HFNR and antiskate on SME 3009/III

When I have my SME3009/III set at approx. 2Gm on the antiskate, and listen to the torture track on the HFNR test record, I get a buzzing in the right channel. Pulling gently on the antiskate string removes the buzz, so I equate that to increasing the antiskate.

Well, if I do that so that it doesn't buzz, the darn thing will not cue up on the outside of the record -- the tonearm flies right off the record. (!)

The antiskate pulley has been lined up (per SME instructions) so that it is perpendicular to the tonearm at the start of the record, and fairly skewed by the end. My thought (and it's dangerous when I think) is that the SME instructions are wrong, and it should be perpendicular when the tonearm is at the inside of the record, not the outside.

Anyone else have this issue?
I posted this elsewhere, but I'll save you the search:

I wish some of you "mavens" would try my "sure-fire" anti skating calibration. It's so easy (and yes Doug, as w/ VTA it gets you in the ballpark, but not all the way "there"), but it gets you very close.

First, as you probably all agree, it's useless to match the anti skate setting to the VTF, which is what most TA mfrs. recommend and how their dials are calibrated. This is due primarily to the differences among cartridges between suspension material (amount of deflection for different materials) and stylus shape (amount of friction in the groove.)

The object of applying AS force is to neutralize the "inward twisting force" of the tonearm toward the record center. This force is the product of a lever arm length -- the perpendicular distance between the centerline of the (offset) headshell and the horizontal tonearm pivot -- times the friction force of the stylus in the groove. BTW, the old time straight transcription tonearms (no offset) didn't have this problem, but of course suffered from more tracking error.

If you will (with zero AS) just watch the position of the cantilever (from the front, with lots of light please) as you lower the stylus into the groove, you will notice that it deflects toward the outside of the cartridge (because the tonearm is pulling inward in the opposite direction.)

As you apply more and more AS this movement will eventually stop. If you apply too much AS, the cantilever will deflect toward the inside of the cartridge. It takes a little observational practice, but not a hell of a lot. The result is a cartridge-specific AS setting which is amazingly accurate. I've tried to dial it in further by ear using a mono female soloist record (keep the preamp in 'stereo' setting) but usually I get it right by eye. I suppose you could do a little better with the oscilloscope technique, but I don't have one handy :~))

PS: Try this technique with (any?) vdH cartridge. vdH is one cartridge mfr. who is very specific about AS settings (and they're QUITE low compared to standard tonearm mfr. recs.) I found (and you will too) that the setting you wind up with (on the AS force dial if you have one) is precisely what vdH recommends.
There is an excellent post on Audioasylum today about this. Go over there and read it. To simplify things, what you want to do with that section of the record is get it so that you get EQUAL (but low level buzzing) on each channel. Getting no buzzing on either channel is going to result in grossly higher or lower antiskating than necessary. Once you've done this, fine tune from there setting by ear if necessary. I would trust my ears more than anything else. The Vdh method above was not even close with my table/arm/cartridge combo; Vdh essentially recommends very little antiskate (as little as 1/3 of VTF running the VTF very slightly higher than recommended) and it definitely didn't work for me. My tonearm is older, though and supposedly has some issues with the antiskate mechanism deteriorating to the point of needing overcompensation (which I also found info on at Audioasylum) which is, in fact, how it has worked out.

Go to the vinyl asylum at Audioasylum and do searches on your tonearm, cartridge and antiskate. But be prepared to invest some time reading; there's a lot of info there and it's a bit of a controversial topic with a lot of varying opinions.
Hdm, the method I explained above is not vdH's method, it mine. With vdH cartridges, it (coincidentally) yeilds the antiskate figures vdH recommends for their cartridges,

It will work quite accurately for any arm/cartridge combo however, so long as all other physical setup adjustments have been properly made, including VTF.

There's nothing mystical about anti-skate. It just a way of neutralizing the inward force created by tonearms with offset headshells, or by S curved tonearms. When there's no sidethrust on the cantilever, it's quite easy to see after just a little practice. And whatever that force happens to be, it's the right one for that arm and cartridge.

As a historical note, anti-skate was not available on tonearms for a long time, and nobody seemed to mind. They simply adjusted the balance control if things seemed a little "off". Getting proper SRA, VTF, stylus overhang, azimuth, and coil loading are far more critical. In fact if these other adjustments aren't perfect first, making antiskating adjustments will be futile IMO.
Listen to Nsgarch. That test record will lead you astray every time. I tried its "torture tests" and the only cart/arm I have that will sail through all four of them is a l962 Empire arm with a Stanton 881 cart on it, and it has no antiskating provisions at all! I have an SME IIIs myself and wouldn't dream of using the record to set anti-skate.
Dopogue, I have to admit I haven't used the test record Hdm refers to. I don't have anything against them, but my position is "why bother when it's so easy to do visually." In fact I have confirmed my visual adjustment(s) electrically time and again, as well as with listening tests (using female soloist mono records with preamp in 'stereo' setting) and the method seems to give consistently accurate results.
Agree with all the above.

Nsgarch's visual method works well. (Note: it's better with high compliance cartridges than low compliance ones.) Fine tune by ear while listening to music.

I do have the HFN record. As I've posted many times on VA, the design of its so-called anti-skate tracks (side one, tracks 6-9) is inherently flawed. Do not use them for setting AS.

If you want to use that silly record to rough in AS, use the three, widely spaced "tracking test" bands on side two. Adjust AS until L/R buzzing is about equal on all three bands. Assuming your VTF is correct, that method is valid. The tracks at the end of side one are not.
I used the Wally antiskating tool to set up my TT. I confirmed the results with the HFNR test tracks. Just out of curiosity I tried Nsgarch's test. This method works as well for me. I even set the antiskate to the opposite setting and saw exactly what Nsgarch was referring to. I can't imagine this not working for any cartridge.
Read the Thornes white paper on setting anti skating. The correct value is 14-16 percent of your adjusted tracking force. Therefore if the tracking force is 2 grams you should sets you antiskate to 0.28-0.32 grams.

Make sure your table is absolutely level in both horizonial and vertical directions
Etbaby, would you mind telling us where we can find the Thornes white paper? Thanks.