I sent an email to Dale Harder who rebuilds classic Ohms and now manufacturers brand new A & F equivalents(HHR Exotic Speakers) with a number of modern improvements. He may be one of the only people who can properly answer that question. Hopefully, he'll be able to chime in with some good info |
Dale did chirp in on this thread earlier regarding his designs and German Physiks specifically and his comments are worth repeating here. Though their Walsh driver designs all vary, I suspect what he says regarding how each are similarly voice coil driven regardless of resulting pistonic or bending wave propogation of sound applies to the OHM CLS as well, but this is just my assumption based on what appears to be similar physics and design elements in play in the case of each. Dale said: "Bending wave schmendig wave piston schmiston its just a matter of semantics at best. So what if the GP vc is elastomerically connected or mine is hard fixed. I can do that with silicone. Bottom line, the vc still controls the movement of the cone and the cone is launched forward in a bending wave fashion or a pistonic fashion depending upon the wavelength. Additionally, the mass of the vc controls the upper end and transient response. GP still uses an Aluminum edge wound vc same as mine. And both drivers are made of Titanium foils exactly the same size. T/S parameters still apply as well." IMHO, each vendors Walsh design varies based on targeted market. GP targets the very high end audio market. Dale Harder is a champion of the concepts of Lincoln Walsh's driver and is a champion of carrying the original OHM designs into the future with a focus on quality and value. OHM has always and continues to focus on providing the best sound possible as cost effectively as possible to the masses. |
Hello Darkmoebius, thanks for the invite to add my two cents. Sorry this has taken so long. After a few email and many days... While this is "some-what of a complicated subject", I dont believe there is a Direct answer, but rather a conglomeration of processes taking place that influences the final outcome. So, in a rather lengthy fashion, bear with me as we try to pick this apart. I would say that a standard crossover being either passive or active in nature is obviously more of an electrical device acting directly upon the input signal rather than a mechanical device, though to be fair it can also impart a mechanical signature as well. That being said, the electronic crossover by nature, IMO, tends to directly interact with the incoming signal and adds or subtracts its influence over a broad range of frequencies, and *in most cases these can be herd* as a distinct change in the original signal. The original frequencies may also be altered by adding odd order harmonics and distortion components and are then passed along down the chain. In the case of the Walsh driver or the DDD driver the electrical input signal has no crossover elements to deal with and therefore remains as true as possible to the original signal. Obviously the voice coil has some losses, however the resistive losses and the conversion losses to a mechanical movement *if implemented correctly* will not alter the signal characteristics, but simply tend to produce a replica of the original albeit somewhat smaller, i.e. less amplitude. There are other factors as well, but we shall ignore those for the sake of brevity. Now the rub, the Walsh and the DDD drivers actually have up to 4 separate modes of behavior all happening at one time. The lowest frequency range of its operation can be described and predicted by the use of the Thiele/Small Parameters. (This is useful in designing the proper cabinetry to mount the drivers) The next frequency range takes place up to the Coincidence frequency. (The Coincidence Frequency is best described as the frequency at which the velocity of the wave traveling through the cone reaches that of sound in air, or approximately 1120 ft / second.) It is this band where the cone tends to move mostly in the pistonic rage. The next range is where Pistonic movement is progressively replaced by Bending Waves. Eventually the range progresses to where the cone is functioning entirely as a bending wave converter. Remember, all of these ranges are overlapping in nature and do not have a very defined band or point that is distinct by nature or equivalent to just a couple frequencies. It is due to this dispersion and the shape of the cone that the Coincidence frequency range is distributed over a broad frequency range rather than occurring at a single frequency. Let me interject for a moment regarding the Walsh driver and the three distinct metal, paper bands. A common misconception is that the low, med and high frequencies all boil off the cone only on the related areas, i.e., the high from the Titanium, the mids from the Aluminum and the lows from the paper. This is actually not the case. The actuality of fact is that the wavelength determines where the wave will exit the cone and it also determines which of the four modes mentioned above the waves are functioning in. The segmented metal/paper cone sections were really derived to offer a substantial change in material density and stiffness, which in turn controls the speed with which the traveling wave moves through the material. It was also an attempt to control modal patterning. Finally, moving beyond the Bending mode of operation we come to the range where the first standing waves begin, and this is to be avoided if possible as this is where modal breakup and loss of cone control begins. OK, now that I have expressed that entire load, you can see that there are an extreme number of factors taking place in this type of driver. Now, in the course of testing a driver such as these, we may feed in signals consisting of sine waves and or square waves (hahaha, just try that with a piston driver) and try to get an idea of the frequency response of the driver or of the overall system consisting usually of multiple drivers. The real rub here is that music is a complex signal consisting of many waveforms and harmonics and is anything but a sine wave. Now, consider this, the many hundreds of waveforms being generated that may make up a single instrument or note are actually boiling off the cone at different places all at the same relative time. Since they are not all exiting the cone surface at the same place and since theoretically they have not been altered in any fashion it would be extremely hard to *hear* a crossover point or region. (BTW, the waves reach speeds of up to 5 times the speed of sound traveling through the cone) And so it is, when we look at a frequency response curve of a Walsh or DDD driver we see for the most part a continuous curve. When we look at a multiple driver system however, we see many curves all interrelated with their crossover points and the regions these respective drivers cover on either side of these crossover points. When I do a test of a Walsh driver using a sine wave generator and slowly look at the waveforms from 20 Hz to 20 KHz there are usually no points in the curve that correspond to say the actual crossing of the paper or metal boundaries. Most of the nonlinearities occur from actual cone construction techniques and these are usually minimal if done correctly. In fact, this is a great way for me to find defects and correct them or to hand tune the cone for best linearity. Now if a cone is made of one single material then there are no physical material changes to act as varying points of density or stiffness, but other methods must be employed to accomplish the same end result. Hmmm, come to think of it, the Walsh style driver does not suffer from Doppler induced distortions either. It has been said that it is possible to hear the crossover point or region on the DDD system because it uses an augmented low end with a standard piston style driver and external crossover. This I can well imagine, but in all fairness I have not heard it for myself. Here is some further food for thought. The very highly acclaimed and very expensive (50K+ and 200K) MBL 101E and the 101 Extremes are one of the few drivers utilizing true coherent 360 degree sound much as the Walsh and DDD. However, these systems utilize FOUR separate drivers with four crossovers. In addition, the drivers are either stacked or separate as in the case of the subs for the Extremes. Now imagine the crossover influence that must be present there vs. a single driver with no crossovers, not even mechanical. So the long and short of it gentlemen is that I have yet to hear or perceive a crossover induced distortion on a True Walsh Driver. PART II Back to the original question.... "wouldn't the transition from a wave-bending mode to a pistonic mode in a single driver result in an audible "crossover" effect, although perhaps much less audible than a well-managed crossover between two separate drivers?"/_ -- could you explain if there actually should be a *theoretical* crossover effect at the transition point vs your first-hand inability to hear one. After talking with DM on a personal e-mail he prompted me to add the following... In order to better answer the original question.... What I believe is happening is that folks are "getting hung-up on the idea that there are actually transitions in the modes of operation." Here again, remember we are not dealing with sine waves but rather complex waves and harmonics at any one give time. While one part of a wave may be indeed transitioning a boundary layer per say, I rater doubt that all of the waves are at one moment. Further, just because the wave is transitioning from say a bending wave motion to a pistoic motion, does not mean that its waveform or sound is changing. The wave is still boiling off the surface of the driver cone in the same fashion and exiting the cone at some finite angle relative to the wavelength. It is still leaving the cone and producing a wave pattern similar to the stone making waves in a pond. The wave is not dependent upon the driver moving air forward or back as in a piston cone of say a regular 12" cone. It is these ripples that are coming from the piston driver moving air in a forward or rearward wavefront that are affected by Doppler shift and beaming. The Walsh style cone does not suffer from either of these characteristics. So, IMHO, I do not believe that you would be able to hear any crossover shift under normal circumstances. I hope this helps to shed some light on the subject gentlemen. Good listening to all. Dale |
Dale, thanks so much for your (one of the most I have seen) detailed accounting of how the Walsh driver works. I've always considered the Walsh driver a truly a unique and vastly under-appreciated piece of engineering work IMHO despite my limited understanding of the physics and engineering behind it. I'm glad there are people like you, Strohbeen and the others out there devoted to keeping it alive and moving forward to the benefit of music lovers everywhere! |
07-22-09: Mapman Dittos |
