I cannot speaker for Frank but judging by the picture I would think so. It's not that different from the Thales tonearms, instead of having a pivot at the headshell, the pivot now is at the base. Again, similar to the Birch design. It's been done before.... if not commercially. The swiveling base is only responsible for horizontal movement so it belongs to the genre what Mark Kelly called "split plane" design, except it has an extra pivot and the armwand is longer. I've been thinking about this type of design for over a year but Frank beat me to it to actually have a finished product. I have my own idea of how to do the geometric guiding so I am curious to see how Frank implement his. Since the pivoting base does not form a straight line with the cartridge cantilever it will exhibit skating force, as I'm sure Conrad Hoffman would chime in to say the same thing. The only one that I know of that has no skating force is the design by member Ralf, aka "Straight Tracker." Anyway, it's nice to see some fresh ideas.
.
No, it won't. If Frank has done what I think he's done, there will be no skating force.
Thanks for chiming in, Mark. Hmm... so if the guiding mechanism includes point B, one of the three points inside the Thales semi-circle, the arm will automatically have no skating force, even if point P is doing the pivoting? Inquiring mind wants to know.
Reusing the image from the Thales design.
An externally hosted image should be here but it was not working when we last tested it.
Reusing the image from the Birch design.
An externally hosted image should be here but it was not working when we last tested it.
.
Hi guys,
Sorry, I haven't had the time to pay much attention to any forum, even this one(my favorite).
The arm shown was my second choice in terms of building a pivoted linear tracking arm. I had conceptualized an arm similar, but simpler than Ralf's(straight tracker), when I found out that he had patented and build it a looong time ago. While my implementation was different in several ways, there were enough similarities to not continue with it, avoiding accusations of plagiarism. Instead, I contacted Ralf and have met him a few times since, discussed his design with him and hopefully helped him a little on the way to commercializing his design(he deserves the recognition).
The arm shown bears no relation to the Thales arm(it is not based on the Thales circle !). As you may have noticed, there is no offset(unlike ALL - but Ralf's - pivoted linear trackers). The displacement required for an arm of 250mm eff. length is only ~30mm. There is only one "conventional" additional bearing that "adds" friction, but only in the order of 2milligrams. The magnetic guide rail system is frictionless, so that the total force required to displace this arm is ~ 5mgr. in the horizontal plane and 1-2mgr(possibly zero in a commercial product) in the vertical plane. It's a close to neutral balance design, a VTA change doesn't alter VTF. At first it may appear as if there ought to be a skating force present - and there arguably can be, but the arm/cartridge doesn't exhibit a tendency to move inwards when properly set up. The cantilever is NOT subjected to any side thrust other than what results from bearing friction and the minute influence of the wiring.
The eff. mass is higher in the horizontal plane off course, but only by about 20-30%(depending upon the mass and position of the counterweight and the damping ratio), causing the arm cartridge resonance frequency to differ by about 1 Hz, lateral vs. horizontal. Damping for the lateral "beam" bearing and the magnetic guide rail is adjustable.
I checked for years what must have been a hundred patents incl. the ones you have listed and couldn't find any that use a magnetic "guide" rail nor any that are based on the geometry chosen. That's why I decided to apply for a patent.
I will be taking the prototype to this years ETF in France( a good week from now) so maybe some of you get to hear and see it
If I find the time, I'll be bringing a second deck with a different pivoted linear tracker to the ETF, cheap and easy to build.
Cheers,
Frank
Sorry, I haven't had the time to pay much attention to any forum, even this one(my favorite).
The arm shown was my second choice in terms of building a pivoted linear tracking arm. I had conceptualized an arm similar, but simpler than Ralf's(straight tracker), when I found out that he had patented and build it a looong time ago. While my implementation was different in several ways, there were enough similarities to not continue with it, avoiding accusations of plagiarism. Instead, I contacted Ralf and have met him a few times since, discussed his design with him and hopefully helped him a little on the way to commercializing his design(he deserves the recognition).
The arm shown bears no relation to the Thales arm(it is not based on the Thales circle !). As you may have noticed, there is no offset(unlike ALL - but Ralf's - pivoted linear trackers). The displacement required for an arm of 250mm eff. length is only ~30mm. There is only one "conventional" additional bearing that "adds" friction, but only in the order of 2milligrams. The magnetic guide rail system is frictionless, so that the total force required to displace this arm is ~ 5mgr. in the horizontal plane and 1-2mgr(possibly zero in a commercial product) in the vertical plane. It's a close to neutral balance design, a VTA change doesn't alter VTF. At first it may appear as if there ought to be a skating force present - and there arguably can be, but the arm/cartridge doesn't exhibit a tendency to move inwards when properly set up. The cantilever is NOT subjected to any side thrust other than what results from bearing friction and the minute influence of the wiring.
The eff. mass is higher in the horizontal plane off course, but only by about 20-30%(depending upon the mass and position of the counterweight and the damping ratio), causing the arm cartridge resonance frequency to differ by about 1 Hz, lateral vs. horizontal. Damping for the lateral "beam" bearing and the magnetic guide rail is adjustable.
I checked for years what must have been a hundred patents incl. the ones you have listed and couldn't find any that use a magnetic "guide" rail nor any that are based on the geometry chosen. That's why I decided to apply for a patent.
I will be taking the prototype to this years ETF in France( a good week from now) so maybe some of you get to hear and see it
If I find the time, I'll be bringing a second deck with a different pivoted linear tracker to the ETF, cheap and easy to build.
Cheers,
Frank
Not automatically, no. If all the pivots are free pivots and the relevant arm sector is free to move longitudinally through the distal vertex of the Thales semicircle* then there will be a skating torque around the main pivot and furthermore it will vary greatly as the effective length changes. This difficulty held up the development of my split plane arm for quite some time.
I finally twigged that these conditions are not necessary. If I am correct, Frank has used a similar technique - I really hope it's not the same! I don't think it is - I played with purely magnetic systems but I couldn't find a design that didn't fall foul of Earnshaw. Frank is a genius with magnetic systems, so I'll be interested to see how he solved this problem.
*Probably better to call it the "pseudo Thales semicircle". My split plane design is based on a geometry similar to that shown in the Birch patent. I can show mathematically that this is related to, but not identical with, the Thales geometry. I am not privy to the details of Frank's design but I think I know enough about these things to deduce the geometry used from the illustrations given. I would have said that Frank's geometry is based on a similar but different modification of the Birch geometry. The function of the magnetic guide system follows from this geometry.
I will however leave it up to Frank as to how much he wishes to reveal about his system.
Glad to see you here at your favorite forum, Frank.
We would still love to see your first choice LT arm, even if it's abandoned so we can get our heads exploded. :-D I know it's different from the Thales design but don't you at least use the Thales circle to confirm the geometric accuracy, those three points around the semi-circle, or you just check tangency at each point of the record? I know the Birch geometry can be accurate only within the record range but not all the way into the spindle. I really like not having a pivot above the stylus as required quality of such bearing is paramount and still cannot be better than no bearing at all at such delicate area of the tonearm. You and Ralf deserve credit for thinking outside of the box.
On a Garrard Zero style design I thought of using magnet as guiding mechanism but it's just too close the cartridge to be effective. Using close to the base makes total sense to me. I have my own idea of a guiding mechanism, with no magnet, on a Birch style design that I hope won't conflict with any patents out there...
Look forward to seeing it! Thanks for chiming in, Frank!
------------------------------------------------------------------
Thanks for explaining, Mark. The effective length change is a bitch! I know I have tossed enough sketch paper because of that!
I am flattered that two of the greatest minds of audio is on this thread.
Agree. If there's a person who can make magnet work, it would be Frank.
For now, I will just called this "pseudo Thales semicircle" the "Birch geometry" and it's one concept that helped me understand the possibility of an alternative to pivoting headshell. It's also great to know from you that skating force is not a big issue with such design as I almost abandoned the idea. It's very simple and elegant and it might not get the kind the attention like the Thales arms. I like simplicity, even if some members don't believe me.
Now, I wish Ralf, Straight Tracker, is here to add his opinion.
.
Hello Mark Kelly
Twigged? I'm always ready to learn new words. Is this the way Elmer Fudd would say "Trigged"?
Sincerely,
Ralf
Twigged? I'm always ready to learn new words. Is this the way Elmer Fudd would say "Trigged"?
Sincerely,
Ralf
Here's a more elaboration on the drawing of the Birch geometry. I hope this makes sense to those who are confused by the Birch patent.
.
An externally hosted image should be here but it was not working when we last tested it.
.
And here's how to find point P:
Take three points on the Thales semicircle (the arc from B to A). Draw the chords from each of these points back to the distal pivot (B).
For a nominal second arm length L2, find the point on each chord which is L2 distant from the end nearest A, back towards B. This is the part of each chord which is shown in the diagram above. Connect these three points with line segments. Bisect the two line segments and find the intersection of their bisectors. That will be the point P.
You can do this on paper or by trigonometry and algebra. I do the second and have loaded it into an Excel spreadsheet. To make it easy, assign either A or B as the point 0,0 and then define each other point in X Y co-ordinates from there.
When you do this you will find that Birch's geometry is less than exact and that certain combinations of Thales circle diameter and second arm length result in much smaller errors than others. The best combinations give errors of less than 0.1 mm for all possible stylus positions. This is effectively zero as I do not know of any practical way to achieve that accuracy with cartirdge alignment.
Take three points on the Thales semicircle (the arc from B to A). Draw the chords from each of these points back to the distal pivot (B).
For a nominal second arm length L2, find the point on each chord which is L2 distant from the end nearest A, back towards B. This is the part of each chord which is shown in the diagram above. Connect these three points with line segments. Bisect the two line segments and find the intersection of their bisectors. That will be the point P.
You can do this on paper or by trigonometry and algebra. I do the second and have loaded it into an Excel spreadsheet. To make it easy, assign either A or B as the point 0,0 and then define each other point in X Y co-ordinates from there.
When you do this you will find that Birch's geometry is less than exact and that certain combinations of Thales circle diameter and second arm length result in much smaller errors than others. The best combinations give errors of less than 0.1 mm for all possible stylus positions. This is effectively zero as I do not know of any practical way to achieve that accuracy with cartirdge alignment.
Last edited:
Sorry for the crude drawing as my photoshop skill is amateurish. I added all the C points to the "chords" to illustrate better.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Of course the shorter the chords are the bigger the guiding pivot semi-circle will be and the more accurate it will be. And if chord is at the midpoint of the Thales semi-circle then, voila, you have a Thales tonearm! Of course, you're back in the soup and end up guiding the headshell again. What a circle jerk of this hobby!
I am still trying to wrap my head around how to define point P without doing it on paper and much of my math has left at the classroom.... I need to buy a beer for my math savvy friends.
-------------------------------------
.
I am still trying to wrap my head around how to define point P without doing it on paper and much of my math has left at the classroom.... I need to buy a beer for my math savvy friends.
-------------------------------------
An externally hosted image should be here but it was not working when we last tested it.
.
Last edited:
Hi directdriver,
Just loose the idea of having to adhere to the guide rail having to be a (perfect)circle segment. You'll be gaining the freedom to chose P somewhat arbitrarily.
I'd also suggest to look at my solution as much closer to Ralf's concept vs. the Birch concept. The "main" arm(not the pivoting lever that holds the main tonearm) is moved in accordance with its magnet guide rail system, but it is not rigidly coupled! For minute movements in the lateral plane it will act like a regular arm with no offset angle and a fixed footpoint. But, as said before, the "rigidity" of said magnetic coupling is adjustable and so is the damping on the pivot bar lever.
The initial, active, arm I worked on before this one did move the footpoint along a straight line, but that line was pointed at a radius midpoint between outer and inner groove. If you think about it, any movement of the main bearing(along the straight line) would cause the lowest average side thrust/deflection of the cantilever(a terrible flaw of conventional servo driven linear tracking arms). More importantly, it derived its positioning from the arm's angle of deflection, instead of having to add an additional cammed "reference" sled to use a deviation from zero type of servo.
I'm afraid B is not where your arrow is pointing. The actual first prototype had the pivoting bar pointing in the opposite direction, - and the guide rail had a totally different profile/curve...
I'm really busy getting my act together for the ETF, so it might take some time, but I may post some drawings regarding possible layouts once back home.
Cheers,
Frank
Just loose the idea of having to adhere to the guide rail having to be a (perfect)circle segment. You'll be gaining the freedom to chose P somewhat arbitrarily.
I'd also suggest to look at my solution as much closer to Ralf's concept vs. the Birch concept. The "main" arm(not the pivoting lever that holds the main tonearm) is moved in accordance with its magnet guide rail system, but it is not rigidly coupled! For minute movements in the lateral plane it will act like a regular arm with no offset angle and a fixed footpoint. But, as said before, the "rigidity" of said magnetic coupling is adjustable and so is the damping on the pivot bar lever.
The initial, active, arm I worked on before this one did move the footpoint along a straight line, but that line was pointed at a radius midpoint between outer and inner groove. If you think about it, any movement of the main bearing(along the straight line) would cause the lowest average side thrust/deflection of the cantilever(a terrible flaw of conventional servo driven linear tracking arms). More importantly, it derived its positioning from the arm's angle of deflection, instead of having to add an additional cammed "reference" sled to use a deviation from zero type of servo.
I'm afraid B is not where your arrow is pointing. The actual first prototype had the pivoting bar pointing in the opposite direction, - and the guide rail had a totally different profile/curve...
I'm really busy getting my act together for the ETF, so it might take some time, but I may post some drawings regarding possible layouts once back home.
Cheers,
Frank
I just reread what I wrote. To avoid confusion, I should add that the pivoting bar does move in accordance with the guide system too, but, for a lack of better words, the main arm moves first, the pivoting bar follows. That would not be the case if all elements were rigidly(lateral bearings aside) connected. Then, all masses involved would move in unison.
Frank
Frank
Hi Frank
Nice to know you also share.
So if I understand correctly, the pivoting lever that holds the tonearm is able to move but in a restricted way. The freedom of movement being set by the stiffness of the magnetic bearing. Did you fine tune it in order to avoid forward, backward movements of the tonearm depending on badly centered discs ?
You're right. Since I do not have a picture of overhead shot of the arm so I intended arrow B to be below the counterweight. It should have NO offset angle. B and C should ALWAYS form a straight line. Sorry for the confusion and thank you for the clarification.
I can't wait to see more drawings and pictures!
Thank you so much for taking your valuable time to answer my silly questions.
An externally hosted image should be here but it was not working when we last tested it.
.
Last edited:
Hi Ricardo,
I'm happy to share as much information as I can afford, in all senses of that word. This arm will be commercially made and has been licensed, so I'm sure you'll understand why I can't publish blueprints(I'm exaggerating, off course
I wouldn't use the term "freedom of movement" when it comes to describing the actual movement of the pivoting bar in operation, but I guess one could put it that way. A simple looking system, but the interplay is complex and can be influenced or yes, finetuned. Excentric discs will nevertheless cause wow, just as with any other tonearm. I don't think there is a way around that one, except for centering the disc.
Best,
Frank
I'm happy to share as much information as I can afford, in all senses of that word. This arm will be commercially made and has been licensed, so I'm sure you'll understand why I can't publish blueprints(I'm exaggerating, off course
I wouldn't use the term "freedom of movement" when it comes to describing the actual movement of the pivoting bar in operation, but I guess one could put it that way. A simple looking system, but the interplay is complex and can be influenced or yes, finetuned. Excentric discs will nevertheless cause wow, just as with any other tonearm. I don't think there is a way around that one, except for centering the disc.
Best,
Frank
Here's how to do the classic Birch geometry without paper.
Choose the Thales semicircle radius, say it's 150*. The beginning is (0,0) (point B), The centre of the semicircle is (150,0) (midpoint). The end is (300,0) (point A, the spindle). Diameter = 2* radius.
Each stylus contact point forms a minor chord equal to the groove radius at that point and a major chord. We find the length of the major chord by Pythagoras so it's SQRT (diameter^2 -minor chord^2). For three groove radii, calculate the length of the major chord. Use three practical radii, say 150, 100, 50. The corresponding major chord lengths are 259.8, 282.8, 295.8**
Now choose the length of the secondary arm and subtract this from each majorchord length. Say our secondary arm is going to be 150 mm long, the remainders are 109.8, 132.8, 145.8. These are the distances from pivot B to the points which will define the carrier arm arc, the centre of which is P.
To find the co-ordinates of each point, we first use the property of similarity of triangles and observe that the ratio of the major chord length to the Thales circle diameter (the sine of the angle at B) is identical to the ratio of the horizontal displacement of each point from B to the remainder length. Thus each x co-ordinate is (remainder * major chord / diameter). The vertical displacement is found by Pythagoras. Our three points are thus (95.1, 54.1), (125.2, 44.3), (143.7, 24.3).
These three points define two line segments (1 -2 and 2 - 3). The slope of each segment is given by (y1 - y2) / (x1 - x2), substitute 2 and 3 as required. We want the slope of the orthogonal intercept, which is -1 / slope segment, which is therefore (x2 - x1) / (y1 - y2). Read that again carefully and notice that the Xs swapped positions and the whole thing was inverted.
Our two intercept slopes are 2.84 and 0.93.
The midpoints of the two line segments are (x1 + x2) /2, (y1 + y2)/2 giving (110.2, 49.6) and (134.5, 34.3). The intercept lines have the general equation y = mx + b. We already know m (the slopes we calculated above) so we simply subsitute to find b: if y = mx +b then b = y - mx so our two equations are y = 2.84 * x - 263.1 and y = 0.93 * x - 90.4.
These two lines will intercept when y = y, which is to say when
m1 * x - b1 = m2 * x - b2 which is
b2 - b1 = m2 * x - m1 * x which is
x = (b2-b1) / (m1-m2).
Substitute x back into one equation to obtain y. Do it to the other one just to check it's right.
Our intercept point is thus 90.36, -6.6. This is point P. Calculate the length of the carrier arm by finding the distance from this to our three original points, using Pythagoras. These lengths are all around 61.7 mm.
This isn't a particularly good example of the Birch geometry, you'll see that the distance from point P to the spindle is 209.8 mm but the sum of the two arms is 211.7 mm.
You owe me a beer.
Next installment (after I've been for a ride before it gets too hot and the UV peaks), I'll show you how to calculate an arm with "floating P" which is the geometry I'm using and possibly what Frank's using as well.
* All these numbers are variables, best to define a cell in Excel so they can be changed.
** I've rounded all numbers to avoid typing. Don't round in your calculations.
Last edited:
Typo: in the post above I meant to type "Floating B" not "Floating P"
A note on terminology. A chord by definition begins and ends on the circumference of the circle. Thus for a given Thales semicircle the chord for each groove radius is fixed in length. Within the Birch geometry the length that changes is that of the line segment between the end of the chord near the spindle (points C1, C2 etc on the Birch diagram above) and the intersection of the circle centred at P and each chord (which I will call points D1, D2 etc).
A better statement of your second sentence might then be "as the length of the segment CD decreases, the length of the segment PC increases and the pivot point P moves closer to the middle point of the Thales semicircle. When CD becomes zero the Birch design becomes a Thales arm"
A note on terminology. A chord by definition begins and ends on the circumference of the circle. Thus for a given Thales semicircle the chord for each groove radius is fixed in length. Within the Birch geometry the length that changes is that of the line segment between the end of the chord near the spindle (points C1, C2 etc on the Birch diagram above) and the intersection of the circle centred at P and each chord (which I will call points D1, D2 etc).
A better statement of your second sentence might then be "as the length of the segment CD decreases, the length of the segment PC increases and the pivot point P moves closer to the middle point of the Thales semicircle. When CD becomes zero the Birch design becomes a Thales arm"