loading, back emf and damping

[Ldg]

.....Anyway, it makes the effects of loading mc cartridges even more mysterious. From an electrical point of view the difference between a 100 ohm load and a 500 ohm load to a 10 ohm cartridge must be miniscule, yet there are people who swear that they can hear a difference. I wonder how much of it is imagined.

If any audible effect of varying loading R is real, it doesn't have a mechanical cause. More likely a magnetic low power non-ideal, IMO.

For completeness and interest, the only MM for which I have published mechanical impedance data is Shure V15 II, 0.048Ns/m@1kHz. Mechanical power 5cm/s@1kHz is c 115uW (versus 375uW for DL103, V15II is lower mech impedance, of course). For MMs, electrical load power is even smaller, c 0.3nW into 47K. So MMs have even less electrical load variation effect on mechanics than MCs. But both MM and MC have effectively zero mechanical influence through electrical load variation.

I don't recall posting these results before, but it is possible I did.

I think any real audible effect in MCs of loading variation is most likely to have an origin in magnetic non-ideals, and their variation with level.

Lastly, I've seen that Shure peak velocity data before, and recall that it doesn't stand scrutiny. IIRC, it was mooted it may include noise pops. I suppose one would argue that is still relevent to headroom. But it doesn't change anything here. And usual caveats about Shure marketing material from that era must apply !

[Larry I]

The electrical circuit of the cartridge can be characterized by inductance, capacitance and resistance. By changing the resistance (loading change), you are changing the filter characteristic of the circuit.

[Rothwellaudio]

The electrical circuit of the cartridge can be characterized by inductance, capacitance and resistance. By changing the resistance (loading change), you are changing the filter characteristic of the circuit.

Yes, I am aware of that, but given the source impedances and inductances involved (12 ohms and 25uH for the Audio Technica ATOC9, for exmample) and about 100pF of cable capacitance, there is no significant resonance in the audio band or above it. Just plug in the values and it's easy to see that adjusting the load impedance has no useful filter modifying behaviour (unlike with MMs).

Non-linearities in the magnetic circuit sounds a more plausible explanation.

[Ldg]

Non-linearities in the magnetic circuit sounds a more plausible explanation.

Yes, if/when any audible effect is real, there's a few pointers to that being a mechanism. You're probably in a good position to comment on the characteristics of such non-ideals, Rothwellaudio. It's a topic I'm very interested in, especially characterising phono cartridge generator performance versus level, and versus output load.

I think one way to look at phono carts is as a transformer with a 'mechanical' primary, rather than an 'electrical' wound primary. Then looking at non-ideals which affect low power audio transformers, especially variation with secondary load, might be revealing. Would be interesting to read comments and pick knowledge on that ?

[Rothwellaudio]

Yes, there will be similaries with a transformer, but I'm not sure how far you can go with the analogy because there are fundamental differences too. In a transformer the current in the primary produces a magnetic field which is concentrated and focussed (kind of) by the core, and the magnetic field induces current in the secondary. More accurately, it's the changes in the magnetic field as the signal fluctuates that cause the lines of flux to expand and collapse and cut through the secondary coil that induce the current in the secondary.
In a cartridge the magnetic field is produced by a permanent magnet and the current is induced as the coils move side to side through the magnetic field. In that respect there's a fundamental difference between a cartridge and a transformer - there's no core becoming magnetized and de-magnetised, which is where most of the distortion in a transformer arises.
However, the current in the coils produces its own magnetic field which must interact with the field from the permanent magnet. But that just brings us back to the issue of back EMF and damping. Or does it? Maybe the field produced by the coil distorts the field produced by the permanent magnet enough to have an audible effect.
You've probably considered all this already, but I'm just thinking out loud at the moment.
Actually, the permanent magnet and its pole pieces would be acting in a similar way to a transformer core and would perhaps have the same type of non-linearities. I'll have to give it some more thought.

Of course, the reason for me asking these questions is to try to get to the root of the cause of any audible differences produced by loading. We make transformers and we also make MC phonostages, and I'd like to make them the best they can possibly be. At present both the transformers and the active MC amps present a 100 ohm load to the cartridge. They could both be a lot lower. Alternatively, the actives stages could be a lot higher, and the transformers a bit higher, but my experiments so far show no advantage in going either way.

[Rothwellaudio]

I found my notes for a worked example, for which there is published data : DL103, 5cm/s@1kHz rms, 0.3mV rms, 40R coil, 0.15 Ns/m @1kHz mechanical impedance.

Mechanical power : 375uW rms (0.15 Ns/m)
Electrical power : 2nW rms (40R)

You can see there are orders of magnitude difference, and mechanical power involved thoroughly swamps electrical power. Electrical loading therefore has no significant influence on mechanical loading.

I've just been looking through some of the figures in this site's database and came across this with regard to Audio Technica cartridges: "AT publishes their dynamic compliance specifications relative to 100Hz . The actual compliance at 10Hz will be higher." This implies that the compliance at 10kHz might significantly lower than the quoted figure, so maybe electrical damping shouldn't be dismissed after all.

[Ldg]

Here's the charts for published mechanical impedance versus frequency for the DL103, and V15II, from Yosh's excellent site :




Compliance at 100Hz is derived from measured mechanical impedance at 100Hz. It is really a measure of mechanical impedance at 100Hz, determined from VTF required to track 50um lateral amplitude at 100Hz.

In any event, you can see mechanical impedance varies significantly with frequency. BUT there are so many orders of magnitude difference between electrical power and mechanical power in play, that it still makes no difference. I chose 1kHz for power calculations as it is sort of median, but actually at many frequencies mechanical impedance is a lot higher (which means more mechanical power in play at a given velocity, of course).

So I really don't see electrical load influencing mechanics, even taking this into account.

[Ldg]

.......Actually, the permanent magnet and its pole pieces would be acting in a similar way to a transformer core and would perhaps have the same type of non-linearities. I'll have to give it some more thought.

Don't many MCs have a permeable pole piece or core over which the coils ride, or more to the point few are truly air cored ? And besides, a complete magnetic circuit is always in play, comprising highly permeable and non-ideal components, i think you are right Rothwellaudio ?

... We make transformers and we also make MC phonostages, and I'd like to make them the best they can possibly be. At present both the transformers and the active MC amps present a 100 ohm load to the cartridge. They could both be a lot lower. Alternatively, the actives stages could be a lot higher, and the transformers a bit higher, but my experiments so far show no advantage in going either way.

Well, at least you're up front about it. Personally, I'm generally not inclined to contribute gratis toward commercial ventures. I think you should contribute some quid pro quo content of some substance about the nature of small signal non ideals in transformers/magnetic cores, Rothwellaudio. Such is probably at the heart of both cartridge and transformer performance. I reckon it's your shout to put up some tangible content on the topic........!

[dlaloum]

Jonathan Carr has provided some input in the past on small signal non-linearities related to permeable core magnetic systems.

He also pointed at the MC designs that have non-permeable cores, his own Lyra designs obviously, but also Denon DL304, DLS1, JVC L1000, Benz Ruby...

And as an aside he mentioned that the same issues (obviously) exist with permeable core transformers.

Presumably it would be possible to create air-core audiophile transformers? (too inefficient?)

Yep your input here would be very much appreciated!!!

[Ldg]

Hmmm. If one does the sums on what inductance a typical air cored MC coil should have, given turns/geometry, it comes out about x10 smaller than typical quoted MC cartridge inductance figures, by my reckoning. To me, this suggests that Rothwellaudio may well be right, and the permanent magnetic path might play a significant role in the flux circuit of even an air-cored coil. Presumably because of proximity. Permanent magnetic circuit is made from permeable materials, obviously. It may not necessarily be devised for low hf loss though, essentially as an inductance core. And it may well have small signal non-ideals.

Perhaps one way of modelling non-ideal output impedance of the cartridge is as a frequency and level dependant series element. And it would be this 'loss' impedance that might give rise to any real audible effect of loading variation in MCs ? In which case, you'd see output impedance vary with frequency, and with level. To a far greater extent than predicted from merely looking at inductance spec and coil resistance.

Reciprically, externally measured impedance at the cartridge pins should vary with frequency, and signal level. To an extent greater than described by simply coil inductance and series resistance. Some tests I made on MMs at very small signals seem to tentatively support this idea.

Just thoughts.

[Rothwellaudio]

Jonathan Carr has provided some input in the past on small signal non-linearities related to permeable core magnetic systems.

He also pointed at the MC designs that have non-permeable cores, his own Lyra designs obviously, but also Denon DL304, DLS1, JVC L1000, Benz Ruby...

And as an aside he mentioned that the same issues (obviously) exist with permeable core transformers.

Presumably it would be possible to create air-core audiophile transformers? (too inefficient?)

Yep your input here would be very much appreciated!!!

All magnetic cores have to be permeable, or else there would be no point in having the core. The distortion arises from core saturation and from hysteresis. Hysteresis is the resisdual magnetism left behind when the magnetising force is removed.
Yes, it would be theoretically possible to make transformers with an air core, and it is done for radio frequency transformers. However, to work at low audio frequencies the number of turns would have to huge and the associated resistance and capacitance of the windings (and leakage inductance) would render it impossible to get anything like a reasonable high frequency response. So in practise it is impossible.

[Rothwellaudio]

.......Actually, the permanent magnet and its pole pieces would be acting in a similar way to a transformer core and would perhaps have the same type of non-linearities. I'll have to give it some more thought.

Don't many MCs have a permeable pole piece or core over which the coils ride, or more to the point few are truly air cored ? And besides, a complete magnetic circuit is always in play, comprising highly permeable and non-ideal components, i think you are right Rothwellaudio ?

... We make transformers and we also make MC phonostages, and I'd like to make them the best they can possibly be. At present both the transformers and the active MC amps present a 100 ohm load to the cartridge. They could both be a lot lower. Alternatively, the actives stages could be a lot higher, and the transformers a bit higher, but my experiments so far show no advantage in going either way.

Well, at least you're up front about it. Personally, I'm generally not inclined to contribute gratis toward commercial ventures. I think you should contribute some quid pro quo content of some substance about the nature of small signal non ideals in transformers/magnetic cores, Rothwellaudio. Such is probably at the heart of both cartridge and transformer performance. I reckon it's your shout to put up some tangible content on the topic........!

Thanks for stimulating and contributing to the debate ld, but I'm not out to steal ideas and make a fortune from them, I'm just curious and always trying to gain a deeper understanding of what I'm doing so I can do it better.
Small signal distortion in transformers is caused primarily by hysteresis, though I don't claim to have any greater knowledge than is already in the public domain. Unfortunately, Rothwell Audio Products is a tiny company without the resources to cast/roll/anneal various iron alloys, so we're at the mercey of what core materials are available commercially. Having said that, the core material is only half the battle - the other half is all about the windings. The windings are where there is scope for experimentation.

[Rothwellaudio]

Hmmm. If one does the sums on what inductance a typical air cored MC coil should have, given turns/geometry, it comes out about x10 smaller than typical quoted MC cartridge inductance figures, by my reckoning. To me, this suggests that Rothwellaudio may well be right, and the permanent magnetic path might play a significant role in the flux circuit of even an air-cored coil. Presumably because of proximity. Permanent magnetic circuit is made from permeable materials, obviously. It may not necessarily be devised for low hf loss though, essentially as an inductance core. And it may well have small signal non-ideals.

Perhaps one way of modelling non-ideal output impedance of the cartridge is as a frequency and level dependant series element. And it would be this 'loss' impedance that might give rise to any real audible effect of loading variation in MCs ? In which case, you'd see output impedance vary with frequency, and with level. To a far greater extent than predicted from merely looking at inductance spec and coil resistance.

Reciprically, externally measured impedance at the cartridge pins should vary with frequency, and signal level. To an extent greater than described by simply coil inductance and series resistance. Some tests I made on MMs at very small signals seem to tentatively support this idea.

Just thoughts.

Permanent magnets are made from so-called "hard" materials whereas transformer cores are made from "soft" materials. That's what makes permanent magnets permanent - it's hysteresis taken to its extreme. Whether or not the current flowing in the coils of an mc cartridge is enough to have any impact whatsoever on a permanent magnet, I don't know, but I doubt it. But if it did, there would be significant hysteresis involved, presunably.

Also just thoughts.

[Ldg]

I'm just curious and always trying to gain a deeper understanding of what I'm doing so I can do it better.

Thanks, Rothwellaudio. I empathisize because that's my philosophy too. I always like to run out of questions before running out of answers.

I think the permanent magnetic circuit may be in play, even for an air cored MC.

As previously posted, one way of looking at is as though the source of flux change is a perfect primary of a transformer. And the MC coil is a real secondary winding, which sits in an (air) gap. Here's a sketch which shows the permanent magnet loop (yellow), and the MC coil. Plus a source of flux changes.



One can see that, proximity and geometry permitting, changes in MC coil flux [inc back emf/loading effects] are mostly conducted by the permanent magnetic circuit. For example, self inductance of the MC coil is determined (potentially significantly) by the permanent magnet circuit. And, being highly permeable, influenced by the non ideals therein.

From a back emf and loading point of view, I think the permanent magnetic circuit is effectively a core, even if the coil is in an air gap.

In any event, using this sketch model, one can evaluate an MC cartridge as a transformer. Albeit a strange one, with high leakage inductance, and potentially significant non-ideals.