by Constantine Soo
For the most part, advancements in the audio hobby are often brought about by revolutionary approaches developed by visionary engineers. Oftentimes, it takes new ideas and talents from the field of electrical and mechanical engineering to realize appreciable improvement in reproduced sound, and papers from these distinguished minds have been continuing to inspire the industry as a whole.
Now, for the first time, we shall read from Robert Palma, a NASA scientist designing rocket/satellite instrumentations, on applying rocket science to our humble hobby. In addition to audio, Robert also shares some interesting perspective of NASA with us.
By the way, the term “Extreme Audio” is a Robert Palma-original.
Many of the missions of “outer space” involve engineering that is beyond the “state of the art” – the so-called “bleeding edge” of technology. In their initial developments, the Global Positioning System, developed at U.S. Naval Research Laboratory in the 1970’s and Space Telescope developed by NASA in the late 1960’s, were examples of this bleeding edge engineering. I bet you didn’t know that these two systems are that old!
While such projects may or may not have been blessed with unlimited budgets, the need for better-than-the-best engineering was a “given”, due to the technical requirements of the system. In addition, economics can drive the technical development to extremes. Here’s how.
Throughout the NASA Space Shuttle transportation program, there have been missions where some space system (satellite) needed repair. Astronauts were sent to do the job. With the exceptions of “space station” types of missions and the Space Telescope program, these “repair” missions, in reality, were astronaut training programs or NASA publicity stunts. The truth here is that the cost to develop special equipment and to train the astronauts for a particular “repair” mission, is considerably more than the cost to build and launch another copy of the broken satellite.
I needed to mention the above in order to make the following point. When designing and developing satellites and rockets, there is a pressing need to affect a perfect or near-perfect design. Engineers and scientists will go to extreme ends to ensure that the first, and often, the only production item of a satellite, etc, works exactly as planned. They will focus, and sometimes obsess on the smallest detail. They know that the tiniest design mistake can bring the whole spacecraft down, and if it stops working, you cannot, economically, send in a repair man.
One of my fields of expertise is instrumentation. This is the measurement of some physical parameter (voltage, current, temperature, pressure, acceleration, etc.) and the conveying of that “value” to another system for processing. Sometimes the sensor that is sensing the actual physical parameter is physically close to the processing electronics. At other times it is not. In this latter case, it is often crucial to “read” the sensors value and not “read” the noise and interference that can be easily picked up in the transmission cabling that connects the sensor with the processing electronics.
One such design was the shock and vibration sensor used on the main engine of a launch vehicle (rocket). This sensor was a strain gauge bridge that produced only micro-volts of signal and was sampled at 200,000 samples per second. It was necessary to transmit that very small, wide-band signal through 80 feet of cabling to the processing electronics. It was necessary, of course, to avoid the pick up of noise and interference along the way. This was a challenging design! It was this design that set the stage for the audiophile product, Digital Precise.
The approach for both of these designs (the rocket sensor and the Digital Precise) was to transmit the raw signal without any additional amplification or encoding. In any system of electronics where there is more than one “electronics box/chassis”, we would refer to as a “distributed system”. In a distributed system, it is easy for “ground noise” to creep into the signal path and corrupt the signal that you are trying to receive and read. Although on a much smaller scale than a space launch vehicle, an audiophile system with as little as two chassis can be called a distributed system and therefore faces the very same ground noise problem.
Ground noise is the result of stray currents flowing in the various chassis. These currents can be the result of: the intended signals that are handled by that chassis, power supplies and power components (relays, solenoids, motors, etc.), and externally induced signals from the immediate environment. Whatever the sources, ground noise is always present, and in a high-tech environment, almost always a problem.
In addition to ground noise, radiated interference can couple directly into the transmission cabling and corrupt the signal. The list of sources of radiated interference is long and growing every day. It begins with the ominous and all-pervading power line frequencies (50/60 Hz), and their harmonics, and goes through the many types of radio frequency transmitters such as two-way voice, broadcast, data, telco microwave, wireless Internet, airport radar, and yes, microwave ovens.
Protecting against radiated interference is a difficult challenge when either the frequency span of the interfering signals is great, or the amplitude of the interference is great.
Another important parameter in the pursuit of minimizing radiated interference effects involves the “field” of the interference. Is it primarily a magnetic field, and electric field or a “far field”, i.e. an electro-magnetic field? The Digital Precise had to deal with all of these.
How did the Digital Precise happen?
One day at lunch, Greg, my brother and president of our new company, Noble Electronics, started talking about chassis-to-chassis SPDIF connections. He talked about various noise problems that were inherent to the system and mentioned calls from his customers of Noble/Boelen Electronics who were asking about a “better mousetrap”.
I asked Greg to describe the various industry approaches to SPDIF chassis-to-chassis connections that were popular at the current time.
He described an approach. I responded, “that’s lame, it distorts the signal”.
He described another. I responded again with “that’s lame, it does not reject electric or magnetic fields”. He described yet another, to which I responded, “that’s really lame, your signal is traversing through a puddle of ground noise”.
At this time, Greg took a clean paper napkin and a ball point pen and pushed it across the table at me. He said “okay, show me a good approach”.
I pushed the napkin and pen back at him, complaining that real engineers only draw with pencils.
He responded, “give me a break, genius, show me”.
Since only a Neanderthal who was having a really bad hair day would refuse his own brother, I said “okay, we’ll do it just like the vibration sensor on a rockets main engine”. Thus born the Digital Precise.
The Digital Precise combines a high degree of balance and isolation to eliminate virtually all ground noise. Transmission line “pairing” (XLR, balanced) brings about magnetic field rejection. Faraday shielding brings about electric field and electro-magnetic field shielding. The inherent high-pass filter response of the components adds additional rejection to power line frequencies.
The care, attention, and even obsession to precision and detail found in the world of “rocket science” is but the same intelligence, focus and passion that underlies the world of “extreme audio”. Yes, they are different worlds, but the relentless pursuit of performance excellence is the very same.
- (Page 1 of 1)