There have been many commonly held beliefs about anti-skating force that have popped up in this thread. Many of those same beliefs are held by some tonearm and cartridge designers and some audio journalists.
Unfortunately, many of them are simply not true. I'll cover a few of them here, but first a primer:
SKATING FORCE IS A FUNCTION OF TWO ELEMENTS - FRICTION AND EFFECTIVE MOMENT ARM (EMA)
I'm sure everyone understands friction, so what is EMA? I went over this in an attachment on
one of my past website blog articles but below is the relevant slide. Just think about having to draw a three line segment series with each joint at 90 degrees. It is important to start the drawing series with the starting point at the spindle and drawn to stylus location. The second line in the series runs from the stylus to a point where the line will be at 90 degrees to a line drawn to the pivot point of the arm. (This second line will also happen to be the SAME length as the effective length of the tonearm.) The last line segment completes the journey to the arm's pivot point. Remember, each join angle must be 90 degrees.
As long as the third segment (the Effective Moment Arm) in that series EXISTS in the drawing there will be a skating force. The longer the EMA, the greater the skating force. In the drawing below you can see how the EMA is longest with the stylus at the outer area of the record followed by a decrease and then increases again at the inner area of the record. If you were to draw this line segment sequence for a tangential bearing arm, you'd see there was no third line - and therefore, no skating force.
MYTH #1: Longer arms do not have skating force
As you can see from above, EMA does not decrease when moving to a longer tonearm but skating does decrease in longer arms because the ratio between EMA and effective length changes. Since 12" arms are about 30% longer than 9" arms, does that mean that skating force decreases by 30% for 12" arms? NO! This is because friction remains independent from arm length and friction is the second element required to have a skating force develop. It takes a more sophisticated logarithm to calculate the true decrease in skating force when moving to a longer arm. I won't touch that one now, but suffice it to say the decrease in skating force is nowhere near 30% less skating force when moving to a 12" arm.
MYTH #2: You can adjust anti-skate by ear by listening for the balance between channels
Cartridges are velocity sensitive devices. The higher the velocity, the higher the output. Given a groove with equal modulation levels in both channels, to say that changing the horizontal force of the stylus on the groove walls will change the tracing velocity with which the stylus will trace one groove wall versus the other just doesn't hold any water. You can claim (rightly) that the higher the horizontal force, the more groove deformation occurs in one channel. You could maybe further an argument for groove deformation vis a vis higher inertial forces resulting in slightly greater excursion distances over the same time period resulting in slightly higher output but I'm not aware of any study that hasn't found mistracking occurs far before measurable channel output changes (much less perceivable ones) . Maybe it's out there, but I've not seen it.
It is FAR, FAR more likely that the reason people hear a change in channel balance when adjusting anti-skating is because they are imparting enough horizontal torque (via either skating force or anti-skating force NET of any arm stiction and the tonearm’s internal horizontal torque force) that the coils rotate with respect to the 45 degree groove wall. In other words, you are changing azimuth!
We know very well that when you change azimuth, it is very easy to generate up to 2dB and more of channel imbalance. I've seen it hundreds of times in my lab. It takes very little rotational force to shift a coil bobbin around its central axis.
There are so, so, so many problems with optimizing by ear that I should write a book about it one day. Notice I didn't say there are problems with IMPROVING the sound by ear. (Improving and optimizing are not synonymous because the latter aims at known goals that will deliver the pinnacle performance using both objective and subjective measures.) While IMPROVING your sound by ear is entirely possible, as I've explained many times before, you will often never know if the NET improvement you just heard resulted in a deterioration of another parameter or two.
There really are 7 targets to hit to get the most out of the groove. My point is: why not take each one down in a UNIVARIATE manner rather than using a multivariate attempt such as listening (or using a test record!
- exception on azimuth, of course). Multivariate tests are almost always the foil to certainty and optimization.
Almost all adjustment functions on a tonearm will change at least two - and up to five- other parameters. I don't want to wonder whether the improvement I just got by listening for improvements when adjusting something resulted in enough of an audible improvement to mask a deterioration in other parameters. By using univariate means of setup and optimization, one need not wonder.
MYTH #3: Since skating force is a moving target, it doesn't matter enough to get things precisely tuned because it will never be “right”
Yes, skating force changes with EMA (as above) and it also changes with the coefficient of friction (e.g., groove amplitude, vinyl formulation, etc., etc.). However, we KNOW THE RANGE OF VARIABILITY. If we know the range of variability, what is the best thing to do to be right most of the time? AIM FOR THE MEAN.
A distribution curve has a high point and two tails. We aim for the high point in the dataset since that is where most of the action is! That is what the WallySkater has been about from day 1 of its development by Wally.
What continues to amaze me about the skating arguments in audiophilia is that these published scientific studies have been known about for decades and repeat tested time and again with the same outcomes. Why does the will to misunderstand - even among some manufacturers - persist?
MYTH #4: I can trust the tonearm manufacturer with regards to the anti-skate setting
The degree to which this is a myth varies by manufacturer. There are tonearm designers who really do understand skating force. There are tonearm designers who understand skating force but can't seem to design an effective anti-skating device. There are tonearm designers who don't understand skating force at all and there are even a couple designers who have defeated the laws of physics somehow because their pivoted tonearms "don't need it".
I won't touch the last example as it would take too much negative energy from me.
Even with well calibrated anti-skate mechanisms, we need to be concerned with the internal torque force of the tonearm WITHOUT anti-skate engaged. Is this force (since it usually exists to at least a small degree) complementary or fighting our anti-skate target? Many designers are quite ham-handed with their application of anti-skate and a lighter approach is needed. One exceedingly expensive design is quite the opposite. In any case, the instructions to apply a prescribed amount of anti-skate force cannot apply to every given arm off an assembly line since every arm comes off that line with varying degrees of internal horizontal torque force - usually caused by the tonearm wires themselves but sometimes the bearings as well.
MYTH #5: Look at the alignment of the cantilever during playback to set your anti-skate
I almost didn't want to mention this one as it is kind of "out there". Assuming you can get a good view of the cantilever, against what fiducial will you align? How will you ensure you've eliminated parallax error? How do you account for tonearm bearing stiction? How do you account for horizontal compliance? At what radius are you doing this measurement? With what groove amplitude are you doing this measurement? (I heard one person's approach that doesn't even involve groove friction; just lowering the arm to a non moving record!)
So many issues on this one. Enough said.
See next post for remainder...