Oooh, this thread has taken some interesting turns since I last visited. In an attempt not to write another "book", I'll make a few comments:
I'm only making that comparison because its the kind of thing that comes up in these discussions. "All belt drives have XX problem" and so forth. No, they don't all have XX problems. I think every drive system is capable of sounding similar if not indistinguishable from another depending on the context.
...
Oh, YES THEY DO ALL have certain problems. Those problems are not insurmountable, and engineers have worked out ways around them, sufficient to lead to the second part of your statement that every system is capable of sounding similar, if not indistinguishable. No drive system is perfect; they all have some (albeit different) problems. It's how you compensate for those problems, the compromises you make, that can make them all sound similar in the end.
That doesn't mean, however, that you have completely overcome the problems. Sometimes you've solved one, but there may remain a problem that another system addresses better. For example, the "flywheel effect" of a large platter goes a long way to overcome the potential for "stretch and rebound" (basically, "cogging"!) sound of a belt drive... but momentum isn't quite the same thing as torque, over time in the real world.
I think there's also a danger in interpreting differences as improvements. When sometimes, they're just differences. You could start at either end point, and think an improvement has been made when you arrive at the other option.
This is very true. Salespeople know this, and set those U-shaped equalization curves to make gear sound more "exciting" to shoppers. It isn't really better, but most people will think so. This is also a reason --beyond actual changes in gear-- why we should allow a "break in period" before writing reviews, unless we are clear up-front that we are only giving initial impressions. Initial impressions are often skewed, because we focus on what is "new and different" to us, which we may like, but often ignore other, equally important things in the process.
This a a slice from Arthur Salvatore's reference components section on his web site. ....
Effective Torque- The Idler-Drive's Unique and Ultimate Advantage
A well-designed and executed idler-drive has more effective torque (ET)* than any other type of drive system. It's that simple....
Effective torque, in effect, is required to properly "initiate" a musical note (or passage). Since a continual stream of new notes must appear while the LP is being played, adequate ET is a necessity at all times. Without the proper amount of ET, the sound will vary, depending on the amount, and type, of notes (resistive changes) coming into play (as various physical changes in the groove). It's this variance, or musical inconsistency (distortion), which good idler-drives diminish better than any other design.
Marc mc
I basically agree with AS on this. If you read the rest of that section, he pretty much isolated nearly all other variables in multiple tests using otherwise-identical gear, before reaching this conclusion. Not 100 percent scientific, but close. I did something similar, and came to a very similar conclusion, although I had a few other considerations and still have one (perhaps two, not sure yet) other possibilities to test.
I think Mr. Salvatore needs to learn a bit about drive systems and motors.
Why do you say that, specifically? I found a few possible nits to pick in his process of reaching his conclusions, but I didn't think this kind of ignorance was really a factor (in fact, it might have been a benefit -?). His conclusions were based on keeping all else equal, and paring away until he was left with this. He got there against his initial skepticism, and by a fairly astute (albeit not perfect) process of elimination, even using the same motor in two different type systems, to eliminate differences due to, say, "cogging" in the motor.
If Arthur Salvatore's can back or has backed up what he is saying with truthful testing then we can assume what he is saying could be true but does he have proof is all I want to know.
Do read his full text on his website, not just the extract above. He deals with a lot of the other questions or issues, and explains how he reached that conclusion. The process may not be perfect, and I believe some additional experiments are needed to "prove" it, but I also believe that, in essence, he is correct. My own experiments point at the same thing, although I went through a different process, and also do not (yet) have "proof".
There is also a lot of common-sense logic to back it up, if you think about the sonic differences being widely perceived that are "characteristic" of idler-wheels vis-a-vis most others, and what is actually happening within the materials and at the stylus-groove interface.
I'll see if I can come up with a good way to explain without confusing the issue anymore. The stylus drag that Salvatore references is something that would be a real phenomena. During dynamic passages of music, there would be more drag than something that wasn't dynamic. We are talking *very* miniscule changes in speed but some claim to be sensitive to miniscule speed variations. So how do you combat it? With a heavy platter like many turntables use, you are reducing the magnitude of the speed variation, but making the variation last longer. With a lighter platter, you have more magnitude of variation but with shorter periods. Designers who prefer both exist, which is subjectively better will likely depend on the listener. The other approach is to add more torque to the drive system (regardless of type) but then you have to deal with a compromise. Increased torque = increased cogging effect so there is a balance that has to be considered. Another balance is with the amount of compliance present between the drive system and the platter. A rubber belt will have more compliance than most idlers, but some idlers will have more compliance than a tape belt or one of the string drives systems. No matter what, you are forced to pick your poison.
I mentioned using an eddy brake earlier to stabilize speed. It's an excellent way of effectively dominating the effects of stylus drag by making the overall drag orders of magnitude higher than stylus drag, with a huge majority of that drag coming from a very consistent source rather than a variable one.
Basically, I completely agree with what you say, but there is one statement that needs further examination and consideration, when put into practice: " Increased torque = increased cogging effect so there is a balance that has to be considered." First of all, does increased torque ALWAYS HAVE TO mean increased cogging, or are there aspects of motor design that can create motors with higher torque, yet not so increased cogging? Are the two automatically inversely related, and is that a straight-line inversion? Secondly, what happens to the "cogging effect", as the motion is transmitted? Do, for example, idler wheels or belt drives transmit the "cogs" to the same degree, or in the same way?
I do agree with the logic of an eddy brake. It is one good way to reduce the relative significance of stylus drag effects, and a better way than relying upon the momentum of a flywheel (which doesn't work quite as well as one might wish).
How about a relatively high-torque motor, one with minimal cogging, using a well-designed idler wheel AND an eddy brake, with a decent platter and a massive plinth designed for vibration dissipation? How about a precisely- and rigidly-fixed tonearm, again with proper vibration controls? Might be heading in the right direction here, towards that dream "giant killer"...