Ahh, a thread after my own heart.
For the last year or so, almost all my efforts to improve the sound of my Technics has centered around refining the way the tape is pulled across the heads and other mechanical aspects.
Where to start?
One of the things that shocked me was the switch from the stock guides (just above the capstan) to the bearing/ceramic guides that's part of Doc's tape path mods. This resulted in an opening up of the high frequencies. Very easily heard and I guess the shock was, with the capstan and pinch rollers between the playback heads and this guide, I had to wonder how much influence this could possibly effect.
I also had some new pinch rollers made since (upon examination with a magnifier) I noticed that the surface of the stock rollers was covered with tiny cracks. When I received the new rollers, I installed them and adjusted the pressure to the factory spec. I gave them a couple of weeks on my machine and no matter how I adjusted the pressure or tweaked the moment of contact (and release), the new rollers seemed to add a few veils to the sound. Very disappointing! When I installed my old rollers, no matter how I adjusted the pressure, they sounded much better. But, I did find that adjusting the pressure a little higher than spec, the sound was the best. The improvements in sound were also in the highs.
I suspect that it's the shape of these rollers that make a difference on the Technics. Here's why;
There was a great discussion of the 3M isoloop machines on either the Ampex or Studer list a while back. This discussion centered around the deformation of the tape at the pinch roller's point of contact. The size and dynamics of this contact patch varies with the rollers tension and the surface of the capstan. This all got pretty technical but suffice it to say that the stretching of the tape by the roller alters the absolute speed as the tape crosses the head gap. One of the things that I took away from this discussion was that the shape of the business end of the rollers has to be a factor.
So, having two rollers, the Technics needs to avoid the stretch at both ends of the isoloop from fighting each other. I think that investigation of the reaction time of the servo might be useful.
Another thing I've addressed have been the bearings used in different areas of the transport. I found that the best bearings seem to actually sound inferior to cheaper, looser bearings in some areas.
I can hear the "What?" ringing off the walls out there on this one.
On things like the capstan and reversing roller, getting the best bearing out there reduces cogging, preventing speed variations and noise transmitted to the tape/head interface. But these applications have considerable mass so the highest tolerances are preferred.
But when it comes to the tape tensioner bearings where the load is very small, a bearing that's relatively sloppy and is loose enough to instantly respond to speed variations and presents the least resistance to the tape is better. My guess is that as minor speed corrections are made by the servo, any drag presented by the tensioner bearings amplifies tape tension variations which makes the tension control overshoot what it really needed to keep tension constant during these tiny speed corrections. Like I said though, just a guess.
One thing that's ongoing with me is tweaking how the tape moves through each guide. Ideally, I'm trying to keep the tape from ever touching either side of the guide. Of course, tape to tape variations makes this impossible to do consistently but having the tape move down the center of each guide much of the time reduces scrap flutter. This is where the word "Obsession" really applies.
Past the above, the things I'm currently working on (or think I'm working on) is improving the way the headblock connects to the machine, and reducing motor noise transmitted to the headblock.
You probably know that I hang out at some of the restoration/archivist sites and I believe that a machines ability to play damaged tapes can be a different kettle of fish. This depends on how badly damaged the tapes are. Machines that have a great number of on the fly adjustments like custom headblocks and speed/tension adjustments are a leg up on any general purpose machine. Some of the archivists have headblocks that have easily adjustable heads which have repeatable settings. Special adapters made for playing cupped tapes can be seated into adjustable swing arms to hold the tape flat as it crosses the gap. I've rigged up my own using squirrel hair water color brushes (insanely soft) taped to the machine face. The results vary widely.
Well, as you can see, I just love doing this kinda stuff but for anyone who doesn't know it already, the way the tape moves across those heads is huge!