"Tracking" speed of pickups?

[Mike Sulzer]

That's mixing time with frequency.

But they are mixed; time and frequency are two different domains for describing the same thing. A slow speaker has a limited frequency response.

[Mike Sulzer]

Isn't that backwards? A high-Q resonator takes time to build up, and a long time to ring back down. Q is basically the ratio of circulating energy in the resonator to incoming/outgoing energy.

Yes, I agree with what you are saying. And I am saying the same thing: when you introduce eddy currents, the resistance across the resonant circuit is lowered. That lowers the Q. The circuit takes less time to to build up and decay.

How do we know that it's a small dip? Given the sqrt[f] dependence of eddy currents, one would expect increasing effect as the frequency increases. Something else must also be happening.

Because measurements show it to be a small dip, and I am not the only one to have done them.

For years you have been saying that the effect of eddy currents increases with sqrt(f). This is certainly true under some circumstances, but it is hardly the whole story. In analyzing a pickup, we make an equivalent circuit. The useful simplest model of the effect of eddy currents in a pickup is a resistor in series with an inductor. The resistance increases with sqrt(f). At very low frequencies, the resistor is large compared to the inductive reactance of the pickup coil; eddy currents do not matter. This is why 120 Hz is good for measuring the coil inductance. The inductive reactance increases faster than the eddy current resistance, which becomes significant, and it tends to load down the circuit. As the frequency increases still more, the resonance takes over and the output rises. At very high frequencies, the inductive reactance of the eddy current effect (the leakage inductance) becomes larger than the resistance and the effect of eddy currents is more limited. (This is the same effect that makes it necessary to have very low leakage inductance in a transformer if it is to have a wide bandwidth.)

[Joe Gwinn]

Yes, I agree with what you are saying. And I am saying the same thing: when you introduce eddy currents, the resistance across the resonant circuit is lowered. That lowers the Q. The circuit takes less time to to build up and decay.

It's a bit more complicated than that. The classic example of the time-delay relay is instructive. When the coil is energized, a large current is induced in the copper sleeve. By Lentz's Law, the sense of the induced current is such as to oppose the magnetic field from the relay coil, so at least initially there is little magnetic field and the relay remains open. However, copper is not a superconductor, so the circulating current soon dissipates, and the relay closes. When the coil is de-energized, a large current is induced as before, holding the relay closed for a while.

The parallel for pickups is if the music field has to get through a brass cover. Sudden changes in field induce eddy currents that oppose the change, slowing things down. This will mainly affect attack transients.

Because measurements show it to be a small dip, and I am not the only one to have done them.

While no doubt true, this is not an answer to the question asked. How did they go about proving that the dip was due to eddy currents, versus a host of other effects? There is a lot going on at once. While there may be models, these must be validated by measurements.

For years you have been saying that the effect of eddy currents increases with sqrt(f). This is certainly true under some circumstances, but it is hardly the whole story. In analyzing a pickup, we make an equivalent circuit. The useful simplest model of the effect of eddy currents in a pickup is a resistor in series with an inductor. The resistance increases with sqrt(f). At very low frequencies, the resistor is large compared to the inductive reactance of the pickup coil; eddy currents do not matter. This is why 120 Hz is good for measuring the coil inductance. The inductive reactance increases faster than the eddy current resistance, which becomes significant, and it tends to load down the circuit. As the frequency increases still more, the resonance takes over and the output rises. At very high frequencies, the inductive reactance of the eddy current effect (the leakage inductance) becomes larger than the resistance and the effect of eddy currents is more limited. (This is the same effect that makes it necessary to have very low leakage inductance in a transformer if it is to have a wide bandwidth.)

Well, I don't have the energy to refight this, but modeling eddy currents requires that the lumped-circuit parameters be a function of frequency.

[Mike Sulzer]

It's a bit more complicated than that. The classic example of the time-delay relay is instructive. When the coil is energized, a large current is induced in the copper sleeve. By Lentz's Law, the sense of the induced current is such as to oppose the magnetic field from the relay coil, so at least initially there is little magnetic field and the relay remains open. However, copper is not a superconductor, so the circulating current soon dissipates, and the relay closes. When the coil is de-energized, a large current is induced as before, holding the relay closed for a while.

The circuit model is derived by considering the changing flux through the loop, or copper sleeve or whatever. The effects you described are in the model; it is not necessary to consider the induced currents separately.

The parallel for pickups is if the music field has to get through a brass cover. Sudden changes in field induce eddy currents that oppose the change, slowing things down. This will mainly affect attack transients.

Again, the circuit model will correctly predict the frequency response. There is no additional time delay for the the signal to get through the brass.

While no doubt true, this is not an answer to the question asked. How did they go about proving that the dip was due to eddy currents, versus a host of other effects? There is a lot going on at once. While there may be models, these must be validated by measurements.

The question asked, the only one I saw, and the one I was answering, was "How do we know that it's a small dip?" I repeat: we know it is a small dip because that is what the measurements show. I am intrigued by this host of other possible effects that could cause a dip in the mid range. Can you list some?

Well, I don't have the energy to refight this, but modeling eddy currents requires that the lumped-circuit parameters be a function of frequency.

Well, the parameters could be a function of frequency, or not. The resistor in the pickup case is because of skin depth effects. The inductor, on the other hand, is just an inductor.

[Joe Gwinn]

The circuit model is derived by considering the changing flux through the loop, or copper sleeve or whatever. The effects you described are in the model; it is not necessary to consider the induced currents separately.

Again, the circuit model will correctly predict the frequency response. There is no additional time delay for the the signal to get through the brass.

The question asked, the only one I saw, and the one I was answering, was "How do we know that it's a small dip?" I repeat: we know it is a small dip because that is what the measurements show. I am intrigued by this host of other possible effects that could cause a dip in the mid range. Can you list some?

Well, the parameters could be a function of frequency, or not. The resistor in the pickup case is because of skin depth effects. The inductor, on the other hand, is just an inductor.

Well, we are back to models of questioned fidelity. And inductance does vary as a consequence of eddy currents.

Where is the experimental verification of this model documented?

[Mike Sulzer]

Well, we are back to models of questioned fidelity. And inductance does vary as a consequence of eddy currents.

Where is the experimental verification of this model documented?

We are back to you ignoring reasonable questions. End of conversation.

[Chuck H]

I think what we've learned is that "tracking speed" has to relate to a pickups reaction to frequencies in peaks and transients. NOT actual signal speed. That is... No pickup can cause a delay in the audible attack and decay. At worst you get a colored attack and decay. Not being a pickup guy I can only assume this can be part of the point just as it is in amplifier design. The intention is to color the sound in a useful and musical way. That in mind I think the goal of understanding this phenomenon has merrit. Now if we could avoid arguing about it...

[Steve Conner]

I like Funky's point about string vibration vectors and so on.

However, I think a Strat sounds "faster" mainly because the single coil pickups have a higher resonant frequency than humbuckers. They let more high frequencies through, and the attack portion of the note is made of high frequencies, so it is emphasised.

There are also cancellation effects between the two coils of a bucker that Mike has written about in great detail. These might well kill even more high end.

[Joe Gwinn]

We are back to you ignoring reasonable questions. End of conversation.

Models can do just about anything, physical or non-physical, and so can be very misleading. That's why one must validate one's models, and why one must be very cautious with unverified models.

That's also why one verifies first, then debates.

[Plectrum]

Great stuff guys.
So, I would say that these words associated with speed are becoming used to describe how "hard" or how much upper-mid and treble content there is when a certain design of pickup is used. Maybe the dynamic element is part of this too, a stock-wound single coil strat unit produces a proportionally large attack-phase compared to many 'buckers so maybe this would be perceived as a"speed" thing too?

It sort-of makes me slightly uneasy, probably for no important reason, but because I can only think people will start to hear actual delays (as we are prone to perceiving the things we expect to perceive, rather than what we actually experience) where there are none - and argue the point - Doom. LOL.

As Steve Conner states: "The human brain is such a fickle thing, though." Truer than most of us realise.
One only has to look at the "Audiophile", HiFi "Aficionado's" various forums and online stores to see how butchered their understanding of what they experience has become.

[David Schwab]

Models can do just about anything, physical or non-physical, and so can be very misleading. That's why one must validate one's models, and why one must be very cautious with unverified models.

That's also why one verifies first, then debates.

Right. This is why I said in the thread about multi coil pickups that you have to listen to them. Then figure out what's gong on. Models and just taking inductance readings is a bit simplistic and doesn't really tell you what the pickup will sound like. If they did, it would be easy to get what ever tone you wanted from a pickup by just matching inductance readings, etc.

[FunkyKikuchiyo]

Great stuff guys.
So, I would say that these words associated with speed are becoming used to describe how "hard" or how much upper-mid and treble content there is when a certain design of pickup is used. Maybe the dynamic element is part of this too, a stock-wound single coil strat unit produces a proportionally large attack-phase compared to many 'buckers so maybe this would be perceived as a"speed" thing too?

It sort-of makes me slightly uneasy, probably for no important reason, but because I can only think people will start to hear actual delays (as we are prone to perceiving the things we expect to perceive, rather than what we actually experience) where there are none - and argue the point - Doom. LOL.

As Steve Conner states: "The human brain is such a fickle thing, though." Truer than most of us realise.
One only has to look at the "Audiophile", HiFi "Aficionado's" various forums and online stores to see how butchered their understanding of what they experience has become.

All great points! What we hear is incredibly important and is (or at least should be) our primary source of data, but sometimes what we think we hear isn't always what we're really hearing. Also, people use some pretty strange words to describe what they hear sometimes, too! I've had to do some pretty exotic translations. It makes for a great exercise in listening to try to describe these things, though.

"Hardness" is another funny thing to hear. Some notes will have a hard sound while others sound kinda like they "bloom" - a very good strat tends to "bloom" a lot. As far as I can tell, this is primarily an acoustic phenomenon and the pickups just need to relay it and not hide it, but it isn't something I've given a lot of consideration to, unfortunately.