[dlaloum]

I really need to spend a few quality hours catching up on that thread...

[neobop]

The thread has degenerated and J. Carr's posts are about the only thing of interest.  Most of that has been repeated here, and I posted there most of the lab test links I posted here.  I think he and I were the only ones who looked at them, and I also think you could be very affective at getting more info or opinion from J. Carr.  He provided the Accuphase sheets, BTW.

A couple of pages back he was commenting on the stylus mounting of the boron tube cantilevered carts - DL-1000a, EPC 305MC.  I found a photo of a Genesis 1000 tip, but I don't know if the link still works.  After commenting that the 1000a stylus was only mounted to the upper wall of an extremely thin tube, he proceeded to say the 305 was worse because the tip was too long and could rotate.  I thought his assessment was self-serving, especially comments about increasing tip mass.  Where does he get off talking about tip mass on a Technics cart like that, when.....?

Anyway, relations were somewhat strained but civil.  Somehow the conversation got onto phase and I didn't pursue the Ortofon papers.  He insists there's a phase shift at electrical resonance.   I still don't think that's possible based on the measurements.  Maybe I'll look at those MM graphs again.
I took a look at the inner groove thread on Karma.  Maybe I missed it, but it seems like nothing new. 
neo 

[*Scotty*]

The inductance of the MM cart alone with give you electrical phase shift, it's unavoidable unless you cancel the inductance out. The inductance plus the cable capacitance forms a second order filter which results in even more phase shift. One of the advantages of the MC design is the vastly lower inductance figure which results in less phase shift in the audio band.
Scotty

[dlaloum]

Like Scotty said - it is physics, phase shift is unavoidable in an LCR circuit...

BUT  that phase shift is identical in properties to cantilever resonance phase shift, and if you balance one vs the other, a flat frequency response is an indicator of neutral phase as well.... Hence achieveing flat frequency response using minimum phase methods is critical.

Cantilever resonance can also be modelled as a circuit I believe, at some stage I should add that to my cartridge model...

[neobop]

Like Scotty said - it is physics, phase shift is unavoidable in an LCR circuit...

BUT  that phase shift is identical in properties to cantilever resonance phase shift, and if you balance one vs the other, a flat frequency response is an indicator of neutral phase as well.... Hence achieveing flat frequency response using minimum phase methods is critical.

Cantilever resonance can also be modelled as a circuit I believe, at some stage I should add that to my cartridge model...

The LCR phase shift is NOT separate from the that of the cantilever.  If it was, there would be a shift at electrical resonance in each of the 5 MM plots in the test report.  Even though we don't know what those MMs were, there is only one shift and it coincides with HFR. 

You can look at it as combining the 2 circuits, mechanical and electrical, because that's exactly what happens.  The LCR circuit has the net effect of lowering HFR.  It's just as CarlosFM (EE) stated on the VE thread.  We can see the physical lowering of HFR in Werner's TNT article Load the Magnets.  Resistance and capacitance are manipulated to augment treble response.  That lowered frequency (16K or whatever) is where the shift occurs.

Gentlemen, we must go with the evidence we have.  The Ortofon tests are the only known measurement of phase vs. amplitude with phono carts.

For years better phase linearity was used as an argument for the superiority of MC.  Why then does Atlas, widely acknowledged as one of the best carts ever, mimic MM phase response?  With HFR @ 20KHz, phase would look more like one of the MMs in the test than the MC200.  Perhaps the strain gauge has better phase characteristics than magnetic carts, at least that's what is claimed.  But I suspect many records don't sound "right" with them.  The mastering is checked with magnetic carts and the imperfections of that playback might be built-in the record.  It's probably sort of the way it is now with some records better served with a forgiving set-up. 
neo 

[dlaloum]

Hmmm yes - we do know that where the Frequency/Amplitude (as the sum of LCR and cantilever F/A) goes off from "flat" neutrality there is definitely associated phase variance.

Given that it is a minimum phase system, it should even be calculable...

But if the response in frequency terms is flat, the then associated frequency/phase will also be flat - with the notable exception of anomalies generated as part of hysteresis, eddy currents and other such - all of which tend to be in the very small scale categories (definitely less than 3db, and in the better cartridges less than 1db) - and well below the phase anomalies caused by even relatively "perfect" phase speakers (eg pure electrostatics) - Which is why I tend to ignore these effects... (also none of the test records claim to be accurate to within 1db... so effects on this scale are not measurable with our current tools)

I have to read JCarr's postings on the effects of imperfect coil shapes, etc... and mull on that particular set of parameters...

I definitely have measured substantially reduced upper midrange losses in switching from a high inductance body to a low inductance one - while keeping the same needle (Pickering XLZ7500 & XSP3003) - so I tend to think there is definitely something that is gained in going to lower inductance.

Eddy currents and hysteresis are both heavily driven by signal strength which in turn is the level of change within the magnetic field (ie how motion is turned into electric signal)... fewer turns of the coil seems to improve things - at least in the case of my example - quite measurably.
BUT - in exchange, you have that rising top end, and the associated phase variance.... sigh.... (which can be corrected by adding more damping.... but the damping has known negative impact on the sound too.... )

Later this year I am going to try the XLZ7500 with digital minimum phase filtering to see how it sounds with flat Frequency/Amplitude as well as Phase/Amplitude response.... (one of those "round tuit" exercises I mean to get to!)

bye for now

David

[neobop]

David,
"we do know that where the Frequency/Amplitude (as the sum of LCR and cantilever F/A) goes off from "flat" neutrality there is definitely associated phase variance."

Looking at the MM plots, there is a small rise in amplitude response at, or just beyond the point of maximum phase shift.  (note: it's difficult to tell exactly, as the phase plot goes off the scale)  Past that bump in amplitude response it either rolls off or continues relatively flat.

"But if the response in frequency terms is flat, the then associated frequency/phase will also be flat"

I don't think this follows.  Just as in the MC plots, phase shift is maxed out at that peak of cantilever resonance (or combined with MM LCR).  On either side of that peak of maximum phase shift there is quite a bit of phase nonlinearity that extends through areas of flat response.  In the case of the MC200 the peak was @ 27K and phase shift didn't reach zero until approx. 7-8K. Some of the MM phase shift extended down to 1-2K.

Another problem - we didn't pay much attention to phase shift on the far side (higher frequency) of the peak.  In the case of the MC200 this is beyond 27K so it's way beyond audibility.  If the peak occurs @ 16K, then phase shift @ 17K will be just as bad as 15K. 

For years phase performance has been used as an argument for MC superiority.  "They" said that LCR determined phase shift frequency and that many MCs had no shift until way into the ultrasonic region based on MC inductance.  This could be way beyond the response of the cart.  MC inductance is measured in uH.  We know this is BS - HFR determines MC phase shift.
What about MMs when LCR can be lower than HFR ?  We don't have enough evidence to prove LCR is meaningless like in the MC case, and we know that increasing capacitance will lower HFR, and because of this the user has another feature to manipulate amplitude response.   Is the downside to this also phase performance beyond HFR ? 

Some of the MMs in the test had HFR or primary shift around 20K.  How is this any different from Lyra Atlas/Kleos ?  It might extend a little lower due to increased MM damping?  So, if we're thinking purely in terms of amplitude/phase, the Atlas is a second rate design.  +6dB @ 20K, where HFR occurs?  It shows how much importance this phase business actually is in the ears of audiofools. 
Lets forget their ears and look at "superior" carts.  How about some Dynavectors or Nakatsuka designs that have far better HF response and resonance in the stratosphere?  They're not perceived as better because of Carr's magic.  Emphasis is on tracing exactitude and excess energy dissipation.  Have you noticed that Anna, Atlas and AT50ANV have titanium or titanium in the bodies?  This is for energy dissipation - send it to the arm instead of messing up the signal in the cart.  Clean up the resonance mess and you have better sound.
neo



 

[dlaloum]

I was about to go off half cocked as usual without giving this due thought...

had another look at those plots, and indeed the 5 MM's show relatively flat F/R with quite widely varying P/R.

So I took another look at the MC plots....

Look at plot 1 and plot 2

1 is textbook minimum phase resonance - amplitude varies with a gentle bell curve and phase has a sharper bell curve with the peaks coinciding.

2 however is strange... for damping, the curves should both become wider and lower, with the peaks still coinciding.... but they don't.
A shift in overall phase peak should imply the addition of another resonance which is summing - this does not make sense...unless the damping mechanism is introducing something else... presumably it is not a "pure" damper - and may use another mechanism to provide a damping effect on amplitude.
Either something is wrong with the measurement method, or someone played rough and ready with some of the plots (perhaps measurement ended at a lower frequency than shown on the plots?) - or I need to review the physics of this effect.....

Maybe I should go ahead with my square wave based phase testing of some cartridges to get some rough experimental results....

[neobop]

Interesting observation.  The MC graphs extend out to 50K and the plots with damping peak somewhere beyond 50K.
The amplitude of the undamped phase peak is +17dB (approx.).   The damped peaks are higher in both amplitude and frequency.  The MC phase curves with damping more closely resemble the MM phase curves, except not as steep.  Remember, an undamped MC is probably a nonentity, so a real world MC actually has phase shift that covers a wider range of frequencies, except most of it is ultrasonic. 

I think the implication is clear.  Damping doesn't just cause the phase curve to extend further in either direction, it also increases the amplitude.  In other words, the curve gets bigger in all directions. 

There's another unknown here and that's the identity of the MM carts.  It looks as if response is falling off beyond 20K in all of them.  I think that's the suspicious factor here.  I suspect a low inductance MM cart with a boron or beryllium cantilever could equal or surpass the MC200 in phase performance.
neo

[dlaloum]

I'm gonna have to study some more physics to understand what could cause that rising phase/frequency plot...

Or review some formulae....

[J-Pak]

Followers of this thread may have heard of the AT33PTG II (a favorite amongst some "famous" needle droppers) MC cartridge. One of my friends added one to his system and I heard it over the weekend. Probably the best tracking I have heard/experienced from a moving coil, though may fall slightly shy of the V15. Stunningly clean on known problem records like some MFSL reissues of Bob Dylan which can have headache inducing vocal sibilance (particularly bad on their 45 rpm reissue of The Freewheelin' Bob Dylan). Sound quality was very good, but a touch too dry and clinical for my taste.

[neobop]

Hi J-Pak,
There surely is much love for the 33PTG II.  They say it digs a little deeper than the OC9II and not as aggressive.  I don't think it was ever a US model, but you can get it direct from Japan.
http://www.joynetmall.com/store/index.php?route=product/product&path=60_64_267&product_id=4124

Joynetmall has a 100% reliability rating.  Still, I don't know about warranty or exchange if it's defective.  That seems like an excellent price. 

Many people like an accurate cart, as opposed to romantic, especially those who use tubes might want a "dry and clinical" sound.
neo

[neobop]

The inductance of the MM cart alone with give you electrical phase shift, it's unavoidable unless you cancel the inductance out. The inductance plus the cable capacitance forms a second order filter which results in even more phase shift. One of the advantages of the MC design is the vastly lower inductance figure which results in less phase shift in the audio band.
Scotty

The notion of more or less phase shift is erroneous.  It's the frequency of primary shift that is affected.  I can't prove that this occurrence isn't at electrical resonance because the test carts are unnamed, but it looks highly unlikely.  A MC shift frequency is 100% mechanically dependent.  We know that MM response peak is lowered by increasing capacitance, or using electrical resonance.  I think this combination of mechanical and electrical resonance defines the frequency of the shift. 

Be that as it may, I remember your showing us how you hook up another identical cart body to cancel inductance, but it was a long time ago.  In order to cancel inductance, two identical inductors have to be directly connected in parallel. 
Here's a link to Werner's TNT article with an electrical diagram.
http://www.tnt-audio.com/sorgenti/load_the_magnets_e.html



Could you please explain it again?  Inductance is in series with the output. 
neo

[*Scotty*]

You need to start with a phono-cartridge and a cartridge body with identical inductance values. The 47 kohms series resistance is removed from the phono-stage input. It is replaced with a 5kohm series resistor,and a 5kohm resistor to ground. The phono-stage input is connected to the junction of these two 5kohm resistors.
The portion of the circuit connected to ground starts at the junction of the two resistors. The non-series connected 5kohm resistor is connected
to one of the positive pins of the cartridge body and the ground pin for that channel is connected to ground. The other channel is a duplicate of this.
 
Basically half of the cartridge's output is taken to ground from the junction of the two 5kohm resistors through the inductor in the cartridge body. This leaves on average 2.5 to 2 millivolts output which seems to be plenty for most phono-stages to work with.
Assuming that the two cartridge bodies have virtually the same inductance, the inductance of the phono-cartridge on the tonearm is cancelled out by the inductance to ground of the identical cartridge body. This gives you the same operating conditions as a MC with any phase shift in the upper mids and high frequencies a function of the mechanical resonance with zero phase shift due to electrical RCL phenomena.
 In addition there is a degree of electrical damping applied to the stylus mechanical resonance due to the 10kohm load the cartridge sees instead of the 47kohm which is standard practice.
In order for this circuit to function properly the basic input impedance of the phono amplifier stage circuit before any resistive loading network is added must be intrinsically above 100k ohms. Some Bipolar and most FET and Tube based phono-stages have this inherent impedance characteristic.
Cables connecting the TT to the phono-stage should be fairly low in capacitance per foot. If the capacitance does not get excessive the simple 6db/oct. filter pole could be above 100kHz. This would make the cartridge's high frequency extension primarily dependent upon the cartridge's mechanical characteristics.
Scotty
Edited for clarity

[dlaloum]

There was also a lengthy thread on VE, started by Luckydog, that looked at using a secondary cartridge body (not matched to the primary) but to use the circuit to cancel/neutralise capacitance - The focus was on all those people out there with phono stages having 220pf or more onboard and cartridges (like AT) that work best at 150pf or less...

The end result was very good and the associated cartridge performance model was then developed to become what I now use as my means of modelling cartridges...

Somewhere in my kit bag I still have my prototype capacifier using a p-mount as a means to plug in the desired "inductor" for the circuit...

The other interesting phono stage / loading setup for cartridges that differs substantially in behaviour is the Transconductance amp.... There was I believe another thread on VE on that topic, which I never joined.... at some point I need to get myself up to speed on transconductance amps - there are quite a few people out there that swear by them for phono stages, including some very well regarded engineers...

bye for now

David

[neobop]

Fascinating stuff.  For those who don't really understand basic loading considerations, read the TNT article.  Separate sets of graphs for varying capacitance and resistance show exactly how this works on a relatively high inductance cart with limited high frequency response.  With a low inductance cart with extended response and is overly bright, decreasing the resistive load value is the "normal" approach, like loading a 440 at 32K, as some do.

Scotty,
Most of us are not EE's or circuit designers (myself included), so some electrical theory explanation if you don't mind.  Here's a link for a parallel inductance tutorial:
http://www.electronics-tutorials.ws/inductor/parallel-inductors.html

Your drawing looks like the dummy cart is in parallel to the operational one, but from the above link:
"Inductors are said to be connected together in “Parallel” when both of their terminals are respectively connected to each terminal of the other inductor or inductors."

Then we have:
"One important point to remember about inductors in parallel circuits, the total inductance ( LT ) of any two or more inductors connected together in parallel will always be LESS than the value of the smallest inductance in the parallel chain."

So, we're not talking about mutually aiding or opposing inductors, but do the 5K resistors make a difference?   The inductors aren't connect together directly. Have you measured resultant inductance?  I'm wondering about final inductance value. 

More considerations later.
neo

[*Scotty*]

The cartridge exhibits a rising source impedance with frequency and the cartridge body exhibits a rising load impedance with frequency so that no change in voltage division occurs, therefore no high frequency rolloff occurs.
 Because the cartridge body to ground effectively compensates for the changing source impedance the high frequency rolloff and phase shift are mitigated.
 Of course these questions would be academic if you couldn't hear the difference, but you can. It sounds a whole lot better with the circuit than without it.
The 5k resistors together load the cartridge with about 10 to 11k which includes the cartridges DCR component. This is a compromise which preserves the signal to noise ratio and allows for an adequate volume level from most phono-stages.
 You have  virtually eliminated the contribution that the phono cartridge's inductance can make to the sound you hear from the cartridge. Electrically speaking you are good to about 100kHz. With the circuit in place, the cartridges mechanical limitations determine the upper limit of your high frequency response.
The circuit has a 6dB insertion loss because you have halved the amount of useable output voltage from the cartridge.
Scotty

[dlaloum]

Scotty, how closely does the cancelling cartridge need to be to the main cartridge?

How critical are other parameters, for example:

Could a Shure 1000e body (530mH) be used as the cancelling cartridge for a V15III (also 530mH), DC resistance for both is very close (1350 vs 1380ohm) but the V15III has a laminated core ... so it will behave differently at higher signal levels, due to eddy currents and any other magnetic effects that are reduced by the laminations...

Also, how does this differ (pragmatically in actual useage) from dropping the capacitive load to a very low level (eg: 60pf)?

Dropping the C, although awkward in many setups will also push the resonance/roll off above the audio range along with its phase effects... and if (big if!) the phono stage has little or no C loading then this can be a VERY easy option... as long as you use ultra short low C cables to connect to the TT.

The trouble I see with these alternatives, is that you are turning your MM/MI high inductance cartridge into something that behaves like an MC (low inductance) cartridge.

Along the way you sacrifice the tuning ability that a high inductance setup provides....

And although I really focused on these approaches a few years back - ultimately I think that these approaches don't tackle the phase issues properly - you need to look at the total phase shift of the system, that means both electrical and mechanical combined.... and it is possible that the mechanical and electrical will partially (or with design/tuning completely?) offset each other....

just thinking out load as usual....

bye for now

David

[*Scotty*]

David, if you have the basically the same inductive values it would certainly be worth trying out. In the case of AT and Grado cartridges, identical cartridge bodies are used in many instances in a given line and the unit to unit variation is very small.  I neglected to mention that the circuit, by minimizing the impact of the inductance also minimizes the impact of the magnetic non-linearities in the cartridges generating system.
 The improvement in the cartridge's overall linearity due to the circuit might obviate the need for tunability. The cartridge's inductance unless nulled out will cause smearing of the highs as well as the upper midrange. I don't see this problem going away even if the magnitude of the parallel capacitance is reduced to an impractically low value.
 There are many sonic rewards to implementing the circuit and the loss of tunablity might be a small price pay for the benefits that are to be had.
Scotty

[neobop]

Thanks Scotty,
It takes a little time to try to figure this out. 
Losing half your output voltage shouldn't be much of a problem, especially for high inductance carts, but knowing preamp impedance w/o load resistors, might. 

It seems that designers use LCR to tune response and w/o inductance you could be left with mechanical response that is less than ideal.  Loading down is a cure-all for mechanical response nonlinearities?   
I wonder what results you would get with 2 - M97 carts.  Even add 1 - SAS stylus to the mix.


Well, on another topic the Signet MR-ML won't seat on the 440 body.  The plastic sides are too tall, like trying to use one of those P-mount (92E) with the big sides, on a 95 only not as extreme.



You can see how the front of the Signet is taller.  The stylus guard makes it look worse, but there's still at least a 1mm gap.  The top on the 440 extends to the front of the cart and even without that I'm not sure if it will fit without trimming.  I was hoping to avoid that, but the best laid plans.....
neo