The true midrange speaker driver

Is there any way of standardizing what makes a midrange driver really great? Do we even know what it is?

I have made my 10 commandments of a true midrange!

What matters and why do certain seemingly unimportant factors play an important role?

Read more about it below!

The midrange driver in a loudspeaker carries most of the information from the human voice and for almost all tonal instruments. It has to deal with both tones and transients at a huge range of levels. Its role is so important that even playing without a tweeter and a woofer, a great midrange would still sound good. Therefore we should really care about its qualities. This is the driver that can really make or break a loudspeaker.

There are quite a few loudspeaker driver manufacturers out there that offer something labeled as a midrange driver. Many of these use technologies that are obviously there to improve the drivers ability to perform midrange as apposed to a woofer. There are also lots of full range drivers out there, obviously optimized for full frequency range performance. Many argue that those are often better midrange drivers than the typical midrange driver. So why don’t the midrange drivers borrow technology from full range drivers? And what would the disadvantages of this be?

Historically there are a few quite remarkable midrange drivers. I would like to use the famous Focal 7k2 as a benchmark for this article to make things a bit more hands on and easy to understand. This is a driver that went out of production many years ago, and a predecessor really does not exist. Just to put things into perspective, please note that this document has no reference to a web page or an e-mail address. It is a really old driver.

Anyway, this is, as you can see, a kevlar driver. That is not a hugely popular material today. It was much more popular in the 90’s. And this illustrates a bit of where I am going with this, because frequency response and asthetics tends to be everything today, and if it is also underhang, and has a cone material with the right marketing word in its name, then you are all good to go. So what else is there to notice about this driver?

The motor
A surprisingly often overlooked part of the driver is its motor. I remember people were bragging about large magnets as a measure of quality of a driver. They might have been closer to the truth than they knew. The Focal has a 134 mm magnet, a pretty huge magnet for a 7 inch driver, especially for a midrange. As you can also see, it has a sensitivity of 95,2dB which is outstanding for an 8 ohm driver of this size. So obviously, high sensitivity has been a target for the team who made this. The inductance is fairly low for a driver with a 40mm coil without shorting rings. It is safe to assume that this is not very far from well saturated, and that really makes a difference. I do not know if this was even a goal, but still they seem to have achieved it. If they had added extensive copper rings as well, the motor performance could have been pretty outstanding by today’s standard.

The coil length is 13mm so it is fairly long for a midrange, but only a single layer, which makes it light. And this is really a key factor on a midrange driver, keeping the components attached to the cone as lights as humanly possible. This really rules out redesigning a woofer to make it into a midrange. If they had swapped the copper wire for aluminium, it would have been even better. They would have needed a marginally wider winding for the same force factor, but the reduced mass means they would not have needed the same force factor for an even higher efficiency. This driver also maybe has even more excursion capacity than you need from a midrange driver. It depends on where you want to cross it, but really optimizing for midrange often means ruling out crossing it over at 100Hz. And since its Fs is around 100Hz, the dominant oposing factor to force factor will be suspension and cabinet, rather than mass, for the lowest octave when crossed over that low.

Damping
One part of the driver that has received lots of attention the later years is the voice coil former. Everyone is moving to titanium, as it is fairly fast (in terms of speed of sound), it is robust, and it is a poor conductor. That means it has many of the same properties as Kapton or polyimide which is also used in the Focal driver. Non conductive bobbins means eddy currents are not generated in the bobbin as it moves in the magnetic field. This leads to higher Qms and lower Rms, also known as mechanical loss. Back in the days, woofers were often offered with aluminium voice coil former, and some were offered with a Kapton option. These versions did typically have higher Qms, lower Rms, slightly rougher frequency response, and still many people think they had far superior midrange performance. This is likely as the difference in frequency response between the aluminium and Kapton bobbins was not caused by removing resonances, just by masking them in the time domain by adding mechanical loss. So the smoother frequency response did not actually translate to smoother sound like one could be lead to believe. This could also be seen on the impedance curve, if it was honestly and precisely plotted.

Look carefully at the impedance curve
Looking at the Focal again, it is easy to see the peak of the fundamental resonance. If we look closely, we can also see a tiny raise around 1–1,5kHz. This peak corresponds to a rise in the frequency response, also very subtile, in the same range. It is likely that this is related to the suspension of the driver. On woofers this is typically far more prominent than we see here. Using a Kapton coil former, an copper voice coil and a rubber surround would probably not have made this possible. But the Focal has cloth surround. More specific, it is a short stroke narrow surround weighing almost nothing. We can see from both the smooth impedance and the high Qms that this is also a very low loss suspension.

In this same time period, some manufacturers chose to use high loss surround and aluminium voice coil former in combination with polypropylene cones. These drivers had extremely low Qms and they behaved almost like a mechanical resistor. In theory this looks like a great idea, but there are two big problems with this. One is that any problem will be so well hidden that it is hard to find, and therefore also just as hard to solve. The other problem is that mechanical resistors doe not return only heat, but translates a lot of energy back as non linear and non harmonic distortion. A neat example of this is the hysteresis loops of magnetic materials as described in Purifi Audio’s brilliant article here: https://purifi-audio.com/2020/04/28/this-thing-we-have-about-hysteresis-distortion/

Heat
We often think more about heat when dealing with woofers as they often seem to require more power to be driven to their extremes. However, a midrange is always used together with woofers, and has to follow them all the way. Many midranges are also crossed over below the baffel step, meaning they have to deliver 6dB more, or handle 4 times the power of a tweeter with the same efficiency for the same sound pressure. Typical energy distribution in musical content peaks around 100Hz, and can be as much as 10dB lower at 1kHz, meaning that a midrange that is crossed over low may have to handle more than 40 times the power of the tweeter. This could translate to some serious heat.

The Focal uses a 40mm coil, and its dimensions tells us a lot about how it might cope with power. We can compare it to the Morel TSCM634 wich is a driver about the same size, but with a huge 75mm voice coil. As we increase the size of a driver, the increased diameter translates to a squared increase in surface are. For the voice coil, however, this is a linear relationship, so an 80mm voice coil is only has twice the surface area of a 40mm voice coil. This means the Morel is not that much bigger. The Morel has a 5,5mm winding length, while the Focal has 13mm, so more than twice that of the Morel. This means that in total, they have about the same surface area to radiate heat.

But the sensitivity of the Focal is 6dB more than that of the Morel. And the Morel is a 4 ohm driver, meaning the Morel has to deal with 8 times as much power for the same SPL as the Focal. So even if the Morel has some shorting rings that might help it a bit, the Focal is likely to stay far cooler. And when the coil starts heating up, the sensitivity of the driver drops resulting in a slow modulation of the signal as heat reacts rather slowly.

Under hung
Another difference between the two is that the Morel has a coil that is shorter than the gap, while the Focal has one that is longer than its gap. It has been suggested that under hung designs (short coil in tall gap) are superior, but evidence tells us that this is far from a universal truth. For a motor that is not saturated, under hung designs struggle with flux escaping to other parts of the gap where there is no force field from the coil. An under hung design also means a short coil that can sometimes have extremely low power handling. On the other hand, an under hung design allows for a very short and therefore lightweight coil. A short coil can use thinner wire, gaining even more in the mass savings department. So the question should probably rather be if the motor is or can be saturated, and if the driver can achieve a sufficiently high efficiency and through that still easilt handle the heat of anything we throw at it, and a bit more, without noticeable power compression.

Inductance
In my article about inductance, I cover the most important bits of this. You can read it here:
https://jrkpost.medium.com/the-broken-transients-94b21e7282da
This is an important basic parameter for midrange drivers and lots of information can be read from how the designer has worked with inductance in the driver. My opinion is that a midrange should have low inductance, added shorting rings to further reduce inductance, and a saturated motor.

The cone
Some manufacturers do struggle to get certain cone materials to work in certain frames. Not that they do not fit, but they can not get them to perform as well as with other frames. This is not at all due to the frame. It is allways a question of what you glue directly to the cone. The specific frame might dictate the use of a different surround that is strengthened and optimized for bass applications. It we look at the dimensions and materials of the surround, and the size, material and number of layers of the coil, we get some good information about what the cone needs to carry around. This lets us know a great deal about how long it will take for it to exit all of its energy.

Imagine a fishing pole. We hold it in our hand, and wipes it around a bit. We stop the motion, and the remaining motion of the pole stops after a short time. We then add a weight to the far end of the pole to mimic a heavy suspension. We make the same motion, and we notice that the pole takes far more time to settle with this added mass. If we analyze the modal frequencies of the pole we will see that they actually depend significantly on the mass we have in both ends. Reducing the mass of the surround has a huge impact but reducing the mass of the coil also has a pretty noticable impact, and often far more can be done to the coil if it is really heavy to begin with. This is also where the amplifier “grabs” the cone, but the drivers resistance and any passive crossover will together produce a lowered mechanical damping factor, meaning that we are not really holding the pole, we rather have a mass at each end, and we apply some force to one of them, which is the coil.

So for a cone to perform at its best we should make sure it does not have to carry around lots of added mass. It has to carry its own mass and any added mass will just add more potential resonances. The Focal driver has a very light surround, and it could have had an even lighter coil but it is still a much lower moving mass driver than most of its current competitors. At 10,4g, for a driver that is in between a 6,5 and an 8 inch driver in size, that is pretty low.

When is too much?
It is possible to get even lower than that. Dayton has an 8 inch full range driver that has lower moving mass (8,4g), and that is not even close to setting any record. To reach these numbers you will off course have to lighten the coil as much as possible, but you also need to lighten the cone, a lot. A very thin cone has numerous resonances that are easily triggered by almost anything. It is like covering the ceiling of an opera house in different sized bells, hoping they won’t start ringing when music is performed. This is an effective way to get lots of high end output from a big cone, but with a less than optimal time domain performance, especially above the primary funamental breakup.

But a midrange does not have to enter its upper modal range, or break up as it is commonly called. We can keep it in the pass band range where the output level does not need to consist of modes. A slightly heavier and far stronger cone does the trick. Carefully balancing it between stiffness and damping is one good way to allow it to be crossed over relatively high. Another way is to make a cone that is so stiff, the pass band ends way above the cross over point. But as engineering requires some particular shapes and masses for a driver to work there is really a limit to how far we can go. Both of these methods allows for a much cleaner midrange. The damping that is naturally within a hard cone is really not a bad thing. We can not take advantage of a cone without damping unless it is infinitely hard, it would simply ring for seconds after you mute the music.

So what we should take from the full range driver is the low Mms in combination with what is often a fairly stable motor. They may actually often be poorly constructed, but the low mass means the job for the motor is so easy. Even the slightly increased mass of a midrange driver should be quite easy for a not overly advanced motor. But for som proper dynamic performance, it really pays off to go the extra mile on the motor as well.

To sum it up:
A true midrange should follow a set of design guidelines. They should be as follows:

1: The driver should not have any non subtile characteristics within its entire working range.

2: The coil mass should be less than 1g.

3: The driver should have a narrow cloth or similar low mass and low loss surround.

4: The total Mms should be less than 0,08g/cm².

5: Any steel present in the gap should be saturated to >1,9T.

6: The driver should use shorting rings inside the gap.

7: The coil should be wound with aluminium ribbon wire on a bobbin no more conductive than titanium.

8: The motor should allow for 2mm of excursion with less than 2,5% deviation from the Bl factor at its resting point, including both the effects of motion and AC current.

9: The efficiency of the driver should be at least 93dB/1W/1m and it should be able to handle transients at >114dB.

10: The cone should have no modes for at least 1 octave past its low pass crossover point and have a loss factor of less than 0,1.

Yes, you are absolutely right, this rules out all midranges ever built on several of these points. Even the first point rules out almost every midrange driver in production. But these are the things that really matter, and should be the guideline for designing midrange drivers. And they are all possible to realize within one single driver.

Let’s call this “The 10 commandments of true midrange”, and label any driver that is able to fulfill them “True midrange”.

Jørn Rune Kviserud
Tonalab
April 13. 2021