Best Sound in a Vehicle - Part I - The Head Unit

Not just any head-unit

Click to launch installation video

To create the very best sound system, every part of the signal chain demands exceptional attention to detail. After years of progressive electronics improvements, I know that small changes aggregate in such a way that the resulting musical experience bears no resemblance to initial impressions from original equipment. This belief guided Paul and I as we went on to construct the best vehicle in the world. We started with the best equipment available at the time and modified it until it sounded better, was more reliable, became much more beautiful and was markedly easier to use. The following paragraphs describe the changes made to the system's head-unit.

Fujitsu Ten Head-unit

Everything started when we received a Fujitsu Ten head-unit for the van. In the US, Fujitsu Ten sold the same head unit under the Eclipse brand, which in 2010 went into bankruptcy.

Back in 1996, the Japanese factory reverted to their proud corporate name for the European market. It had something to do with the fact that the word Ten means Heaven in Japanese.
Needless to say, everybody over there wanted American Eclipse head-units and hated the idea of owning a Fujitsu Ten. But I knew they were both made with the same exceptional quality. So, I had no problem using one as the starting point of our new source unit. After all, it was Fujitsu Ten who manufactured many of the best head-units at the time. 

Today, CD players use digital processors capable of much higher resolutions. But at the time, there was nothing better than the Eclipse... I mean Fujitsu Ten.

Transport

The truth is that Paul and I were mainly interested in the Eclipse's transport and D/A Converter. These were as good as there was at the time. Moreover, the areas of improvement that represented the low hanging fruiit where elsewhere. 

We knew that by reducing power source ripple and minimizing the transport's electric load, we could end up with a very good and stable digital unit. 

To reduce ripple noise, we used a very large capacitor/inductor network. An 8 AWG inductor is first placed in series with the load on the positive cable going to the head unit. Next, a large capacitor is wired in parallel running in between the positive and negative power cables. You then repeat the same inductor in series plus capacitor in parallel filter. While overkill, this setup proved quite effective at stiffening the head unit's power supply. 

To reduce the transport's electric load, we added a custom external power supply that drove an external pre-amplifier. The power supply was wired separate from the transport and had its own capacitor/inductor network.

After the modifications, the CD player lost all of its operating features. AM/FM radio, Volume, Balance, Fader and Tone controls were no longer active. In a nutshell, all that remained was the transport. All the duties of the Display, Controls, Volume and Output, were handled by separate devices placed elsewhere in the van. 

Center Console

The CD transport, the pre-amplifier and its power supply were installed within a new center console. Their 3" height determined the minimum internal dimensions possible for the console. Since we wanted the console to have the least effect over the sound of the floor speakers, we needed it to be very small. In the end, we made it about 3 3/4" high. 

After we finished, there was really nothing visible that would indicate that the electronics were contained within the unusually low console. There was only a very small slot facing the driver side.
A slot was precision cut out of a black acrylic layer. So, considering that the interior of the van was all black, it was difficult for anyone to know that the CD player was housed there. This was the maximum expression of a stealth installation. 

Despite the unusual nature of the location, CD's were surprisingly easy to insert without looking. The reality was that placing the transport there was ergonomically superior to the more common dashboard location. In any case, it was very cool to take a CD and insert it into the invisible slot. It was as if the discs magically disappeared within the center console.

Pre-amplifier

If there is something that we all learned from Richard Clark was the value of a constant high voltage output from a head unit. This allows the signal to display a much higher degree of immunity to induced noise than is normally the case. Furthermore, a constant signal level means that processors behave much more linearly than they would otherwise. On its own, music is very dynamic. This poses a massive challenge to any power supply. But if you then add a volume control, you ask for the impossible. The dynamic range and response speed of such supply would be unique to say the least. Without such demand, dynamics improve as the signal is pushed hard through signal processors. 

The preamplifier took the output directly from the OpAmp buffering the D/A converter located on the unit's transport circuit board. As it is normal for me, we changed the buffer from the piece of garbage use by Eclipse to a superb Burr Brown. At the time, Burr Brown was yet to be bought by Texas Instruments. With Analog Devices, they represented the highest standard in performance. In comparison to other OpAmps, Burr Browns were sweet, spacious and large. But don't ask me which one we used. I can honestly not remember. All I remember is that it operated with high bias and that it was orgasmic in the way it played music.

By using a very small high quality cable, we reduced the signal's path length and bypassed  all the unnecessary radio and preamp features offered by Eclipse. In other words, there were no Volume, Fader, Bass boost or other features that could distort the signal. The difference in sound when one does this is incredible. The degree of transparency achieved is much higher.

To create the preamplifier, we started with a Lanzar six channel line driver. We disconnected the internal power supply and connected the board to our new external +/- 15V regulated power supply sporting a very low output impedance. Then, the input and four of the outputs were removed. Finally, two single channel Burr Browns were used in place of two of the original dual channel OpAmps. This meant that we ended up with a true dual mono line driver. Our goal was to improve stereo separation. In high definition systems, better separation and symmetry result in solid imaging and an expansive stage.

We then cranked up the gain until the Burr Browns begged for mercy. The line run hot at all times. There was no attenuator at the head unit. Instead, we installed ALPS motorized attenuators after the processors and right before the amplifiers. With what I know now, I would have used a better volume control. The difference between the ALPS and the one I have in my system now is like night and day. The ALPS lacks transparency. Here, consider the fact that I paid about 40 times more to build my present volume pot than to buy the ALPS already made. So the comparison isn't fair.
Likewise, comparing the ALPS with pots commonly found in car audio equipment is like comparing an MP3 with a High Resolution signal. There is just no comparison. The ALPS is much better than regular car stuff. Several layers of grain were removed thanks to the ALPS.

LCD Display

Can you imagine what it would be like to try to navigate CD tracks without a display. We took the front circuit board from the Eclipse unit and placed it behind a custom piece of sculpted acrylic in the upper center section of the dash. Because the acrylic piece was painted black from behind, the display was completely invisible when off. As the head unit was energized, the display would appear, seemingly out of nowhere. This perfectly matched the disappearing CD trick described above. 

Unfortunately, to simply wire the front circuit board that housed the display to the body of the CD player would result in failure. How do I know? Well, because we failed when we tried it. What happened was that the long cables acted as capacitors. So, after we would remotely activate the "track up" feature, for example, the stored capacitance would keep it on; thus stopping all other features from working. Incredibly, the cables held enough capacitive energy to last for quite a while. This meant that we had to use a different solution. Let's just say that we ended up using telecommunication micro-relays to make the controls work. 

Control Panel

The one part of the head unit that users experience directly was the control panel. This one was located in the best ergonomic location possible. While seated, the vehicle's driver would just slowly rest his right arm and, where the hand landed, that was where the controls were. 

We created a durable membrane that was stretched inside of the panel's top lid. Behind it, an aluminum frame holding low-travel (0.5 mm) precision switches held the rest in place. By pressing the membrane over each of the illustrated button images, the switches below would be activated. Each of these switches would then control the micro relays located behind the display board. Then, the volume up and down switches controlled the motors of the attenuators located at the amplifiers.
In this way, all volume, track and power on/off functions were centralized within easy reach of the users. This meant that a seating listener would insert the CD through the stealth slot near his knee. He would then operate the system through the control board while visually confirming everything through the display located just below his sight. When compared to other vehicles, the experience of listening to the van was indeed exceptional right from the moment the disc was inserted and the system turned on.

Ventilation System

Part to impress the judges and part to prevent any possible issues with heat, Paul and I created a ventilation system that would keep the transport, the power supply and the pre-amp running cool. We used very quiet fans that could be turned off during listening sessions. 

Fuse Protection

Finally, we had to have the best protection. While most sound-off competitors create installations without access to the radio's fuses, our van was designed to be much better. All fuses where placed right where the driver's left hand would hang when not holding the steering wheel. In this way, the van's conductor would have access to all components of the head unit without having to exit the vehicle. Everything was easily reachable from a comfortable sitting position. 

Ergonomics

And there was nothing as comfortable as seating in the van. We raised the floor to ensure maximum leg comfort. We selected firm, wide Recaro seats for the beast support. These seats were located very close to each other to improve speaker path length symmetry. Finally, all windows and windshield were darkened with 30% film. To me, there is nothing like the impression of peace and quiet that results from the psycho-acoustical response to darkness. Think of light as energy. Then think of this energy as noise that interferes with the music. Light energy is, after all, uncorrelated to music.

How does it compare?

While Pioneer had already created its fantastic ODR system by the time we built the van, business issues meant that we would not be able to use it in the UK. Nonetheless, I am comfortable stating that our final head unit largely outperformed even a hot-roded ODR. Chuck Barbosa used such ODR, modified with Burr Browns and superior components. But the musicality and the dynamics from our Eclipse unit were simply out of this world in comparison. The only feature where the ODR won hands down was in time compensation. The Eclipse didn't offer any. Thankfully, time compensation was not needed. Paul and I selected speaker locations with the longest path lengths possible within any car. The resulting symmetry was unique. Then, speaker locations compensated for the natural delays affecting all speakers driven with low-pass signals from a crossover. You see, time delays are more valuable as the system quality decreases. This is why high end home speakers like Wilson Audio's achieve the most three dimensional musical reproduction without digital time compensation. 

Thanks to speaker locations, the van's boundaries disappeared when playing music. The stage's width and depth seemed to extend into the horizon. It was like listening to an orchestra in an open amphitheater rather than the usual enclosed theater. In comparison, the best cars at the time sounded like lunch boxes. Without proper speaker alignment, this would have never been possible. But rather than using a patch for the problem, Paul and I eliminated the problem right from the beginning. 

In the end, Paul and I bridged the gap created by a massive ocean that separates our two nations. We brought the best from British and American know-how. We created a vehicle that although American in heritage (it was a Chevy Astro van with a Corvette engine after all), it was very British. The sound system demonstrated a complete lack of fear of innovation. It was quirky as many British masterpieces are. It was as elegant as a James Bond super car. Finally, it created beautiful music as only the Brits know how to. Oh, how I miss that sound!



February 14, 2014 update:

After publishing this article, I received what I thought was a great question from a close friend whom I admire very much: über world champion Ron Baker. I hope that you find our exchange insightful.

Ron Baker:
"Speaking of head units, would you say that the head unit and processor are the more important items in the signal path?  Then, within the head unit, I would suspect that the most important part is the transport. Finally, the processor would do the heavy lifting."

Alberto Lopez:
"Ron, I am not sure that any one part is superior. I think that when we listen to reproduced music, we in fact hear a layered picture made up from the combination of the characteristic sound of all components in the signal chain. Improving one makes the new picture better but it isn't free from the problems of the other parts. 
If my thesis is correct, then it would make sense to try to improve every part of the system. Moreover, it would also suggest that fewer, as opposed to more, system components in the chain should perform much better. Both of these observations are supported by my empirical observations."