Post by Owen Y on Jun 7, 2017 11:29:12 GMT 12
This is a perennial topic on hifi forums, so I thought I'd open a thread (or two) for everyone to discuss if they wish. Also to gather the info in one place as a resource for members to find. There are numerous related topics (eg. VTA/SRA, anti-skating) but we should start with optimal tracking alignment.... (I certainly don't know everything about the subject, but have probably spent more time than is healthy, thinking about it over the yrs ) Geometry of a pivoted tonearm tracking the record groove 'tangent': H.G. Baerwald's 1941 paper on 'Optimal Pickup Design' is regarded as one of the starting points for our modern understanding. Baerwald used mathematical analysis to show that for pivoted tonearms: - 'Weighted' Angular tracking Error is most important (Angular Error divided by groove radius) because distortion is inversely proportional to groove radius.. - There are 2 points of zero tracking error across the groove area of a record. - These 2 'null' points depend only on the minimium & maximum record groove radius. - There will be a corresponding Offset Angle & Overhang. Also: - Groove friction increases with angular tracking error. - Skating force increases with stylus friction (ie. angular error).
|
Post by Owen Y on Jun 7, 2017 12:06:37 GMT 12
Baerwald suggested that Peak Harmonic Distortion should be minimised across the record, ie. at the 3 peak distortion points (60.3, 89.3, 146mm). The familiar fish-hook graph (shown here for a 9" (229mm) tonearm): It was B.B. Bauer I believe, who suggested that hi-fi record inner groove radius should be 2.375" (60.325mm), together with outer groove radius 5.75" (146.05mm). The resultant 2 Null Points for Baerwald's optimisation are at 66.0mm & 120.85mm radius.
|
Post by Owen Y on Jun 7, 2017 17:00:59 GMT 12
'Löfgren' alignment:Actually, I understand that it was Prof. Erik Löfgren in Sweden in 1938, who first published the idea of groove radius-Weighted Tracking Error. Löfgren also proposed an alternative 'B' alignment, to reduce the 'annoyance factor' of distortion over the long. slow portion of the record between the 2 Null Pts. He suggested that the trade-off of short durations of higher distortion at the inner & outer groove peaks is preferable..... The 2 Null Points for Löfgren's 'B' alignment are at 70.3mm & 116.6mm radius.
|
Post by Owen Y on Jun 9, 2017 9:12:22 GMT 12
J.K. Stevenson also proposed an alternative alignment (in 1966), with zero tracking error occurring at the 60.3mm inner groove. This was based on the idea that inner groove distortion was more annoying & that orchestral finales, crescendoes most often occurred there. Of course, this results in slightly higher distortion elsewhere. This is a comparison of these 3 alignments in terms of Tracking Error Distortion:
|
Post by Graham on Jun 9, 2017 9:40:21 GMT 12
Excellent work Owen. This final graph clearly shows why most people, myself include, consider Baerwald alignment to be the best compromise of the 3 options. It also demonstrates how much of a compromise cartridge alignment really is with a pivoting tonearm. A parallel tracking arm of course doesn't suffer from any of this error or distortion, but unfortunately introduces other complications.
|
Post by Owen Y on Jun 9, 2017 10:08:34 GMT 12
Yes everything involves compromises. Linear tracking tonearms - as I've mentioned elsewhere, IMHO the main advantage of Linear trackers, is not (only) zero tracking error, but the elimination of Skating/Anti-skating side-forces. Didn't you DIY a linear tracking tonearm Graham? (I went as far as trialling a linear bearing, based on a Japanese DIYer's idea, but I'll open a separate thread for DIY tonearms.)
|
Post by Graham on Jun 9, 2017 12:20:51 GMT 12
Agreed, having the only force being applied to the stylus parallel to the cantilever must be a huge advantage. However, achieving that is the challenge with linear tracking arms. IMO most end up being far too complicated and even with theoretically zero friction linear air bearings there is still the quite high lateral inertia force that the poor little cantilever has to contend with. My DIY effort must be somewhere on the old Audioenz files.
|
Post by Owen Y on Jun 9, 2017 15:08:17 GMT 12
Chaps - sorry to rant on about this, but this stylus-strain concern is a misconception IMHO.... There are 2 mechanical things happening: 1) Stylus tracing the microscopic groove modulations. 2) Stylus moving across the record, from outer to inner groove. 1) Tracing groove modulations: When the stylus is tracking groove modulations, being moved up/down/left/right, at the frequencies of these modulations (say 15hz - 20+kHz), the tonearm & cartridge MUST appear completely rigid, as 'seen' by the stylus. Problems can occur at low freqs & this is where Effective Mass come into play - ie. the combination of the mass of tonearm/cartridge/screws & the compliance of the cantilever suspension must have a resonant freq that is lower than the lowest groove freq that we wish to track. (At system 'resonance', ability to deliver power is lost - same as loudspeaker driver resonant frequency.) It doesn't matter if the Effective Mass of the tonearm/cartridge is 10gms or 100gms, as long as Resonant Freq of the system is below say 12Hz. Problems can also occur at HFs, eg. bearing 'rattle'. 2) Moving across the record: The tonearm moves across the record of course, either arcing or linear tracking - but very SLOWLY. As long as bearing friction is low, then this will occur without problems or strain on the cartridge cantilever. 'Inertia' is the mechanism at play here. Typically, the mass of just the counterweight of a pivoted tonearm (eg. 150gm) will be higher than the mass of an entire linear tracking 'slider' + cartridge (eg. 50-80gm). Rotational Inertia (M x R^2) is at play. With linear trackers, overall (arm slider) mass is less & arm length is usually much less than pivoted arms. Friction is so low in either type of arm that no 'strain' problems should occur at the stylus - except maybe with eccentric records & pivoted arms present much higher Rotational Inertia than linear trackers. Air bearings, of course, have pretty much zero friction. Happy to hear any different views on this.
|