I will respond to some of your points if I can, Basement. First off, I myself have no direct experience with using fluid damping (yet - I expect to before very long), so all of my information comes from what I have gathered on the subject from various sources, and a lot of my assumptions in the above posts also stem from nothing more formal than just my own common good sense and scientific/mechanical aptitude/intuition (which, believe me, is merely a little better than average, and not the result of extensive education or training). In short, I am no expert.
Your supposition about out-of-phase movement of the cantilever relative to the cartridge body/tonearm, and its causing of positive amplitude errors, is interesting. But do you know for a fact that it is true? I would liken this view of the situation to pumping a swingset for increased altitude, or the cracking of a whip. The other view would be that of trying to push a tackling dummy on a muddy field and having your feet travel the other direction instead - it would result in less motion of the dummy. We could call the first example the "whip-crack" model of relative motion as it affects signal amplitude, and the second example the "traction" model.
The traction model seems fairly straightforward - any unintended deflection of the arm causes a signal undershoot. The whip-crack model would be a good deal more complex. When attempting to push the tackling dummy in the mud, your result will be similar no matter when you engage in pushing. But when cracking a whip or pumping a swing, timing is everything. Get the frequency wrong, and you'll get negative acceleration, not positive. So presumably, the whip-crack model would often show a similar result to the traction model, and only sometimes result in signal overshoot, depending on the conditions of groove frequency, velocity, and amplitude. I do not know which, if either (or even both, under differing dynamic circumstances), of these possibilities is valid.
As far as the Townsend damper design goes (something this thread has enlightened me to), I would surmise the reason he moved the damper out the arm's radius close to the cartridge is because that is where the arm's motion will be greatest in amplitude and velocity, relative to the same damper installed closer to the pivot point as is common. This is very ingenious, as it both increases the effectiveness of the fluid's damping properties, and simultaneously eliminates the cartridge's leverage advantage vs. the normal arrangement, while also ridding the cartridge/damper interface of several inches of potentially resonant and flexible intervening armtube length. It does sound ungainly as hell though (and scary to boot - who wants all that viscous fluid above the record surface?)
I'm not sure I understand the reasoning for the conclusion you draw in your last sentence, since you don't seem to have provided an explanation of why you feel this to be the case (mass being superior to fluid). I would comment regarding your observation about the increase in average arm mass, that this has to be taken in the context of average cartridge compliance - these two parameters evolve hand-in-hand as they must, but it is the compliance issue that leads the dance. My own feeling, as it has developed over the course of this thread, and for reasons I explained in my previous posts, is that a theoretically ideal system would probably be infinitely low in mass in both planes, and hence inertia-less in all directions, and entirely resistively damped, whether by fluid or some equivalent. Of course such is not possible, but it does tend to point away from increased-mass solutions - not that they won't be effective in some ways as well, and maybe even just as good for practical purposes, though I am somewhat skeptical of this right now. |
Ok, we have a couple of suppositions here, and I will try to address them as I see them.
First, it seems we all agree that the headshell/arm should position the cartridge directly centered over the groove, and should maintain that position as it plays the entire record.
Next, we all seem to agree that it is possible for the cartridge compliance to move the arm out of the ideal position, especially if the arm has insufficient lateral mass, or no fluid damping.
Now, our contentions are that with either increased lateral mass, or fluid damping, this problem can be mitigated/eliminated. But, both solutions introduce their own potential problems which need to be dealt with.
It initially seems as though the fluid damping creates less inhererent possiblity of problems, because it does not have a mass that could be set into motion, and thus go out of control. However, the problems introduced with the fluid damping system, as I see them, are these: The paddle must be set into some level of accellerated motion before it can begin to work, therefore it will allow some arm motion to occur before the unwanted movement becomes damped. If the arm does move before damping occurs, then it has to move back, and could be slowed by the damping during its return to the center. This creates a bell-shaped amplitude curve regarding its movement away from center and back to center. Granted, this will be a small amplitude, but it will be there none-the-less, due to the requirement for the arm to laterally accellerate to a certain speed before damping occurs. So therefore I conclude that fluid damping has static/low amplitude limitations, and works best in higher amplitude/ high acceleration conditions. And, in all cases, allows some unwanted movement to occur before coming into play, as it is a dynamic system.
Increased lateral mass has what appears to be more difficult inherent problems associated with its use in this application. This increased lateral mass can be accellerated out of control, if the mass is insufficient to perform its intended task. Momentum would take over and create mayhem. Return to center under these conditions would be out of the question, as the arm would be flying across the record surface. However, there is something here that works in favor of increased lateral mass. It is a static system that raises the "moment of inertia" to a point where the cartridge compliance cannot overcome the static moment of inertia. In this case, the cartridge is stabilized over the groove center, with no movement needed, or allowed, to work. So the stabilization, if sufficient, can be complete, with no lateral accelleration(movement)of the arm needed to bring the stabilizer into play.
So, what we have here, as I see it, are a dynamic control system, and a static control system. The dynamic system requires some movement of the arm to work, but then comes into play very aggressively to limit movement. The static system doesn't require any movement to work, and in fact cannot allow it, or it will spin out of control from momentum. Or at best, create a nasty "swinging door" effect, which we definitely don't want.
Now, my assessment of these systems, is that we really don't want to allow any deviation of the arm from the center of the groove, so that the stylus will do all the moving, not the arm. So, from an absolute performance point-of-view, the static system of increased lateral mass allows no deviation, assuming sufficient mass. But, if the mass is insufficient, or if some unforseen large accelleration enters the system that can overcome the static moment of inertia, it can have a major disruptive effect. The dynamic system, while allowing a small degree of deviation from the ideal, will control any large accelerations very well, and will never get out of control(except at the very low accelerations, which it does not control at all, due to the "damping threshold").
So, where do we want to go with this? Do we allow some movement, and then quickly stop it, or do we allow no motion, and possibly get out of control if a large force enters the system? Or is there another idea, or combination of ideas that would better resolve the problem without causing additional ones?
There are some other issues that enter into this discussion also, and they are, how are the other aspects of arm function affected by these mods? There is no question that the added mass system can cause a change in vibrational modes that may or may not be benign. The fluid damping does not have this characteristic. It has no potential problems in this area, and even may tend to damp some larger vibrational modes. It is less "cartridge dependant" for its correct operation. On the other hand, if the cartridge is selected with all the correct capabilities(low compliance), then the increased mass of the static system, may enhance the vibrational and resonant modes to not need any damping, or need less damping. And if lead is used, especially in the proper locations, damping may occur by the lead material itself. I believe that this is happening on my tonearm, with this mod. Another benefit that I believe I am getting from this is the reduction of bearing chatter, due to the increase of the static "chatter threshold" by placing the lead weights directly on the bearing axle. This greatly increases the mass of the axle, and is much less likely to chatter in the bearing, because it is too heavy for the vibrations to excite/move it in the bearing races. This does not come into play with a unipivot, obviously, since the unipivot has tons of PSI on the pivot tip already. But interestingly, the added mass(PSI) on the tip, is what causes the unipivot to be chatterless.
I could ramble on about this, but please give your comments on what I've said so far. |
Very well broken down, Twl. I agree about the fundamental aims you suggest. Your analysis of the probable dynamic response limitation of a fluid damper reinforces the reasoning behind Mr. Townsend's implementation I was expounding on above.
But I must point out a flaw, or at least a simplification, in your characterization of mass-damping as a "static" system. Assuming negligable bearing friction (and this might not always necessarily be the case, but for now I will assume that a premium bearing's friction will fall below the level where it would play a larger role in dynamics), the horizontal mass of the tonearm does not entirely resist the arm's deflection by the cantilever, as you state, but conforms to the basic principle 'for every action, there is an equal and opposite reaction'. In other words, the motion of the stylus will produce deflection in the cartridge/tonearm which is inversely proportional increasing mass of the cart/arm (and also to increasing cart compliance), but this delfection will not = zero, just as in the fluid-damped example. (Had it = zero, you then would have been implying that there existed some 'mass threshold' for deflection, which your unmodified arm hadn't yet crossed, but that had been crossed once you added the extra mass. Of course, and again assuming bearing friction isn't mucking the system up at this point, this isn't the case; there is no 'mass threshold' - it is a continuum, and you have simply moved down the scale in the direction of less deflection.)
What the levels of allowed deflection would actually be in each example of these respective damping systems is unknown to me, and I would guess is likely to remain so. Obviously, in each case there are many variables to be played with that would affect the answer. To achieve an optimum balance of dynamical attributes for any given cart/arm combination, the fluid system allows for relatively easy adjustment via modification of the fluid level, so the mass-damped system should probably feature a method whereby the user can either change the amount of applied mass, or more likely adjust its distance from the pivot point, like a common counterweight adjuster, either by sliding or helical means. (Twl, did I just suggest a major price increase? :-) |
The effect of the cantilever deflection in the reverse is a real thing. I've read about it and observed it. When you get your fluid damping devise (where are you getting it, by the way, so we may obtain our own), you may be able to observe it as well. I am blessed with near sightedness, and if I may pause to brag, I can count the seven strands or litz wire in cardas. To make it easier to observe, I get my mag-light and look head on at the cantilever, observing its revative movement with the cartridge. Then also I sometimes find a spot in the background and observe both the movement of the cartridge and the cantilever as it traverses the record. It is much easier to see with a record that is more off-center than usual, but it is rare that I find a record that I can't observe as being off-center. I have, and do, observe less movement of the cantilever when I damp the arm.
I'm going to jump around- first, yes the last sentence is speculation, I'll explain why later,
I am seeing two separete reasons/uses for damping fluid.
The one reason, is the centering of the cantilever in the coils. The other is the taming of unwanted frequencies.
On the first point on fluid, I am suggesting that the cartridge is a more accurate transducer with its cantilever centered in the coils.
On the second, There is both the issue of the speed at which the cantilever moves (which translates to unwanted bass information, the movement of the cantilever transmitting a bass frequency, which is the 'bell curve' stated above) and the control of the other frequencies higher in the audioband, the resonences.
Now I bounce again. You are correct in your statement of the tonearm being of mass according to the complience, as one reason tonearms are heavier than they used to be, and mass having other purposes. The current thinking in design is to keep the moment of inertia as low as possible in all planes, and at the same time, add as much mass as possible while keeping the moment of inertia low. Tonearm makers are also adding mass at their bases (no doubt allowable by better tables). This can be likened to the effects of a heavy platter being a better sink for resonences, as opposed to just better speed control. Or a heavy suspensionless table as opposed to one that seeks only to isolate.
The other trend, although not new, is the use of new lightweight materiels. The reason for this is not simply to make the arm lighter, but to make the arm stiffer by using more of these stiffer materiels. The stiffer the arm is, the more energy is channeled to these massive energy aborbing bases, and the more solid the headshell is at the cartridge as a result of all this.
Now I bounce back to the meat of our discussion. If we explore the use of damping for the second reasons I stated above, and we explore what is happening at the cantilever as far as movement (I like the bell curve analogy) we may be able to get better results by substituting mass for the reasons we use fluid in that area. The evidence I use to support this is that 1) arms have been getting a lot heavier, and that they seem to be attempting to put it where the moment of inertia would be lowest, 2) The graham arm has its weights slung out at an angle for stability and proper tracking of the cartridge offset angle, and may be enjoying the effects of greater horizontal mass as a side effect, 3) the reputation of linear trackers to have good soundstaging qualities, as they have a disproportionate horizontal mass to their vertical mass, perhaps an overlooked side effect in their quest for accurate transcription.
All of these arms use fluid, however, and in the case on the linear trackers, before it was made available on the E.T., the heavier armtube had a reputation for snapping cantilevers, for the first reason I stated above for the use of fluid. The wheaton also uses fluid, (a real heavywieght), slung out in what seems effective for tracking more than transferring resonences, and of coarse the immedia and graham both depend on it for proper operation.
Further evidence I suggest, and this is perhaps the most compelling, is that the rb-type arms are the lightest in this class, and that they have all benifitted from adding more weight from aftermarket counterweights, And in the deliberate attempt to add weight only in the horizontal plane, seems to have shown results disproportionate to simply adding more weight, and was added in what was stated in the beginning of the thread as perhaps an oversight to why the graham works so well. (the rega does not need counterwights for lateral stability the way the graham does).
I would correlate that fluid on a rega is a rarity and that it is also a lightweight.
So my thoughts now are, do we need to emply fluid on the rega to explore this to a higher limit, and also, if we choose to use more mass, if we would be better off not using fluid for certain applications. There may be trade offs as to how much we allow the cantilever to move in relation to the cartridge, wheather or not the cartridge/arm would work better being in a static position over the record, and reap the benifits of stability, at the expence of letting the cantilever out of center,(or evan if we could get dangerously close to causing damage).
What is particularly fascinating to me is that I have never heard of adding weight for the purpose of modifying the behavior at the cartridge end, but certain evidence shown here seems to support it. That is why, although I believe that to allow the arm to freely with the groove, as opposed to remaining static, is better, I am willing to question it. |
Thanks, Zaikesman. I agree that with the low bearing friction, there may be some movement allowed with the mass system. I cannot say that absolutely no lateral movement is present over the groove. But I can definitely say that it is much reduced, because of the results of greater dynamics, crisper detail, and bass I got with it. I may not even have the best ratio, because I did not try a bunch of different weights. But by luck, I got a pretty good result on my first try.
I have thought of making the weight on a threaded shaft, so it could be adjusted for distance from the pivot. This would complicate the prototype, but production would be just as easy. You may have caused a price increase :^). It would certainly be more applicable to a wider range of cartridges with a system like that. But the spread would have to be equal on both sides, or you'll be changing the anti-skate force.
What type of arm are you using, Zaikesman? If you have a way to fix these weights onto your arm, I could send you a set for evaluation on your TT. I know that they easily go onto a Rega arm. I would like to get some feedback on this. I am already sending a set to Nrchy, who has a Rega RB900. And if Basement wants some to try out on his RB300, he can have some too. They don't cost much, I just get them at the fishing store. If you think you could somehow get them onto whatever type of arm you've got, just email me your address, and I'll send a set to you.
By the way, I think that there is a sort of "mass threshold". As an example, if I am lifting weights, and I keep increasing the weight, at some point I won't be able to lift it. The static moment of inertia will be too high for me to overcome. That is the "threshold" that I am looking for with this system. If the mass is higher than the cartridge can overcome with its suspension, then the theoretical infinite mass can be approximated. As long as the arm can still move freely to track the groove spiral. Since the spiral tracking occurs over a long arc, the low friction of the bearing should allow this to occur, but on the quick dynamic spikes of the groove info, this mass should be sufficient to virtually eliminate arm deflection, if the mass is calculated correctly. Do you agree with this hypothesis? |
