way to stop the minor problem of the cone imploding... say a 12 inch driver with Sd of about 500 cm2, that's something like 500 kg (over 1100 lbs) of force on the cone...
"Problems: air leakage, pump noise, exact matching of amp and speaker, high impedance voice coil, voice coil heating. If you try to make a sealed loudspeaker unit then it would have to be quite large to get enough bass."
Welll, if it doesn't leak or returns to equilibrium during the off state, then leakage or pump noise isn't a problem. The voice coil does not need to be any higher impedance here than usual, there is still an output transformer, in fact that's the whole point of this exercise to neutralize the DC in that so that no air gap is needed in it (for SE operation).
The voice coil heating from DC current is a real and near killer issue though. The only way I see to control that would be to make the voice coil/magnet assembly a controlled air leak over the coil with a fan to pressurize the assembly and provide air flow. One could run a 2 inch flex air hose to the speaker from a remote location to eliminate the fan noise, but that's not too practical. The inclusion of a fan does at least allow variation of the pressure so effectively the DC bias requirement could be adjusted to match an amplifier DC bias requirement. There is also ferro-fluid that could be put in the magnet/VC gap for heat conduction to the magnet assembly.
-----------------
The large rear containment issue I believe goes away here. There is no radiated rear wave, we are converting that to back pressure which simply replaces the inductor energy storage of a SE amplifier's gapped OT or parafeed's B+ inductor. The back pressure build up is required and equivalent (we are not fighting it, it is helping us return the speaker cone forcefully, just like the now missing half of a P-P amp would. The equilibrium back pressure need only be equivalent to 1/2 peak sound pressure normally developed by the speaker at Max output). ( note: There is no electrical signal here, like in a P-P amp, to push the speaker cone back.) This does end up looking like an inductor electrically, so it produces AC "magnetizing" current just like a real inductor. But since the OT being used here has no air gap, we have just replaced one inductor with another (equivalent). (well, no air gap P-P OTs do have inductance, but with a very high L compared to SE ones, so they draw minimal magnetizing current)
---------
"However, I would suggest that the system be designed so that the pressure in the enclosure is lower than the surrounding air, to take advantage of the change in effective volume. "
The approach of using a partially evacuated chamber is interesting and should be equivalent as far as the SE DC offset requirement goes. The DC bias would simply move the cone from one extension limit to the other extension limit and produce the required back pressure drop. When not in use, the cone would return to the 1st extension limit with zero pressure differential and a tiny air leak could restore any lost air volume to the "sealed" chamber.
Cooling of the VC still remains a show stopper though. Perhaps a ferro-fluid in the Magnet/VC gap could transfer the heat to the magnet frame. Still a substantial amount of heat to dissipate. Lets say the VC has a 2 Ohm DC resistance, and the DC bias (equal to peak maximum AC signal) is 6 amps. Then I^2R = 6*6*2 => 72 Watts!! And thats continuous (class A mode), even with no sound output. Sound signal AC losses just pile even more onto that.
Anatolyi's idea of applying this to an electrostat is looking quite attractive about now. That would only require a DC voltage bias, which could come off the tube plate directly. No OT at all, but still SE operation. Not sure how or if that can work though without an inductor somewhere.
------------
"I feel I ought to point out that the proposed scheme does not magically remove the requirement for the driver to be mounted in an enclosure that suits its parameters. In other words, "Hoffman's Iron Law" still applies. "
See above on the back pressure utilization, but also, we are under the opposite constraint here actually. The chamber length must be small enough so that no radiated wave (dynamic pressure differential along the length of the chamber) can develop. We need to avoid resonance effects by keeping the length dimension (x2) UNDER a 1/4 wavelength. This does conveniently eliminate the speaker cabinet. A full vacuum would also solve the back wave problem, but as you said, rather impractical. Perhaps an easily compressible vapor could substitue for vacuum.
Welll, if it doesn't leak or returns to equilibrium during the off state, then leakage or pump noise isn't a problem. The voice coil does not need to be any higher impedance here than usual, there is still an output transformer, in fact that's the whole point of this exercise to neutralize the DC in that so that no air gap is needed in it (for SE operation).
The voice coil heating from DC current is a real and near killer issue though. The only way I see to control that would be to make the voice coil/magnet assembly a controlled air leak over the coil with a fan to pressurize the assembly and provide air flow. One could run a 2 inch flex air hose to the speaker from a remote location to eliminate the fan noise, but that's not too practical. The inclusion of a fan does at least allow variation of the pressure so effectively the DC bias requirement could be adjusted to match an amplifier DC bias requirement. There is also ferro-fluid that could be put in the magnet/VC gap for heat conduction to the magnet assembly.
-----------------
The large rear containment issue I believe goes away here. There is no radiated rear wave, we are converting that to back pressure which simply replaces the inductor energy storage of a SE amplifier's gapped OT or parafeed's B+ inductor. The back pressure build up is required and equivalent (we are not fighting it, it is helping us return the speaker cone forcefully, just like the now missing half of a P-P amp would. The equilibrium back pressure need only be equivalent to 1/2 peak sound pressure normally developed by the speaker at Max output). ( note: There is no electrical signal here, like in a P-P amp, to push the speaker cone back.) This does end up looking like an inductor electrically, so it produces AC "magnetizing" current just like a real inductor. But since the OT being used here has no air gap, we have just replaced one inductor with another (equivalent). (well, no air gap P-P OTs do have inductance, but with a very high L compared to SE ones, so they draw minimal magnetizing current)
---------
"However, I would suggest that the system be designed so that the pressure in the enclosure is lower than the surrounding air, to take advantage of the change in effective volume. "
The approach of using a partially evacuated chamber is interesting and should be equivalent as far as the SE DC offset requirement goes. The DC bias would simply move the cone from one extension limit to the other extension limit and produce the required back pressure drop. When not in use, the cone would return to the 1st extension limit with zero pressure differential and a tiny air leak could restore any lost air volume to the "sealed" chamber.
Cooling of the VC still remains a show stopper though. Perhaps a ferro-fluid in the Magnet/VC gap could transfer the heat to the magnet frame. Still a substantial amount of heat to dissipate. Lets say the VC has a 2 Ohm DC resistance, and the DC bias (equal to peak maximum AC signal) is 6 amps. Then I^2R = 6*6*2 => 72 Watts!! And thats continuous (class A mode), even with no sound output. Sound signal AC losses just pile even more onto that.
Anatolyi's idea of applying this to an electrostat is looking quite attractive about now. That would only require a DC voltage bias, which could come off the tube plate directly. No OT at all, but still SE operation. Not sure how or if that can work though without an inductor somewhere.
------------
"I feel I ought to point out that the proposed scheme does not magically remove the requirement for the driver to be mounted in an enclosure that suits its parameters. In other words, "Hoffman's Iron Law" still applies. "
See above on the back pressure utilization, but also, we are under the opposite constraint here actually. The chamber length must be small enough so that no radiated wave (dynamic pressure differential along the length of the chamber) can develop. We need to avoid resonance effects by keeping the length dimension (x2) UNDER a 1/4 wavelength. This does conveniently eliminate the speaker cabinet. A full vacuum would also solve the back wave problem, but as you said, rather impractical. Perhaps an easily compressible vapor could substitue for vacuum.
Last edited:
An alternative cooling scheme for the voice coil comes to mind. The VC/magnet does not have to be in the sealed chamber, that can be on the other side of the cone. So forced air flow thru the VC is possible without affecting the sealed chamber. Still there's the issue of a quiet air flow source though.
A more distant solution to VC heating would be a superconducting VC. More likely to materialize (in say 10 years) though would be a graphene ribbon VC. Very low loss and very light weight too. Even the diaphragm could be graphene possibly. A single broadband speaker maybe. Of course once we have graphene ribbons, all kinds of odd speakers will be possible.
A more distant solution to VC heating would be a superconducting VC. More likely to materialize (in say 10 years) though would be a graphene ribbon VC. Very low loss and very light weight too. Even the diaphragm could be graphene possibly. A single broadband speaker maybe. Of course once we have graphene ribbons, all kinds of odd speakers will be possible.
Last edited:
What about elimination all transformers, coils, all together?
Make a vacuum tube, triode, with flexible radiating anode.
Make a vacuum tube, triode, with flexible radiating anode.
Too bad nobody is going to hear it, since it's a bell in an evacuated bell jar............😕
No, it is a jar from which an air had been evacuated.
Think about a capacitance manometer. Similar, but with filament & grid.
- Status
- Not open for further replies.