STILL waiting

I was at a book store last night and found a pretty good book on car audio systems, published by TAB. It discussed the BOSE systems only briefly, but in sufficient detail that I was able to derive certain conclusions with sufficient confidence to publish them here. Specifically, I will discuss in some detail the architecture of this system to support the following conclusions:

1. It should not be at all difficult to keep the stock head end but replace the amp/eq and the speakers.
2. It will be highly impractical to attempt to replace either the stock speakers or the amp/eq but not both.
3. Another very practical option is to replace either the front half or the rear half of the system with a new amp and new speakers, while leaving the other half BOSE.

Before I start to explain these conclusions, let me first say that I am not by any means an advocate of BOSE, but I do think that my reaction to BOSE is less emotional and more pragmatic that the reactions that have been displayed by many people. I don't personally care for their home systems, and I regard their "surround systems" for the home as a farce. Nevertheless, I find their automobile systems to have certain pragmatic advantages, and although I don't believe that they could compete with custom installations exceeding a grand in cost, I feel that the trade-offs are not as severe as the audiophiles typically represent, and that they are a pragmatic, cost-effective engineering solution for the automotive environment.

According to the book that I read, the BOSE systems invariably use a "differential" signal between the head end and the amp/eq. What this is, simply, is a pair of wires for each signal, as opposed to the more usual single wire per signal with a common return or "ground". In a differential signal, in the simplest case, the two leads each carry the same voltage as the other, but of opposite value, i.e., +V and -V. The reason for this is that any noise that is present will have a nearly identical effect on both lines, and thus will cancel out when the difference between the two lines is taken. The approach is similar to the use of coax when in fact the shield transmits a signal of opposite voltage than the core conductor, but not that when coax is used in home stereos for the "line-level" signals that use RCA connectors, that this is ordinarily not a differential signal, and the shield merely serves as a shield and common return path or reference point, being grounded at one or both ends. For computer geeks that have looked at the difference between RS-232 and RS-449, they know that the difference is mainly that instead of there being only a single common ground, that there is a dedicated pair of wires for each of the signals, and that the sum of the voltage on each such pair is 0. Note also that the differential pair does not absolutely have to sum to 0; the sum could instead be a fixed DC bias, or a "floating ground" as it is sometimes called. For example, if both signals have a DC bias of +6 volts, when the difference between the two is taken, the DC bias cancels out. What is important is that one of the two deviates from the DC reference by a given positive voltage, the other deviates from the DC reference by the negative of that same absolute voltage.

The total number of signal wires from the head to the amp should then be exactly 8: since the head has a functioning fader control, a differential pair of wires must exist for each of RF, LF, RR, LR. The really compelling question is what are the impedance requirements of the output and what is the votage range or maximum voltage of the output. Here is where I will make a bit of a leap into what I do not know absolutely, but what seems entirely likely to me because it is in agreement with the assertion that the head unit is the same whether it is BOSE or not, and also because it is an entirely pragmatic solution. What I believe is that these differential outputs are capable of driving an ordinary 4 Ohm car speaker directly, while being acceptable as the inputs for the BOSE amp/eq. Consider that first all that the current that flows through the voice coil depends on the difference in voltage at its two ends. Also, an identical DC bias on both lines presents no problem whatsoever, because it does not cause a voltage difference across the voice coil. (In a common, non-differential configuration, if the signal line contains a DC bias, it will not cause a motive force on the voice coil, because only a time rate of change of the electrical field produces a magnetic field, but it will cause DC current and heating nonetheless, so either the same DC bias must be applied to the "ground", or else a blocking capacitor must be wired in series, and if it is an electrolytic capacitor it must be oriented in the correct direction.) Thus, as long as these differential outputs are of sufficient voltage to get adequate power out of a 4 Ohm speaker, and will supply that sifficient voltage at that nominal impedance, then these signals are perfectly adequate for driving the speakers directly.

One consequence of this is that the voltage levels will exceed the usual input voltage for an amplifier, but the compensation for this in the amplifier is trivial. Another possible and interesting consequence is revealed when you ask what is the maximum output level of the amplifier that is built into the head end unit. One of the major differences in these built-amps and separate amps is that the separate amps have DC-DC convertors to increase the output voltage above the 12V battery. The reason is simple: power is V^2/R, which limits the power consumption (not acoustic output) of a 4 Ohm speaker to about 40 watts. That is actually quite a bit for the interior of a car for us old peoples, but not nearly enough to transmit the deep bass through the tires and the earth effectively the way that the kids like nowadays. So there are basically two solutions: either increase the voltage, or else decrease the impedance of the driver. The conventional solution for separate amps is to employ a DC-to-DC convertor, but BOSE uses the other approach. Drivers with very low impedance preclude the need for a DC-DC convertor in the outboard BOSE amp/eq. Consider what is needed to convert 12V DC to the 20V level that is needed in order to consume 100 Watts in a 4 Ohm voice coil. A voltage transformer is needed (one of those things that has a lot of wire wrapped around a loop made of ferrite, or rather, has two ferrite loops wrapped around a bundle of loops of conductor, or whatever). A voltage transformer, however, works according to Maxwell's laws, which state that a time rate of change of an electrical field produces a magnetic field that oscillates in an orientation that is orthogonal to the orientation of oscillation of the electical field ...) Anyway, the point is that you have to first convert the DC to AC, which requires an oscillator capable handling the high currents, and then the transformer that must also handle the high currents, and then you have to convert the output AC back to DC, which requires some diodes capable of handling the high currents, and some filtering capacitors capable of handling the high currents. An alternative solution would be to put a great big DC-AC converter capable of providing several Ohms at 120AC house voltage, and then use regular home components. Anyway, the bottom line is that since head-end units don't have DC-DC convertors, they can't produce deafening volume levels from 4 Ohm speakers. You might then ask why 4 Ohm is the standard for car speakers. There are probably a couple of reasons for this, one of which is that at one point it was generally believed that 40 Watts of power consumption per driver was way more than enough. Another reason is possibly that it is difficult to manufacture an efficient speaker with a nominal impedance less than 1 Ohm.

In any case, if what you want to do is replace the amps (and the speakers), this should not be difficult. There are several manufacturers that make little convertor thingies to convert speaker level signals to line level signals, and the cost is about $30. To do both the front and the rear you would need two of these. Contrary to what some people will claim, these passive gadgets should cause much less signal quality degradation than the amplifier itself. I can't say for certain, but it seems to me that a differential signal should not present any special requirements for these gadgets, so that any of these gadgets should work and you should not need one made specifically for BOSE. You could easily determine this experimentally (I may do this sometime after it gets warmer) because you will not damage the head end by connecting it to the input of this gadget, and then you can measure the voltage level at the output and verify that it does not exceed the maximum input voltage for your amp.

Replacing the speakers while keeping the BOSE amp/eq, or replacing the BOSE amp/eq while keeping the speakers, is another matter however. The maximum output voltage of the BOSE amp/eq will not be sufficient to get acceptable power from a 4 Ohm speaker. Conversely, if you connect a standard amp that expects 4 Ohms to a BOSE speaker that may present less than .5 Ohm, either the output transistors in the amp will fry or else the protection circuitry in the amp will kick in, but it will not work regardless. Additionally, the EQ that is done by the amp in all likelihood is specific to the speakers, as opposed to merely the acoustics characteristics of the interior of the car.

Another alternative that may be a good solution for anyone that dislikes only part of the system is to replace only the part that you don't like. For example, if the tweeters sound harsh to you or do not seem to reach the highest frequencies, you could replace only the front portion and leave the rear portion alone. What you would do is simply disconnect (or sever) the appropriate input from the BOSE amp/eq and instead connect one of those signal convertor gadgets, an equalizer if you like, one or more amps, and speakers of your choice. You would thus disable whatever speakers are driver by whatever individual amps are fed from the input that you disconnected. Someone reported that the amp is a 6-ch amp. I can not say how these are divided among the speakers, but suppose for the sake of discussions that one stereo pair provides the front speakers, the other pair the rear passenger door speakers, and the third a woofer that has dual voice coils. Since there are only four inputs, in this case I would expect that both the rear passenger door speakers and the subwoofer are fed via the same input, so that the fader control would work as expected. In this case, if you severed the input for the rear, you would lose both the rear door speakers and the subwoofers, and would need to provide new replacement speakers for these as well as one or more amps and possibly a crossover.

Lastly, I want to say something about the tweeters that BOSE uses. Many people criticize BOSE for using paper cone tweeters while every other manufacturer seems to use exclusively domes. There is a common misconception that the reason for the dome has to do with the dispersion of the high frequencies, but according to what I have read, the dome shape has no significant influence on the radiation pattern. The radiation pattern is determined almost entirely by the relationship between the wave length and the diameter of the driver. The real reason for the dome shape is that, for a given weight, it is more mechanically rigid, and thus produces less distortion, than a cone. A lot of dome tweeters are in fact a combination of dome and a little bit of a cone around it. Historically, cone tweeters had a little dome over the voice coil, that was called the dust cap, but that contributed to the radiating surface nonetheless. A dome tweeter is realy just a cone tweeter that has had the voice coil enlarged so that the dome extends all the way to the edge of the cone. The original reason for doing this was simply that the improved rigidity allowed a softer material to be used than was possible with a cone, and thus better intrinsic damping and less distortion. In my personal view, the single advantage of a dome tweeter is that it allows a soft, damped material to be used, and if I did happen to like the sound of most of the metallic domes that are being used in home speakers nowadays, I would wonder why they aren't simply cones. The console stereo that my parents bought back in the early sixties had metal cone tweeters and they sounded, well, sort of metallic. Anyway, it just isn't true that in general dome tweeters are superior to cone tweeters. Once again, BOSE has opted for a pragmatic, cost-effective solution. Having said that, I will say that if I change the BOSE system at all, it will be to change the entire front portion so that I can put in some tweeters that have the sort of sweet sound that I love.

anonymous

Try:

realistic, matter-of-fact, practical, sober,
unfantastic, unidealistic, simple

In the context of your article,

cost-effective, best compromise

also work.

lionsfan54

Surround sound a 'farce'? You haven't heard the right system yet

STILL waiting

multi-channel surround sound is the BOSE system compliant with? Dolby Digital (5.1), I suppose? The BOSE "surround sound" system is not even based on true multi-channel source. It is a synthesis of something that is supposed to emulate something that I cannot even fathom. whatever it is supposed to be, the whole idea is so technically absurd that I can only call it a "farce".

lionsfan54

I agree with that sorry, I thought you were trashing all surround systems

San Dimas S4

wwwhhhhoooo dude! I don't understand