Tonearm Geometry and Setup

While this subject matter seems to be of 'crucial importance' for our hobby our discussion about this issue is fragmentary and split over different threads in our forum. As if we are talking about some 'abandoned child' that nobody wants. I thought that this 'child' needs its 'own home' in metaphorical sense or its own thread in the
usual sense. While tonearm geometry seems to be very 'simple' in mathematical sense there are many unresolved questions in the actual sence.
My question for the start of this thread is:
'why are we not free to choose the zero points on the record radius independant of the tonearm used?'

48f03962 30db 4aad a0a9 eab5c5d5ce40nandric
I thought that these were the product of the width of the record, which is a constant, the length of the tone arm, which varies, and the chosen formula [Loftgren etc] for producing the least distortion. This would appear to offer zero degrees of freedom so I must not have understood the question.
Stanwal, rather than not understanding the question, I think perhaps you answered it. The zero points are, as I understand it, calculated as you say -- so we are only free to "pick" a zero point to the extent we are "free" to pick one of the several formulas intended to identify the zero points leading to least distortion. True?
Dear Rdavwhitaker, true. No the question is, what are the best points and what does mean "least distortion"..... least distortion where on the records radius .... the groove is not homogenous, as the radius is decreasing ever more with lesser distance to the label and thus the situation for the stereo stylus is getting more and more "unbalanced.
I am of course not (pre)supposed to answer my own question
but considering Stanwals comment I need to put my question more clearly. The context is this: 'we', or at least I, got the impression that I can choose the 'zero points' depending on which part of the LP I want the least distortion. Baerwald is about the best solution for the whole record radius (aka 'the average') while the other provide better solution for specific part(s) of the record. Ie Stanwals 'least destortion' in general will not
do. This thread as well as the question of 'choice' is caused by some remarks by Dertonarm in the 'Copernican thread'.

Dear Nandric, "we" (... we, the people ...) do indeed have a choice which of several options we do choose.
It does depend on personal preferences, stereo or mono, vintage of the majority of the records one owns/plays ( how close cut to inner label ), effective length of tonearm and expectation in sound reproduction/ability of a given audio set-up to recreate soundstage dimensions and acoustical positioning.
Most will fare well with Baerwald/Loefgren A IEC, owners of mostly modern pressings with large wax will do great with Loefgren B IEC, collectors of early vintage stereo records of the 1950ies to mid 1960ies will prefer Baerwald DIN or UNI DIN.
There is no absolute - and there is NO mathematical "best".
It is - as long as this was about mono reproduction only ( and mono only it was when Baerwald / Loefgren A was postulated ) and when one assumes that the music engraved is VERY homogenous without huge climax' towards the end of a piece/movement.
But since music is anything but homogenous in volume or structure nor are all records cut following IEC, we have do address our needs individually.

Nandric, you said:
why are we not free to choose the zero points on the record radius independant of the tonearm used?

You are, of course, as Dertonarm says, free to choose whatever null points you wish. The origin of the null points for the geometries used are based on the inner and outer radii of a record which give a constant, the linear offset, which in turn gives an offset angle and effective length. To be strictly correct, any calculation is only accurate for one record, and is, to a greater or lesser extent, a compromise for all others.

So, a designer of tonearms has to decide first and foremost what are these inner and outer radii. Are they IEC or DIN, or the median of his own collection, or of a surmised potential buyer? In the past, 7" singles had to be considered so the range of possible radii was extended if they had to be accommodated as well as LPs.

In my case, the Odyssey arm used nulls similar to Lofgren A/Baerwald DIN. That was a design choice, as was the nominal effective length of around 230mm. If I was designing an arm today, I would make it have an adjustable offset angle in addition to an adjustable base, thereby allowing the user to choose the alignment. Also, the effective length used to be governed by the space available on turntables. This is less so now, so longer arms are more common, and they have advantages in reducing tracking error and skating forces.

(Though, as an aside, an 8" arm plays a 7" single with less distortion than a 12" arm playing an LP)

All these features in an arm are merely for facilitating setup, they do not define the arm except in the single case of the hypothetical standard LP, whatever it may have been.

Some corollaries of this are:
1. Some LPs may be more useful than others at showing the effects of particular alignments.
2. When evaluating an alignment choose a track at a distortion maximum not a null.
3. When comparing alignments you shouldn't use the same LP, as it will probably suit one alignment better than another unless you wish to use that LP as your standard.
4. You will have LPs with loud passages at the beginning of a side which may sound worse with certain alignments which favour end of side.
5. The cartridge offset angle need not be the same as the tonearm headshell offset, nor the mounting distance be the same as the specified tonearm alignment should another be used.

In all this it is illustrative to play with the fantastic calculators on Vinyl Engine which allow a reverse engineering of the geometry to try non-Lofgren alignments and look at how small are the variations in distortion with small changes in nulls. And how quickly the distortion rises on the inner grooves with LofgrenB. All a lot easier than calculating in the old way...

I got it. I have to do alignment/arm adjustment each time and somewhat differently before I play a record because they are all in one or more ways different. That's, of course, if I want the best results. And the better I can hear things the worse it is for me. Now that's a big fun and true joy.
Welcome to the audiophile madness club, everyone!
Dear John, If the designer is free to choose between Lofgren and non-Lofgren geometry than he obviously will do this according to his own preference or assumptions regarding the (potential) customer.But this way he already preselected the 'optimal adjustment' for the customer?
Kessler&Pisha stated in the context of 'Overhang' that many
Japanese designers 'overlooked' the dependance from or correlation with the eff. lenght. So, for exapmle, Dyna 505 has 15mm overhang while 17.164 should be optimal. Ie
the eff. lenght corresponding with 15mm overhang is 274mm
(10.787 in.). This lenght will not fit the turntables disigned for 9" tonearm ( Tonearm geometry and setup; Audio, January 1980). My 'intuition' is : you can't have
it both ways; the choice for the designer as well as the choice for the user. Ie the choice of the designer limits the choice of the user. As far as I know our choice for
Bearwald means the least distortion for the whole record
radius (aka 'avarage') while Stevenson means the least distortion near the lead out 'part'.
This means to me 'one way or the other' and this precludes
our 'free choice' or at least limits our choice. Otherwise
I am not able to make any sense of 'different tonearm geometry'.


For those who think to be good in math but also think to
have good nerve there is an article from Stereophile by
Keith Howard on the net: ARC ANGELS; Optimizing Tonearm Geometry. Look for : tonearm geometry.
I myself should keep my promisse to myself made after
my high school : 'never ever mess with math '.

What a cool question.
Ebm, To which question are you refering?
Dear Nandric,

If the arm has a range of adjustments then the alignment can be whatever the user chooses - in that sense the arm does not drive the geometry.

The Dynavector arm in your example could be mounted a couple of millimetres closer to the spindle and so have the "correct" overhang, as overhang is simply effective length minus mounting distance. The question then is: can the cartridge be aligned to the "correct" angle?

Had the headshell a facility for doing this, then there is no problem. The more scope for adjustment, the more the onus for choosing an alignment falls upon the user, as it should, given that it is his or her own records that are going to be played.

For me, as a design factor, the chosen alignment is the least of it, there are other far more relevant aspects to consider which are more important to the function of the arm. Of course there are always useful features which might conflict with other desirable features, perhaps, for example, ease of adjustment versus rigidity.

Regarding choice of alignment, LofgrenA/ Baerwald gives the lowest distortion for the 3 maxima, and Lofgren B, lowest average distortion across the side. Stevenson IEC is a special case of LofgrenA, in effect with an inner radius of around 55mm, making the inner null at the IEC minimum. Choosing different null points means that there will be an equivalent LofgrenA or B to match, except that the inner and outer radii to obtain the lowest maxima or lowest average distortion may vary from normal. For example, Dertonarm's Uni-DIN uses nulls of 109.52 and 63.49. This is equivalent to inner and outer radii of 58.4 and 129 for Lofgren A, which shows that beyond 129mm the distortion increases for an LP towards the outer radius.

If you have a wide ranging collection of LPs, then it is reasonable to use an alignment which takes all this into consideration. Personally, I would chose an alignment such as LofgrenA DIN, as it is a reasonable compromise, and try to align the cartridge as well as possible. Then forget about it (unless your favourite old record distorts on the inner groove. At which point you make a null there, and play with the calculator until you get something that is acceptable for the rest of your collection...)

As an aside, in this regard, with the potential of computer generated graphics, I am surprised that we are all still aligning to the nulls, given the surfeit of protractors and super protractors on the market.

I'd have imagined that by now we should have had a protractor with the adjustment grids at the minimum and maximum radii, offset to the appropriate angle to give the nulls in the correct location for the chosen alignment. The further apart the alignment points, the more accurate the alignment Perhaps one of the younger computer literate gurus could oblige.

Dear Gordon, Thanks for your explanation. In some earlier
thread you refered to my 'point' about different tonearms
each of them adjusted to the 'needs' of the LP's in our
possession. Your arguments in this thread assume just one
tonearm in correlation to the 'best' possible adjustment.
I have no intention whatever to spare you but on the contrary will try to get as much info from you as I can . We are allowed to ask questions in this forum I think.
My first is reg. those max distortions. You forget to mention how 'bad' 1% or more is. Then there are many of us with two or more tonearms. My quess is that they all are adjusted in the same way. Ie with the same geometry. From what I think that I can deduce from your arguments we should adjust them differently. Dependig on the records we

Kind regards,
Dear Nikola,
My comment a while back was rather tongue in cheek, not myself having a surfeit of decks.

I make do with one set up and don't worry too much about it, though I do like my anti skate set up well...

"Best" and "bad" are a different matter. Especially regarding distortion, as is well documented. So 1% (or 0.5% or 0.1% or less) can be regarded as acceptable or not. I suspect that vertical mistracking of one form or another rather than distortion per se due to the horizontal alignment is the main culprit regarding inner groove distortion, that and poorly set up anti skate.

However if I had four decks all identical, I may well set them up with different alignments and mark the albums appropriately. Or even completely different decks to suit different types of music. Why not?

But I doubt it. I would probably sell three of the decks and buy a guitar...
Dear John, There is this hilarious English phrase for the
German tourist: 'Don't mention the war!'
Well you deed mention the anti-skate...The most controversial subject in our forum according to my knowledge. My 'Solomon' solution is 'in between'. I use my test-records to get some idea about tracking ability but never push above 60 micron. Ie if there is no 'buzz' from the at 60 micron I live the anti-skate 'there'. I come a warning from Van den Hul across not to try to get,say, 80 microns which the cart can 'perform' because
one need to increase the anti-skate to get this 'result'.
Something like : better no anti-skate at all than to much.
What is your method?

Kind regards,
Nikola, I normally use a test record with tracks of increasing level, and listen with headphones, as that way I can hear better when there is mistracking. You have to hear the mistracking to know when it is equal on both channels. then re-adjust the VTF.

Having said all that, if, with whatever set-up, the sound feels solid and the image doesn't wander or get a bit watery, then fine. But I still feel more comfortable knowing that the forces on the stylus are equalized as best they can be, as the skating force is always there as long as you have a VTF
Dear John, Thanks for the x time . 'As skating force is
always there...I feel comfortable that the forces on the stylus are equilized AS BEST THEY CAN BE'. You are of course refering to the fact that (anti)skate force is variable across the LP radius. To my knowledge only Sony
provided for a 'varible force' depending on the radius(?).
There are also some claims for the Triplanar (which I own) reg.
variable skate force but I can't see that, so to speak.
Is the variable anti-skate in this sense so difficult to design? The ususal 'solution' is just a (small)'hang'- weight.



A system like the Sony would be ideal especially if it was programable, such that the force could be adjusted to vary at different points - I don't know if this is possible.

What is perhaps more interesting is that a mechanism of that kind might be used "in reverse" to measure the varying inward force trying to rotate the arm, and the settings stored for the next time the record is played.

As you have steered the thread towards antiskate, can I recap and condense what I said in another thread?

The skating force basically follows the tracking error curve of the arm (not the tracking distortion curve) which varies across the record. staying more or less constant within a few percent of around 25% VTF, reducing slightly during the first half of the side.

Depending on the overhang used and the arm effective length, at the inner null the force starts to increase rapidly. So it would appear that the ideal would be to follow this curve, reducing slightly then increasing.

However, it has been shown that there are other factors which increase the friction force as the radius decreases, which would skew this result and make it more likely that the force should at least be constant then increase, or even gradually increase all the way across the record. When I designed the RP1 that's what I did, as I thought that, like tracking distortion it seemed to be worse towards the inner grooves, and less of an issue further out, and should be weighted that way. My tonearms had an antiskate mechanism, using lever and thread, which allowed for varying the force, and the ratio to some extent across the record, or it could be removed completely. But it was not any more sophisticated than that - it was all "suck it and see". So, really, a tonearm should perhaps give the user the option of trying different things.

The actual amount of antiskate needed is very much dependent on stylus profile and downforce, as the skating force is proportional to VTF. Obviously, with enough downforce the stylus will never meet a modulation large enough for it to lose contact with the outside wall, or even throw it up out of the groove. However there will still be unequal forces on each side of the groove unless there is some kind of compensation. Does it exist as a force. It exists, yes, it definitely does. Whether its effects are noticeable to the listener is one thing, but why some people prefer no antiskate, or removal of the mechanism, that is for them to say. But the force is still there, and it has to be dealt with somehow. As Dertonarm says, a longer arm has less skating force than a shorter one. In that, we agree.

So as I see it, the bottom line is that there is a turning force (a torque) which tries to rotate the arm inwards. This torque has the effect of reducing the VTF on the Right channel and increasing it on the Left. Without compensation, one channel has more VTF than the other.