The Shelter should not be super-sensitive to VTA... it sounds like you are having bigger problems with your sound.
Use the HiFiNRR record to dial in the antiskate. Then test it with some real-world records that have big peaks in the inner grooves (most opera and orchestral records should work).
What phono pre are you using, and what loading for the Shelter?
I'm not sure whether improperly set anti-skating can cause significant permanent damage to records unless they are played repeatedly. Your stylus pressure of 1.9 grams seems quite standard. I have seen improperly (anti-skate) compensated styli skitter out of a groove and rip across a record with about this much stylus pressure. Of course, if the stylus pressure were high enough, this wouldn't happen, but groove damage becomes more likely.
Your comments seem to suggest that damage to the LPs is a possibility. If so, perhaps the damage occurred only where the Shelter's stylus touched the groove, and this may be a different place than where the Clearaudio stylus touches the groove. I don't really know if the styli actually have different shapes, but if they do, perhaps one is narrower and contacts the groove walls further down, i.e., deeper in the groove, than the other stylus. With this, the Clearaudio is tracking the groove walls at a level that wasn't damaged, while the Shelter would be replaying the groove where it was damaged. Other possibilities exist depending on stylus shape specifics.
I hope you are right about the whole effect being only psychological. Good luck.
This is something that is very hard to diagnose for certain, without being there. I think that you are going to have to make the determination yourself, from your listening and adjusting process.
Even though damage to the record is unlikely, it could have happened. Playing ok with anouther rig and cartridge does not preclude possibility of damage, because the cartridge may trace a different area of the groove, depending on the stylus shape and dimensions, as Jameswei said.
I would also check the azimuth because the machining of the collet could have changed vertical alignment, and re-check all the cartridge alignments. The large VTA movement you did when you re-installed the machined collet could have affected the overhang adjustment slightly.
That sounds like a very nice rig, so it should be awesome sounding. Good luck.
Not much to add to what the others have said, except to suggest re-cleaning the affected LP's. Perhaps the excessive sidewall pressure ground some dirt into one sidewall. Certainly couldn't hurt to try.
Regarding VTA, I haven't heard the 90X but unlike Patrick we find our 901 to be VERY sensitive. Arm height changes of .04mm or so are clearly audible on the large majority of records. I am building a list and we're re-setting arm height with every change of label or record thickness. Boy that's fun :( Worth it when you find the magic spot though.
John, because I have some experience with the 'Not/Illustrious combination (you of all folks knowing why), I will probably mention some things that you have already done but that others might find useful. I would first suspect the azimuth as Tom has mentioned. Check the arm pod of the Hyperspace first to ensure that it is plumb and level with the platter plinth. Next, make sure that the collet ring is seated "plumb" such that the arm pillar is perpendicular to the platter. After these are successfully accomplished, check the azimuth (with a small spirit level on top of the headshell is possible). The Illustrious CAN be adjusted for azimuth by CAREFULLY loosening the hex nut located behind the headshell, CAREFULLY rotating the headshell, and the CAREFULLY re-tightening the nut. I am told that the re-tightening is most crucial so as not to collapse the arm tube.
The wire cradle for the anti-skate "bob" (according to OL) should be at approximately the angle as shown on the template provided. HOWEVER, there are many who believe that this cradle is best set to be perpendicular to the arm when the cartridge is positioned at the outer edge of a record. Too, the cradle is held with a tiny hex screw and it can get out of vertical - loosen the hex nut and make sure that the wire is fully seated in its "hole".
Heres to hoping that you get back to your original sound. Oh, one last thing. Make sure that you tighten the two grub screws on the collar JUST tight enough to keep the arm from falling. Otherwise, if one is tightened more than the other it can throw of the pillar and put it out of perpendicularity! :-)
Hmmm, I have also heard others, in addition to Doug, comment that their 901 was sensitive to VTA. My 501 is far less picky about VTA than it is about VTF. Others have shared with me similar experience regarding the 501. No clue about the 90X, though.
Thanks for all your help. Patrick, I did use HiFi News test record to reset the anti-skating bias. Tom, since the inside of the collett was fitted to a mandrel that was cut to fit on the lathe, the machining to the outside SHOULD be perp. to the pivot. Doug, I did reclean the records. I will say that when the problem first occured that the anti-skating bias was so out of whack that the arm would travel very rapidly to the outside of the record. I doubt if there was much force at all on the inside groove wall. I have noticed that small changes to the VTF do make a difference. After the tracking problem I set the VTF to 2.0g. I also rechecked the overhang with my protractor. I have since reset it to 1.85g. I have also noticed that amount of torgue (sp?) on the two set skrews for the pivot collett does affect the sound. Since machining the collett to allow the arm to sit lower does affect where the set skrews bite into the collett (and their leverage on it) I'm starting to wonder if this might be a factor also. Also, even with the 0.100" I had machined off, the cart. still appears to the eye to be slightly high in the ass. Should I have some more taken off? I guess I could blow all this headache off and go back to listening to those little silver M_F_ers.
The 901 likes to be very slightly down at the back. It sounds a trifle shrill set up the way you describe. Again, I don't know about the 90X but it's a fair assumption that it might prefer a similar orientation. At the very least you should have the ability to try it. I'd find a way to get the arm a bit lower.
Just my $.001.
That's where I am heading. In the initial setup I had the cart. loaded with 300 ohms. Probably to kill some of that shrillness. Last night I was listening to the new Peggy Lee release, Latin A'la Lee and switched to 100 ohms. Bang! Peggy Lee started to sound like Peggy Lee. Imaging and focas improved. Tried some more records and noted that there was still some of that shrillness. Removed the adjuster to lower the arm by another 0.060". Better, but still not there. I am going to take the collett back to my machinist (how I wish I were still a chemistry grad student at Stanford with access to a machine shop) and have as much as I can removed. Dropping the arm 0.100" doesn't change things much. Also, I find that there is a VERY fine line between too loose and too tight with the two set skrews that fix the arm to the table. Thanks again for everyones input.
John - I am distressed that you are having such difficulties with the Illustrious/Hyperspace combo. On my Spacedeck w/Mat 1 the VTA was a snap. (maybe try a thicker mat of some sort to see how it changes things). Have you considered contacting 'Not with respect to the issue? Hard to imagine there would be that much difference in pods. Know you probably don't want to spend more for a different pod - but maybe they'd have suggestions. It just doesn't seem right.
Yes, the two grub screws make a difference with respect to tightening. JUST enough to hold in place and equal if possible. I spoke to Tom Fletcher once and he told me, and has told others, that all "bits" on the 'Nots should be just snugged so that if all were loosened just a tad, the table would fall apart. I went through and followed this advice and things opened up tremendously when compared to when I had everything cinched down pretty firm.