Hello all - Chakster mentioned the Stax UA-7, which I'm currently using with a handful of different cartridges whilst mounted on a TD-124. It's geometry is a bit unusual, putting the inner null point substantially inside even the Stevenson solution (at 56mm), even though it has a longish (for a "9-inch" arm) 240mm effective length. I like this tonearm's general design and versitility, the low pivot friction, the consistency of its anti-skate mechanism, the graceful aesthetic, and the overall quality of fit and finish. I'm consistently happy with the sound quality from a wide variety of different cartridges. But I've never understood why they chose the geometry they did . . . mainly because I have yet to find a record that has modulation so close to the center as to justify it.
In exploring other geometries, my main fork in the road was whether or not the stock 20-degree offset is to be observed . . . and if this is the case, one cannot pull the inner null point toward the outside of the record without also pushing the outer null point towards the center, and if the tracking error is to be kept fairly constant at the outside of the record, then there must be a corresponding increase in both the effective length and the mounting distance. If one alters the offset, then it's possible to move the null points both further apart and further away from the center . . . that is, towards the Baerwald solution.
I personally consider the act of changing tonearm geometry in the same vein as that of all equipment modifications . . . it's worth it to the extent that one is solving specific issues with a reasonable amount of elegance and precision. But there's a point after where a certain amount change starts to fly in the face of the original design, and then you have to ask yourself . . . "If I'm changing it this much, do I really like this thing enough to own it in the first place?" On tonearm geometry, it's obviously a matter of degrees (couldn't resist the pun) . . .
I ended up mounting it at the stock mounting distance from the spindle . . . I wanted to try the stock geometry out of respect for the designers and so I could thoroughly correlate the sound with my own preconceptions of what the "correct" geometry should be. I've found that it leaves little to be desired with elliptical and spherical styli (i.e. Shelter 501 and Denon carts), and it also sounds great with an integrated-style headshell for B&O cartridges (MMC20CL and rebuilt MMC6000) . . . the latter with a custom lighter counterweight, which combined with the headshell, gives an appropriate reduction on mass for the higher compliance cartridges.
For the fancier MCs (Cadenza Bronze and Delos) I've pulled the overhang up to 15.5 mm and the offset to 21.7 degrees (2.5mm and 1.7 degrees from stock) - this is a solution "in the Baerwald direction", and the results are outstanding. I arrived at this by playing with the fit of a few cartridges in the stock headshell, and coming to the conclusion that there was only so much extra mounting room before it really started to "feel wrong" and become a decidedly inelegant setup. I do understand the concern that altering the overhang creates a variation of azimuth with vertical deflection of the tonearm, but when you actually see it in practice, 1.7 degrees is extremely small to affect this compared to the length of the arm and the deflection of a moderate record warp.
I will echo the sentiment that whatever the geometry is chosen, precise alignment is an absolute must. I'm very much an arc-protractor guy, and I have them made as required for every turntable/tonearm geometry solution I set up. My approach is to calculate the geometry with an Excel spreadsheet, generate an HPGL file using my PC-board design software that has the appropriate arc placement and null points, and have them machined to order by Front Panel Express, usually for about $30 apiece. I verify the spindle-to-pivot distance on existing mountings usually with a 12" dial caliper and a bit of basic math . . . for new mountings I have a variety of dummy tonearm posts machined locally, combined with the caliper and/or mounting-distance bars/plates sourced locally or from Front Panel Express.
The key to precise alignment is to use methods that allow each dimension or angle to be verified against each other, to make sure your measurement techniques properly fit the specific hardware, and to understand what the required tolerances are for each dimension. Being a mm away from perfect on the spindle-to-pivot is VERY different from being a mm off on the overhang.
In exploring other geometries, my main fork in the road was whether or not the stock 20-degree offset is to be observed . . . and if this is the case, one cannot pull the inner null point toward the outside of the record without also pushing the outer null point towards the center, and if the tracking error is to be kept fairly constant at the outside of the record, then there must be a corresponding increase in both the effective length and the mounting distance. If one alters the offset, then it's possible to move the null points both further apart and further away from the center . . . that is, towards the Baerwald solution.
I personally consider the act of changing tonearm geometry in the same vein as that of all equipment modifications . . . it's worth it to the extent that one is solving specific issues with a reasonable amount of elegance and precision. But there's a point after where a certain amount change starts to fly in the face of the original design, and then you have to ask yourself . . . "If I'm changing it this much, do I really like this thing enough to own it in the first place?" On tonearm geometry, it's obviously a matter of degrees (couldn't resist the pun) . . .
I ended up mounting it at the stock mounting distance from the spindle . . . I wanted to try the stock geometry out of respect for the designers and so I could thoroughly correlate the sound with my own preconceptions of what the "correct" geometry should be. I've found that it leaves little to be desired with elliptical and spherical styli (i.e. Shelter 501 and Denon carts), and it also sounds great with an integrated-style headshell for B&O cartridges (MMC20CL and rebuilt MMC6000) . . . the latter with a custom lighter counterweight, which combined with the headshell, gives an appropriate reduction on mass for the higher compliance cartridges.
For the fancier MCs (Cadenza Bronze and Delos) I've pulled the overhang up to 15.5 mm and the offset to 21.7 degrees (2.5mm and 1.7 degrees from stock) - this is a solution "in the Baerwald direction", and the results are outstanding. I arrived at this by playing with the fit of a few cartridges in the stock headshell, and coming to the conclusion that there was only so much extra mounting room before it really started to "feel wrong" and become a decidedly inelegant setup. I do understand the concern that altering the overhang creates a variation of azimuth with vertical deflection of the tonearm, but when you actually see it in practice, 1.7 degrees is extremely small to affect this compared to the length of the arm and the deflection of a moderate record warp.
I will echo the sentiment that whatever the geometry is chosen, precise alignment is an absolute must. I'm very much an arc-protractor guy, and I have them made as required for every turntable/tonearm geometry solution I set up. My approach is to calculate the geometry with an Excel spreadsheet, generate an HPGL file using my PC-board design software that has the appropriate arc placement and null points, and have them machined to order by Front Panel Express, usually for about $30 apiece. I verify the spindle-to-pivot distance on existing mountings usually with a 12" dial caliper and a bit of basic math . . . for new mountings I have a variety of dummy tonearm posts machined locally, combined with the caliper and/or mounting-distance bars/plates sourced locally or from Front Panel Express.
The key to precise alignment is to use methods that allow each dimension or angle to be verified against each other, to make sure your measurement techniques properly fit the specific hardware, and to understand what the required tolerances are for each dimension. Being a mm away from perfect on the spindle-to-pivot is VERY different from being a mm off on the overhang.