Thanks for that!
My inner groove verticles look like do-do, I need to figure out why. I might try the 10Meg loading and measure the cable C (though I figure it's around the 150pF mark).
EDIT - Due to the nature of tip geometry I don't think the TEK scope rise/fall time computation is accurate (it assumes a nice R/C shape). Or more accurately the implied BW from the rise time would be far off.
My inner groove verticles look like do-do, I need to figure out why. I might try the 10Meg loading and measure the cable C (though I figure it's around the 150pF mark).
EDIT - Due to the nature of tip geometry I don't think the TEK scope rise/fall time computation is accurate (it assumes a nice R/C shape). Or more accurately the implied BW from the rise time would be far off.
Last edited:
Scott's square waves
Scott,
In addition to indicating something that resonates at 40kHz, there are other features of your square wave plots that look interesting and could have mechanical explanations.
For instance, three of the four plots show oscillations of greater amplitude on the foot of the square wave than the crown. As I understand it, the square wave output is generated by cutting a ‘triangular wave’ into the vinyl. Let’s assume that tracking and skating forces are static over the period of measurement. That would mean that the vector sum of the forces on the stylus (longitudinal, lateral and vertical) must add to give the same magnitude vector sum as the tracking and skating forces, though acting in the opposite direction.
Those stylus forces aren’t static – the stylus moves – they vary in direction and (it appears) magnitude. In fact, when tracking a triangular cutting, it is reasonable to expect that the lateral and vertical components of stylus forces and velocity will be different when the stylus traverses an “up” slope compared to when it traverses a “down” slope. That might help explain the different oscillations on the foot and the crown of the square wave.
In the same vein, the difference in the output from inner and outer grooves could be due to differences in skating force. As I understand it, Kogen found that skating force varied with radius (though different experiments have produced different results!). Maybe that’s true of your set-up playing that disc!
Scott,
In addition to indicating something that resonates at 40kHz, there are other features of your square wave plots that look interesting and could have mechanical explanations.
For instance, three of the four plots show oscillations of greater amplitude on the foot of the square wave than the crown. As I understand it, the square wave output is generated by cutting a ‘triangular wave’ into the vinyl. Let’s assume that tracking and skating forces are static over the period of measurement. That would mean that the vector sum of the forces on the stylus (longitudinal, lateral and vertical) must add to give the same magnitude vector sum as the tracking and skating forces, though acting in the opposite direction.
Those stylus forces aren’t static – the stylus moves – they vary in direction and (it appears) magnitude. In fact, when tracking a triangular cutting, it is reasonable to expect that the lateral and vertical components of stylus forces and velocity will be different when the stylus traverses an “up” slope compared to when it traverses a “down” slope. That might help explain the different oscillations on the foot and the crown of the square wave.
In the same vein, the difference in the output from inner and outer grooves could be due to differences in skating force. As I understand it, Kogen found that skating force varied with radius (though different experiments have produced different results!). Maybe that’s true of your set-up playing that disc!
I don't know what the c 40kHz monotonic content is, but it is big and unusual IME. Because the cart coil output is velocity sensitive, a generator doesn't have to vibrate with much amplitude to produce big content. I suspect that result might be cartridge/setup specific, and many sources seem possible IMO, including an electrical resonance.
Triangular groove shapes are great for analysing tracking mechanics, because they are simple, repetitive, produce constant velocity in both directions, and have demanding curvature in both direction at both apexes. Velocity sensitive cartridges transcribe them as square wave outputs.
groove angle = velocity = output level
groove curvature = acceleration = output slew rate
Attached is a old proof of mine from the archives that relates slew rate to acceleration and curvature. It also shows that, for a given programme content slew rate, curvature increases with the inverse square of linear velocity ie spindle radius.
That's all there is to it !
Where there are instantaneous deviations from an ideal square (trapezoidal) cartridge output waveform, one can know generator's velocity at that exact moment from the cartridge output. For example, if the output non-ideally approaches zero, the generator has unexpectedly slowed down. One can tell the exact location on the groove where this happened, and usually deduce what might be going on. Unexpected changes in cart output toward zero mean the generator has slowed down, unexpected increases mean sped up - such things either mean mistracing (stylus not following locus) or something has flexed.
Asymetrical skate force arises for 'inward' and 'outward' slopes of the triangular groove shape. This is because tonearm drag is typically applied at an offset angle, the tracking offset, say 21 deg to tangent. Say the triangle wave slopes at 15 deg to tangent, then for inward sloping sections drag force is applied at 6 deg (21-15), whereas for outward sloping sections it's applied at 36 deg (21+15) deg. This is the biggest factor in asymmetric mechanics: skate force is not instantaneously symmetrical. The outcome is why those square wave's aren't exactly symmetrical.
All very interesting, and the crux of what is really going on IMO.
LD
Triangular groove shapes are great for analysing tracking mechanics, because they are simple, repetitive, produce constant velocity in both directions, and have demanding curvature in both direction at both apexes. Velocity sensitive cartridges transcribe them as square wave outputs.
groove angle = velocity = output level
groove curvature = acceleration = output slew rate
Attached is a old proof of mine from the archives that relates slew rate to acceleration and curvature. It also shows that, for a given programme content slew rate, curvature increases with the inverse square of linear velocity ie spindle radius.
That's all there is to it !
Where there are instantaneous deviations from an ideal square (trapezoidal) cartridge output waveform, one can know generator's velocity at that exact moment from the cartridge output. For example, if the output non-ideally approaches zero, the generator has unexpectedly slowed down. One can tell the exact location on the groove where this happened, and usually deduce what might be going on. Unexpected changes in cart output toward zero mean the generator has slowed down, unexpected increases mean sped up - such things either mean mistracing (stylus not following locus) or something has flexed.
Asymetrical skate force arises for 'inward' and 'outward' slopes of the triangular groove shape. This is because tonearm drag is typically applied at an offset angle, the tracking offset, say 21 deg to tangent. Say the triangle wave slopes at 15 deg to tangent, then for inward sloping sections drag force is applied at 6 deg (21-15), whereas for outward sloping sections it's applied at 36 deg (21+15) deg. This is the biggest factor in asymmetric mechanics: skate force is not instantaneously symmetrical. The outcome is why those square wave's aren't exactly symmetrical.
All very interesting, and the crux of what is really going on IMO.
LD
Attachments
Last edited:
Looking at the Shure plots and given the fact that the V15 was famed for a flat FR when loaded correctly that looks close to how I would expect it. Attached is the Bob Cordell simulation of the V15 from LA vol 4 which shows a teeny peak above 10kHz.
As the load R goes up then you would expect a larger HF peak and we see that.
I don't understand the tails on some of the inner groveplots tho, like some odd phase shift going on. Mistrack, or the 29KHz resonance of the V15?
All fascinating
BTW blast from the past ebay auction of the day. Technics SL150 Mk2 Black Widow Tonearm Shure V15xMR | eBay
Infinity black widow +V15. Before I got in to this stuff, but suspect that might be as good as it got, before the baby, bathwater and bath were thrown out the window.
As the load R goes up then you would expect a larger HF peak and we see that.
I don't understand the tails on some of the inner groveplots tho, like some odd phase shift going on. Mistrack, or the 29KHz resonance of the V15?
All fascinating
BTW blast from the past ebay auction of the day. Technics SL150 Mk2 Black Widow Tonearm Shure V15xMR | eBay
Infinity black widow +V15. Before I got in to this stuff, but suspect that might be as good as it got, before the baby, bathwater and bath were thrown out the window.
Attachments
Hi Scott,
The thin wiring, connecting your Cart to the external cable can easily have a C of 100pF, so in your case the sum could be as high as 250pF.
In the meantime I have figured out that John was using a TEK 2002.
The rise and fall times of John's recordings have roughly the same sub 20usec as yours. But John's TEK 2002 has an 8 bit A/D, giving less amplitude resolution as your recordings.
On the other hand, your Tascam equipment has a -3dB bandwidth at 40Khz, very close to the resonance in your recordings, whereas the TEK has a much higher Bandwidth of 70Mhz.
So both recordings are somewhat restricted either in amplitude resolution or in bandwidth.
But at the end it is obvious in John's recordings that the load resistor and Cabling C do play an important role and should be properly chosen to isolate mechanical from electrical resonance.
Hans
It is obvious that loading is important, which isu why its worth revisiting the damped loading concepts up to and including LDs transimpedance input.
But also I think this proves that, for good cartridges (V15V, AT, Ortofon etc) the mech resonance does not appear to be an issue that needs correcting for.
The unanswered question is whether flat to 40kHz response is helpful. MC advocates sometimes claim that is why their preferred cartridges are better. I am starting to think that the playing field is level
But also I think this proves that, for good cartridges (V15V, AT, Ortofon etc) the mech resonance does not appear to be an issue that needs correcting for.
The unanswered question is whether flat to 40kHz response is helpful. MC advocates sometimes claim that is why their preferred cartridges are better. I am starting to think that the playing field is level
The liner notes point out that this LP was cut at 15 degrees and that apparently several of the others in the series were not. Makes things even more confusing. I will have another picture of the 300Hz verticle sine wave to show that I can't figure out.
The lateral 300Hz tracking tests had warp sensitivity (they go up to ~90u peak displacement) and the record unfortunately has some visible warp. I assume this is tracking force modulation again.
EDIT - Another note they make is that since there is no equalization these LP's can be used at any TT speed with frequency/velocity scaling appropriately noted.
Last edited:
If those measurements are good, 20uS rise time on that STR112 test equates to about 500G acceleration. Close to the limit of tracing performance for many setups but perhaps that's the point.
500G acceleration equates to groove curvature radius of about 54um at a spindle radius of 15cm (outer tracks) and about 9um curvature radius at 6cm (inner tracks), at 33 rpm. Tight for inner tracks, and 500G involves relatively big forces for things that flex........sometimes slew rate is limited, sometimes the stylus lets go and there is a recovery period for a short period after the acceleration.
There is another mechanics aspect which is seldom (never?) mentioned: 'jolt'. The rate of change of stylus acceleration, or in groove shape terms, the rate of change of curvature, which is fierce in this test at the apex. In signal terms, the rate of change of slew rate.
LD
Last edited:
Did the STR120 500Hz-50kHz constant velocity today after checking my electronics. The recorder measured -2dB @42kHz and aliased nicely from a signal generator
so since it's only capable of 96k I figured I would still be able to detect any sharp resonances out to 50k using a single swept tone.The result was interesting, no sharp resonances but some 2-3db soft peaks and the output went to 48k and turned around back to 46k just as expected.
So for Bill this might be a better choice especially using 192k sampling. These tracks are repeated outer, mid, and inner at 8dB lower level, they say to make the minimum radius of curvature equal to a "typical' 18 micron stylus @20kHz. Maybe LD can comment.
This Grado measured like most I have had 2-3dB dip at 5k back to +0.5 at about 14k then a slow roll-off L, R, lateral, vertical, no difference. No idea where this comes from. I could wildly speculate about a lot of popular mics that have presence peaks at 5k, a recording engineer friend immediately commented after about 10 seconds listening to some of my bird recordings. It's probably just the way I set them up.
Last edited:
Yes, there are several terms for it in use.
It doesn't help that there is no defined meaning of a constant relating jolt to force..........classical mechanics stopped there, like a book missing the last pages.
Displacement F = kx (spring constant)
Velocity F = kZ (mechanical impedance)
Acceleration F = ka (mass)
Jolt F = kJ (?????)
Yet we all know what jolt feels like, and it has real effect on mechanics of systems, even here..........but there is no constant defined AFAIK. It would have units Nm-1s3.......so what would its properties be, how would it feel, what would it be called? I think we all know what it feels like to increase the acceleration of a car, for example......
LD
Last edited:
Mathematically the second derivative of the motion is singular, in reality I guess you need to integrate across the singularity conserving energy. In SPICE some tricks are needed to conserve charge or using ideal sources in a sim can create singularly impossible voltages/currents.
Last edited:
Now I may be being thick here, but the 'jolt/jerk' we are talking about is when lateral stylus velocity is zero (so like in the standard textbook pendulum where kinetic energy is zero and potential energy is maximum)? I am sure there is a term for it for racing shock aborbers, where hitting the curbs is an important trade off to adjust for. I have a friend who used to design them as well as doing mad things like working out how to get tanks to pull 1G on cornering. will ask him.
I would think it was like bottoming out your suspension in a pot hole. The energy lost in the rubber stop blocks, etc. is a very non-linear function and happens in a short time. The net result is that a big enough pot hole and there is catastrophic failure. Bent pitman arms, broken ball joints, bent rims, here in New England you get them all.
Last edited:
Hydraulic bump stops help a lot there as they are adjustable! But not fitted to road cars. Of course its not the wheel going into the hole that is the problem, its coming back out of it!
Still needs the same suspension solution (acceleration sensitive) so its very compliant the higher the acceleration is and becomes stiffer as acceleration decreases. Active suspension cartridge anyone? Suddenly optical reading looks easy.
Edit: Pseudoplastic in the cartridge suspension might help. Shear thinning would make compliance higher as acceleration increases. A few engineering challenges to keep it in place tho. Sure someone must have done that at some time...
Still needs the same suspension solution (acceleration sensitive) so its very compliant the higher the acceleration is and becomes stiffer as acceleration decreases. Active suspension cartridge anyone? Suddenly optical reading looks easy
.Edit: Pseudoplastic in the cartridge suspension might help. Shear thinning would make compliance higher as acceleration increases. A few engineering challenges to keep it in place tho. Sure someone must have done that at some time...
Last edited: