Just to illustrate, here's an estimate of likely flux density changes arising in MM/MI and MC cartridge cores which shows it in both cases to be very small in the context of being unaffected by the large scale non-linearities arising in B-H curves of highly permeable alloys.
Apols it's a bit mathy, but hopefully if the terms used in the JC extracts make sense to you, so should this !The aim is to estimate the size of flux changes in typical MM/MI and MC cartridges, and relate this to published B-H curves for typical core materials and geometries:Definitions :
Coil self inductance L [H]
Number of turns in coil N 
Current in coil I [A]
Induced voltage across coil V [V]
Induced voltage per turn in coil Vt [V] = V/N
Total flux linking coil Ø [Wb]
Flux Density in coil core B [Wb/m²]
Time t [s]
Angular frequency w [rads/s] = 2*pi*frequency
Coil core area A [m²] Relations :dØ/dt = dI/dt * (L/N)
as an inductor, rate of change of flux is proportional to rate of change of current, the constant is inductance per turn.V = L * dI/dt Vt = (L/N) * dI/dt
voltage per turn is proportional to rate of change of current, the constant is inductance per turn
Then Vt = dØ/dt
; ie coil volts per turn is exactly rate of change of core flux.Calculations :
For an MM/MI N = 1000 , number of turns in the coil is about 1000 ref this thread : viewtopic.php?p=273610
For an MC, N = 50, number of turns in MC coil is about 50, same reference above.
Assume an output of 5mV (MM), or 0.5mV (MC), which is near full scale output ref 0dB@1kHz sine wave. We wish to find Ø, the peak amplitude of total flux in the core, and H the peak flux density in the core.
dØ/dt = Vt = 5mV/1000 = 5E-6 V
Because the stimulus is a sinewave, the peak value of Ø = Vt / w Then for MM peak Ø = c 8E-10 [Wb]
and for MC peak Ø = c 1.6E-9 [Wb]
For MM core section area A might be 5mm², for MC A might be 3mm²
Then peak core flux density B for MM/MI is estimated to be Ø/A = 0.00016 Wb/m²
And peak core flux density B for MC is estimated to be 0.00053 Wb/m²And the amplitude of flux density changes in coil cores associated with full scale output of the cartridge at 5cm/s@1kHz is estimated to be circa
0.00016Wb/m² for MM, and 0.00053 Wb/m²Discussion
A datasheet for what seems a fairly ubiquitous permalloy core material is listed here : http://cartech.ides.com/datasheet.aspx? ... &c=TechArt
B-H curves are listed, which show the onset of saturation being at flux densities c 1 Wb/m².
Then the working range of flux density changes within coil cores for MM and MC cartridges is estimated to be 0.016% and 0.053% of the working range of that permalloy material. Associated with cartridge outputs representative of 5cm/s@1kHz, 5mV and 0.5mV outputs respectively.
Therefore, it appears that flux density changes represent a very small part of the working range of the core material, and therefore approximate a linear B-H transfer function. Even allowing for normal peak programme level overloads, there appears to be several orders of magnitude margin. Sufficient to support the notion that non-linearities in the B-H transfer curve aren't significant at any level or frequency.
Interesting ? I'd welcome this being checked over, but i think it's right. Comments and questions welcome, as always.