Cartridge loading

"[Lowering the R_load]... of course will limit the ability of the cartridge to trace higher frequencies. "
I fail to see it either, unless the below reasoning is wrong. I think of a simple cart model as a damped harmonic oscillator, excited by an external force of freq. f (the movement of the diamond induced by tracing the groove with a tone of freq. f). Loading the coils creates a current flow in them, which results in a damping force, opposing the movement of the coils in the magnetic field of the pole pieces. This is just a well known electromagnetic breaking force, proportional to the velocity of the movement (in turn proportional to the frequency) and inversely proportional to the R. It is just plainly ~f/R, like any other linear damping force. It adds to the total damping force, acting on the cantilever (the rest comes e.g. for the mechanical damping in the suspension). Lowering the R, just lowers the output across the entire spectrum but the nature of the output (its functional dependence on f) does not change at all. No additional damping of higher frequencies beyond the normal behavior of a damped oscillator. Just the damping coefficient increases.

I’m much more intrigued by @intactaudio dave’s observations of lowering the IMD. Have you tried plotting the IMD vs. R dependence?
Cheers

I think you might be over-thinking this.

@atmasphere  I doubt it, I'm quite ok with physics and I apply it to the situation.

You got most of this right, right up until your conclusion. Think about a generator, one with no load and one with a load, which will be harder to turn? By your logic above (if I’m reading it right) somehow the loaded generator is easier to turn, which certainly isn’t going to happen.

Seems you did not understand what I wrote: lower R presents obviously more breaking force, opposing the stylus movement. This is the electromagnetic induction law in action: the current (flowing through R) creates the magnetic field that opposes the stylus movement. This force behaves like ~f/R.

I think where you’re getting into trouble is the idea that the output goes down with reduced R load, which it does. The problem is: a certain amount of energy is used to make the stylus move. Where does that energy go? It is of course applied to the input load of the preamp in the form of a voltage. Now if you decrease the voltage by reducing the R load value, where is that same energy going? The Law of energy conservation says it has to go somewhere! It does not just ’vanish’. It is dissipated in the load and also by the cartridge coils themselves, both in the form of heat.

Just to be precise, the energy is not presented in a form of a voltage because voltage alone cannot perform work. The energy is presented in a form of a heat, dissipated in the combined resistance of the circuit (R_load, the coil DCR, the cables etc), caused by the induced voltage applied to the resistance. This is ok. The question is so what? To speak of energy conservation, you have to look at all the forces acting on the stylus:
- the driving force, coming from the diamond tracking a rotating, modulated groove, say at freq. f; this force is the source of all the energy flows- the restoring force of the suspension- various damping forces, including the electromagnetic one ~f/R

In a simple case of a linear suspension, you can solve it (for the speed of the stylus as the signal is proportional to it) and you will see that the movement has two components:
1) the transient, exponentially decaying self oscillations of the stylus; the frequency is the usual cart-tonearm combo but decreased due to the electromagnetic damping; the decay time is inv. proportional to the damping so ~R in the EMF part
2) the steady state, the forced movement with the freq. f, dictated by the tracked groove;this is the signal we want; the amplitude of this movement will have an R dependence too

The only *qualitative* change in R can happen is in the 1st, spurious part. Lowering  the R changes the suspension character from underdamped to critically damped to overdamped. But this is not the signal we are trying to get! This is an artefact added to the real signal of freq. f.
Of course if the motor is so weak that the stylus tracing a HF track into low R will make it slow, we are in trouble but let's assume a healthy TT design.

But I think you will find if you do some measurements that the output does not go down as fast as it appears you are thinking. This is because the stock 47K load is easy to drive and the output of the cartridge will stay pretty constant until the load is decreased to some point below 10x the impedance of the cartridge; IOW probably less than 100 ohms.

I suspect you are trying to describe a behavior of the 1/R function. Yes, far from zero it flattens out so changes say 1k to 47k can be negligible, but the closer to zero the steeper the changes.

Summarizing, you seem to selectively use the bits of the whole picture of the stylus motion and draw conclusions from them, like the fact that lower R changes the compliance, but you completely neglect that there is a very strong driving force here coming from the rotating platter, and this is the force setting the stylus into the motion.

Since we have been in agreement all along on the first two bits, maybe its this last bit that is the stumbling block. I used to load MM cartridges to critical damping by simply ringing the cartridge/cable combination with a square wave and observing the resultant output and taming it with a loading resistor. MM cartridges have a lot more inductance so its easy for that inductance to ring. But attempts to do this with LOMC failed, simply because with any loading I could not detect anything other than a nice looking square output since the inductance is so low. So I am challenging the idea of critical damping of the mechanical aspect of the suspension, not because I don't think it can happen but more because I'd like to see the evidence.

I don't say that achieving critical damping is doable or even desirable from the sonic perspective.

I've also noticed that while I can cut a 35KHz groove on my Scully lathe, depending on loading you can't always play it back, depending also on the cartridge. 

I think here lies the answer - "depending on the cartridge". I can imagine that for some cartridges, or better yet, some cartridge/tonearm combinations, the extra damping from a low R, combined with  other factors may compromise the tracking. What I fail to see however is that this should be some universal law.

So now I am curious- at what frequencies did you make your measurements?

I did not do any measurements. I tune R_load by ear, usually preferring lower values, and have never experienced any HF mistracking.

Dear @atmasphere please treat all my writings as a friendly argument in a search for the truth, so there is no need to apologize :)

If you have to use loading to achieve proper sound, it is a flag that something could be amiss: Instability in the phono section, a mismatch between arm and cartridge, that sort of thing.

I'd respectfully disagree. I'd say much depends on how the cart is designed. Electromagnetic damping due to a low R is actually quite an attractive (at least on paper) way of damping. Look at it from this perspective: linear just by the physics, no deterriortion, easy to implement and easy to control *by the user* with a great accuracy by changing R. Mechanical damping on the other hand, could be more challenging to implement with non-linerities, aging effects, no user control etc.

But the inductance of a low output moving coil is so low that resistive damping has little effect, as in a nutshell the inductor does not ring at audio frequencies or anywhere near them

But what I meant is the mechanical damping of the stylus, not electrical damping of the LC tank.

But damping of the mechanism (cantilever and suspension) is a different matter, and its pretty safe to conclude that if it is damped, high frequencies will be attenuated.

Ok, so if by tracking problems at HF you mean attenuated HF then I agree. But this behavior applies to any damping, also purely mechanical, not only to electromagnetic. The simplest damping is a force proportional to velocity of the cantilever movement. For periodic movements this means proportional to the frequency, so the damping force is increasing with the frequency of the played groove. Electromagnetic damping is just an example of this type of force. And this is what anyone who has ever played with loading hears - lower R = less HF, more pronounced LF. The question is so what if at the end you can reach a natural balance? If you cannot, sth is wrong or substandard.

But I’m not sure how important this is. If the cartridge is properly set up in the arm and the arm is able to track the cartridge correctly, **and** if the phono preamp is unresponsive to RFI and is also inherently stable, then IME the stock 47K load has yielded the best results. I do think its an interesting topic though and think it bears more research.

Again, I think this the last sentence is more of a personal preference than any rule. I do not see anything bad in designing a cart in a such a way that it relies on a certain load resistance to help mechanically damping the cantilever.

@intactaudio As I said in proper design it is obviously possible to reach a neutral balance and this is what we all do adjusting the load. Is the article available online?
In general, there are dozens of situations where a wise pre-loading linearizes a mechanical system. Think e.g. of a motor, where a pre-load helps to smoothen out coging

A MC cartridge is inherently a current generator

Why? By the Faraday law of induction, changing in time magnetic flux generates voltage (EMF) across the coil not current. Current flows when you close the circuit with a load.

I don't agree. Coil moving in a magnetic field can generate only one thing - voltage. This is the lesson of papa Faraday :)

bydlo and larry, I wouldn't argue with what you both say, but I would point out that the capacity of a LOMC to make signal current into a very low resistance load (meaning a load that is equal to or much lower than its internal impedance) does not usually parallel its capacity to make voltage into a "high" resistance load (meaning any load that is about 10X the coil resistance). (I am not getting into the argument between Raul and Atma-sphere.) For example, my MC2000 is rated for 0.05mV at the standard stylus velocity. But its internal resistance is only 2 ohms. Thus it can generate 25uA of current into no load or probably anything much lower than 2 ohms. For comparison, my Audio Technica ART7 has twice the voltage output of the MC2000 (0.12mV) but also has an internal impedance of 12 ohms at 1kHz. Thus the ART7 is less efficient at generating current (10uA), when forced to do so, than the MC2000. Viewed this way, the MC2000 more than holds its own for current output, among very LOMC cartridges.

@lewm What you are describing is the fact that the lower output resistance of a voltage generator, the more current efficient it is. This does not change the fact that by the laws of induction MC and MM carts are both voltage generators.

A secondary issue is whether or not the cartridge was built with some form of electrical damping in mind with the intent of thus creating mechanical damping. This seems problematic as the compliance of the cartridge is directly affected- that being the case a range of loading values would be specified for a given effective mass and we certainly are not seeing that in the spec sheets!

@atmasphere  I'd imagine the compliance change may not be big in the manufacturer specified load range so no one cares. Plus  load optimization is usually done by ear anyway.

@intactaudio

Now, as the force against the groove wall lessens
momentarily,...

Very much agree with that. EMF damping I’ve been talking about. Thanks for sharing Dave.
@atmasphere Don’t really buy the RFI argumentation, sorry. Putting aside simple and super cheap solutions like grid stoppers existing for decades, I cannot see how the RFI conspires to always give the effects exactly mimicking underloaded (high R) cartridge.

@antinn Fantastic source, thanks Antinn for sharing! But the cart model of Shure brothers analyzed there is *unloaded* (plus the current source is strangely drawn with series (??) instead of parallel source impedance). The R on the schematics is the mechanical damping of the suspension, not the loading R.

@antinn

That article was modeling the mechanical resonances as an equivalent electrical circuit. I was trying to show that the mechanical resonances were more complex than I originally thought, and if the cartridge goes into physical resonance (which is more likely at high frequency), then there will be a high harmonic distortion (EMI/RFI) produced.

Indeed it's more complicated than just a mass on a spring as there comes into play the vinyl resonance (I admit I never thought about that).
But getting a mechanical resonance radiate into RF range I think is not easy. Ultrasonic yes, but RF we are talking about some 40-50kHz at least.

Its my understanding that LOMC, are poor voltage generators, but good current generators which is why the new generation of LOMC preamps are current sensing.

Again: MC and MM cartridges work by the electromagnetic induction. What they generate is voltage. What I mean is if you play a cartridge unconnected what you will measure on its pins is a voltage but no current flow as the circuit is open. So cart is a voltage generator. Now, you in the electric circuit theory you can exchange current and voltage sources on the paper. You can redraw any circuit using voltage sources to current sources and vice versa. BTW the current source at teh Shures schemtics is wrongly drawn. Series resistance makes no sense with current sources. What I suspect the "current generator" thing mean is that LOMC's have a low internal resistance so for a given output voltage they will push more current. And this is of course a fact.

If a capacitor saturates because of the high harmonic current, and there is no dampening resistor, the only thing that is left is the capacitor equivalent series resistance (ESR). If the circuit current suddenly increases, this will by the back emf, dampen the motion of the cartridge armature (stylus-cantilever-suspension), and may depending on where on the record this is occuring (i.e. radial velocity), cause miss tracking. So long as the current was not outrageous and the capacitor is not damaged, and most will self-heal, once the current drops, the capacitor returns to function. So the event can be very transie

Sorry, cannot understand what you mean here. Which capacitor? Can you point me to the relevant passage from your links?