Not having heard the MANGERS yet, I only wanted to interject an observation I've made relating to pistonic diaphragms in general. The conventional dynamic loudspeaker has a diaphragm that is supposed to be perfectly rigid. As this is not actually possible with known materials, some distortions are inevitable. They take the form of bending waves and nodes like the choppy waves on the surface of a swimming pool after a person does a cannon ball into it.
I believe it can be demonstrated with laser interferometry experiments that a SLOW amplifier, one that has POOR DAMPING and TRUNCATED high frequency response, will excite these modes less in a poorly designed piston-type radiator. A voice coil attached to a narrow band of material on any "cone" or other common geometry not only has to accelerate the diaphragm, but also must yank it to a halt when the sound stimulus is supposed to be decaying/ending. With a power amplifier that has a high degree of control over the motor system, the equivalent is like a MAC truck hauling a trailer hitting a concrete bridge abutment. The area of the cab will stop in a hurry, but the rest of the vehicle will do a lot of moving before it’s over. A Fostex white paper discusses rate of propagation in materials being related to internal damping possibilities in a diaphragm, and its an interesting intersection that sapphire and diamond diaphragms are at the same time stiff and high-velocity with respect to internal sound propagation. If a given amount of damping is provided by added materials like the surround, voice coil interface, and the air load, the TIME period for the decay of undesired energy from the diaphragm is shortened, thus adding to “clarity”. This does not equate with total energy, but that may not have as much bearing on psycho-acoustic perception. Such diaphragms are likely resistant to the charms of weak amplifier coupling, as they don’t “benefit” from it.
I have a great deal of respect for Mr. Pass, and watched with some appreciation his arc of experience with this issue. I am bemused by the appearance of his missing this "link" as to why a "poor" amplifier might actually sound better. Driving speakers with diaphragm designs that "freak out" when presented with an iron grip on their voice coils by a high damping factor, high rise-time, wide band solid state amplifier using high local and global negative feedback to get there may be part of the explanation. I think the "distortions" are not the vanishingly low harmonic fuzz you get from imperfect group delay in feedback servo loops, but instead the complaints of diaphragms abused by their own voice coils. A tube amp may have an output impedance EQUAL to the voice coil impedance, for a damping factor of ONE. Try this experiment some time; lightly tap on the woofer cone of a speaker with no wires connected. Then do it with a 4-8 ohm resistor across the input. Then, short the input together on the speaker terminals, and tap again. Each report from your test bell will sound different, and your finger tapping sound is EXACTLY like the decay the speaker adds right after being stimulated by musical impulses. The smooth relaxed sound so many tube and SET aficionados care for may be the avoidance of stimulating break-up modes in conventional dynamic speakers. These distortions, unlike simple harmonics, are non-mathematically correlated, and occur in time AFTER the musical signal, rather than with it. Reducing them may indeed be a technical justification for seemingly reduced performance systems sounding better, in some combinations. Conversely, solving the dynamic speaker design problem by reducing these modes in a diaphragm deliberately may allow us to move away from pretty glowing tubes, and their other limitations.
All this leads up to the Manger. Instead of pretending it is a rigid piston like dynamic drivers to, it is a thin flappy membrane carefully tuned to produce MAXIMUM POSSIBLE NODES of bending from the voice coil motion. One can think of it as the worst pistonic speaker design ever, with near-zero rigidity. As such, it need not suffer from bending modes as distortion; it is the OPERATING PRINCIPLE instead, on purpose! Like the pretty ripples in a circular pool of water when a pebble is dropped exactly in the center, it is like a cross section of pressure/rarefaction in a spherical acoustical volume around a point source radiator. Up to a certain frequency defined by the voice coil diameter and material tension, it should be hemispherically radiating irrespective of its membrane diameter. The point is, the OPPOSITE effects of damping and bandwidth from an amplifier apply in the case of the Manger. A poor damping factor will make the amp unable to fully modulate the center of the "pool" of the manger, and thus not introduce the correct counter action to the high amplitude of modulation of the diaphragm, on the DECAY of musical signals. Think of an SUV with 4-wheel drive on ice; much easier to get going than to stop. The primary damping of the Manger is the air load it plays into, besides the voice coil and periphery materials. If you have a lackadaisical amp that used to go easy on floppy cone speakers, it will have no chance to get anything meaningful out of the Manger.
I applaud Mr. Pass for doing the practical footwork that bolsters my theory. I plan to do a physical demonstrator setup with a pair of Mangers and a variable damping factor apparatus, along with a representative pistonic system alongside. By introducing several series impedances in sequence, it should be possible to DEGRADE the apparent quality of sound from a Manger, and in some ways IMPROVE it from the pistonic.
