GoodSoundClub - Romy the Cat's Audio Site

In the Forum: Audio Discussions
In the Thread: Constructing LF modules to the limits

Post Subject: The tone at LFsPosted by haralanov on: 9/25/2011

noviygera wrote:
I am very curious what "properly made" means when perusing the tonal attributes of LF. What properties or specifications other than driver diameter do you evaluate when perusing a better tone?

Herman, you are asking me where does the tone comes from in context of bass reproduction? Well, this is a multifaceted subject that takes huge amount of explanations in order to be understood properly. I will mention the very basic things that affect tone.

What we have in a bass driver? We have a moving cone. So if we want to have good and complex tone, we have to ensure the cone will be excited in accordance to the signal of the amplifier’s output terminals. And when we see all the influencing factors, we realize we are sunk deep in the swamp. The cone is excited by the voice coil movements. These movements should mimic the voltage waveform of the signal at the output of the feeding amplifier. Here is a short list of the factors that distort that accurate mimicking.

First of all, the voice coil should move in an absolutely uniform magnetic flux during its movement, no matter how loud the driver is being pushed. It is some kind of ideal theoretical situation which is never realized in practice. Most of the magnet systems have so unstable magnetic field in their gaps, that it is very funny to call this field “permanent/constant magnetic field”. One needs truly heroically constructed magnet system to assure stable working point of its magnet and to avoid field modulations in order to ensure a stable platform for achieving accurate VC movements. There is another problem here. Achieving stable magnetic field does not ensure this field is evenly distributed through the height of the gap. In most of the cases the induction varies to a large degree in the different sectors of the gap. The difference in the field at the bottom and at the top of the gap sometimes exceeds 40%!! You can easily imagine how much distortion could be produced with that kind of motor. So one needs smartly designed motor, where the magnetic field does not collapse when the field of the voice coil pushes against it, and also it must be the same in all points in the gap. This is not cheap to be produced. Until now we looked at only half part of the problem with achieving accuracy for VC movements. Let’s see the other half of the problem. Most of the bass drivers have their VCs longer (sometimes considerably longer) than the height of the magnetic gap. It creates a whole new world of problems and distortion forming mechanisms. One does not need to be very smart to realize that with that type of motors, the VC doesn’t “see” constant number of magnetic lines when moving back and forth. This is another process causing distortion. But wait, there is even more into it! The voice coil has parasitic electrical parameters. One of them is called inductance. The self inductance of the coil acts like a build in low-pass filter inside the driver. The problem is that when the coil moves, during its movements it sees different amount of steel around it. Steel affects the inductance value of the coil and if the steel is not of a constant amount around the coil when it is moving, the VC changes its inductance constantly, depending on its position in the gap. That means there is constantly changing crossover frequency due to inductance variations. It is called inductance modulation and affects the upper working range of the driver. So if one is looking for a bass driver which he intends to use with first order filter, he has to ensure there are no inductance modulations in order to prevent this new type of distortion mechanism.

How to eliminate all these variables in order to ensure proper excitation behavior on the cone? Well, there is only one really efficient way to do that. The solution is called underhung motor and it is used by some manufacturers in their drivers. Here we have a VC that is shorter than the height of the magnetic gap where it is moving. It sees constant amount of steel during its movement and that kills the inductance modulations. Also it is driven by a constant flux during its entire movement. It lowers some types of distortion to nearly zero level and eliminate transient compression. And this is very important.

At the second stage, we realize that although the coil moves in stable and constant magnetic field, its behavior is hugely dependent by its electrical interaction with the driving amplifier (that’s why the tone is dependent from the amplifier/speaker interaction). We need the current flowing in the voice coil to change in the same manner as the voltage waveform at the output of the driving amp. This is very difficult to happen because there are eddy currents flowing in the iron (this is the former material of the magnetic circuit) and there are hysteresis loses in that iron and all this distorts the current flowing in the coil. The human ear is very sensitive to that type of distortion, so all the good speakers have very serious magnet systems where some efforts are taken to break up these eddy currents and to lower that nasty sounding distortion mechanism. Speaking of speaker-amplifier interface, I have to mention that every amplifier need exact amount of wire length in the gap in order to drive the VC properly. That’s the reason why most (practically all) of the bass drivers do not sound good with solid state amps. It is not because the solid state amps are bad, but because they need low inductance/resistance voice coils in order to work/sound properly at LFs. There is also another phenomena: imagine you have two bass drivers with identical moving parts and with identical magnet systems. The difference between them is in the type of the voice coil. The first speaker has 2 layers with 30 turns per layer (60turns in total) and the second speaker has 1 layer of 60 turns of exactly the same wire. The DCR of the two coils is the same, but the 1-layer coil sounds much better compared to the other. But why?? Do not ask me why – I do not know. I only know it affects the tone a lot, and I simply get use of it, without any explanation why it happens.

When all of these 5-6-7 different types of distortion mechanisms are eliminated (I mean lowered), one has the opportunity to hear the unmasked/undistorted tone of his bass. I deliberately used the word “opportunity” because now we have to convert those accurate VC movements into sound waves and if we fail to do it, we will not get good tone out of our bass drivers. Good example of this kind of failure is JBL 1501Al bass driver. It has very very very smartly designed underhung motor, but its moving system is crap and the result is a lack of tone in its sound.

OK, now let’s imagine we have kind of ideal situation and our voice coil follows exactly the signal waveform and amplitude. We have to convert those movements in air molecules excitation, and this excitation should mimic exactly the VC movements. We further realize that we are not only deeply sunk in the swamp, but we are on the bottom of the swamp.

What we need to have uncompromised moving system? We need to attach to the VC a cone that has density equal to the air density and at the same time to have as much radiating area as possible in order to achieve good efficiency. That type of cone does not need any suspensions or electrical damping to reproduce the signal precisely. But unfortunately we do not live in an ideal world, so we need to think for a while how could we get as close as possible to the above mentioned criteria. So the first thing we need to do is to find a material having mechanical impedance as closer to the air mechanical impedance, while being rigid enough in order to maintain accurate behavior. We realize the only suitable material is long-fibered paper. Now we need to “invent” smart cone geometry to make the cone rigid enough by using smart construction geometry and not stiff materials. There are a LOT of tricks here, but I will not mention any of them. All this affects tone a LOT. But we need our cone to flex very gradually in order to avoid nasty break-up peaks, because every material on this planet flexes at some point when there is applied enough mechanical acceleration to it. Just a quick illustration: take a look at the following video (the part at 4:27-4:35):
http://www.youtube.com/watch?v=QFlEIybC7rU&feature=related
You can clearly see how this obviously rigid baseball bat flexes like an elephant ass when there is enough force applied to it. The same happens with speaker cones. The whole trick is to take control over these break-ups. It influences the tone a lot when the driver is being used over a wide range (this is the case with 1^st order filtration).

Now let’s imagine we have two identical voice coils. The first one drives 8” cone and the second one drives 18” cone. What happens? Since the bigger cone has more radiating area, it reproduces the LFs more efficiently and its tonal balance is centered at lower frequency compared to the small cone. The small cone has higher amplitude for the upper range of the input signal and its tonal balance is more upward oriented (Paul already mentioned that, and he very intentionally used irony when calling a 10 inch driver to be a “bass” driver) although it is perfectly suited to reproduce low frequencies if having low Fs. So properly designed big cones have better tone and higher efficiency at lower frequencies compared to small ones.
The less radiating area of smaller cones, means there is less coupling to the surrounding air. That means there is less air control to the cone movements and now these movements must be controlled by different ways. One needs harder suspended cone in order to control these movements. Of course this type of cone control has negative consequences to tone, because it acts with no own intelligence – it just blindly resists to these movements, applying a force in order to turn back the cone at its rest position. This process inserts its own tone and adds its taste to the real tone. Air loading does not have this behavior defect and it controls the cone in a smart way, applying no own intelligence, thus allowing one to hear more complexity of tone at LFs.

There is one drawback of using too big cones for bass. This drawback is due to the raised moving mass compared to a smaller cone. This mass stores kinetic energy and if this energy is not damped by the air/suspension, it makes the VC to generate more back voltage and sent it directly to the driving amplifier in different point in time (after the original signal that feeds the VC). The amplifier should be able to absorb that energy. Sometimes this is a big stress for smaller amplifiers and they go really crazy while trying to absorb that energy, especially at Fs of the system, where the returned energy is returned in different points in time. The more powerful the amp is, the better it resist that sent back energy coming from the bass driver’s VC. By the way, that’s one of the main reasons why low powered amps cannot drive the big woofers with confidence:
http://www.goodsoundclub.com/Forums/ShowPost.aspx?postID=9390#9390

Now when we have good driver with high tonal potential at LFs, we have to find a way to use it. We have to prevent acoustic shortening effect. There are a lot of possibilities how to achieve that. Every solution has its own character, because its affect the cone behavior. The best way to ensure undistorted cone movements, is to load the cone symmetrically. The air loading from behind the cone should be the same as the air loading in front of the cone. This is satisfied with infinite baffle configuration. In this situation there is a lot of air behind the cone and it breathes freely in the air, providing pure tone to the listener. Every closed volume of air behind the cone manifests itself like a breath-blocking device, no matter how damped it is from the inside. Make a simple experiment: take huge amount of damping material and insert it in your wardrobe. Now put your head inside and listen carefully. You feel your head is inside a small and highly damped chamber. Very unnatural feeling, isn’t it? Now take your head out of your wardrobe and go outside in your garden. How do you feel now? Do you feel the freedom and lack of closed and stuffed chambers? Yes? Well, the same is true in context of your bass driver operation. You can always recognize there is trapped air volume behind the cone. This affects tone too, although there are no air-born resonances, due to the heavy internal acoustic damping.

The vibrations of the acoustic shield (the “box”) also destroy the Tone. If one wants to hear real bass tone, he must take all those sound-poisoning vibrations out of the picture. The material that really do this job well is the sand. I know there are some people who state, the acoustically dead/inert cabinets destroy tone. And the fun part is they really do not imagine it – it really happens sometimes, but there is a reason why it happens. The reason is because they are Гевреци/Gevreks (Morons) and they use tonally dead drivers. What else could they expect, besides dead tone when using tonally dead driver in acoustically inert enclosure?? So they “fix” the problem by using mechanically resonant enclosures, calling them “musical instruments”!! These wide-band resonators really make their dead drivers to sound more alive and with more “tone” but in reality it is just the opposite situation – they have nearly zero tonal complexity, because the bass tone is highly polluted with cabinet colorations. The lower the parasitic resonances, the more complex the tone becomes.

The type of filtering also affects LF tone. With first order low-passing, the bass channel sounds like the missing part of my widerange/upperbass channels combo, but with sharp filter it sounds much like a self contained sound that is not part of nothing. It lives its own life and although it could be +/- integrated with the other channels, this integration is light years away from 1^st order filtering integration abilities.

There are many other factors, affecting the tone at LFs, but the people who are really interested into this, are going to learn those factors by themselves.

Best regards,
Petar Haralanov

Rerurn to Romy the Cat's Site