Welcome to today’s Tech Talk. In this installation, we’re continuing our dialog with Doug Button, Chief Engineer in the Corporate Technology Group at HARMAN Professional Solutions.
Last time we spoke with Doug, we delved into the way the evolution of technology often takes the easiest route and can miss opportunities for meaningful progress. Today, we’re focusing on the JBL CBT 1000, the latest of the JBL CBT Series and a project with which Doug challenged himself to deploy an older style of circuitry in a new way. His goal was to build a customizable line array column loudspeaker for such venues as theaters, auditoriums and houses of worship that doesn’t require computer configuring.
[MM] Why did the thought process behind the JBL CBT 1000 include envisioning optimum sound coverage that wasn’t computer reliant?
[DB] The idea behind the whole CBT Series is that many computer-based loudspeaker systems, while offering infinite flexibility in a variety of environments, are not as user-friendly as people would like. More and more, we hear audio technicians say they don’t want to have to be IT professionals to run a first-rate sound system.
The evolution of sound systems has vastly improved the quality of what we hear. But, for the last 30 years, the systems have become not only more and more computer-based, but also more complex and prone to problems and malfunctions. One of the risks you get with any computer-aided system, or any system with a computer, is that something might go wrong with the computer. There are gremlins in some of these systems, which really can be a pain. They can be really, really challenging.
About 10 years ago, when we first started working on the CBT Series, I spent a lot of time building computer-based systems and thought they seemed more complicated than necessary to accomplish the simple task of pointing sound at an audience. I kept wondering if it really took all this computer power to do it.
The big “what if” was whether we could develop a series of sophisticated line arrays while eliminating the computer from the user experience.
[MM] So, you’re saying that the CBT 1000 doesn’t require a computer?
[DB] We initially used a computer to help design the loudspeaker based on traditional parts, but that was it. Although it’s made of simple parts, the circuit in the CBT is very complicated, and we wouldn’t have been able to optimize it without a computer. However, while a lot of new systems need a computer to re-configure the loudspeaker every time it’s set up, the CBT never needs one again.
[MM] What motivated the development of the CBT Series?
[DB] We found there was kind of a sweet spot from a technology standpoint to be able to build a passively based loudspeaker [using an external power amplifier] with the flexibility of a digital system. My colleague Don Keele had written extensively about the CBT concept, which involved bending an array and using amplitude shading [turning down the drivers as you move outward from the middle]. Don and I found we could use digital delay to accomplish the same thing, but it would have required an active system.
We came up with a new approach that could mimic the behavior of the active CBT, but with passive components. We implemented the idea in the first version of the CBT system, which came out about seven years ago. The original CBT 100LA-1 has two basic configurations—wide and narrow—that are easily switched between and make it more versatile than a single-purpose loudspeaker. It was one small step toward a system that could be easily adapted to any environment. Additionally, we developed the CBT 70J-1 that similarly has the two coverage patterns.
Graphic user interface with orange jumpers that switch the banks of capacitors.
[MM] Mechanically, how were you able to achieve the shift using a switch?
[DB] It switched a large number of the components in the circuit. At the time, it seemed pretty complicated to switch all those circuits just to make two different coverage patterns. But, working on it through the years, we figured out that we could make a matrix from the circuit and have as many as 16 different coverage patterns just by changing a small patch bay of jumpers. That resulted in the development of the CBT 1000, which has 16 different coverage patterns, giving customers almost as much flexibility as a digital system.
[MM] Why would a user need 16 different coverage patterns?
[DB If you’re setting up a sound system in a relatively wide short room, it has completely different requirements than a speaker being used outdoors, having to throw 300 feet. We wanted a loudspeaker that could handle those extremes plus pretty much every possibility in-between. We figured 16 different amounts of projection would be enough to cover most applications.
There is also an equalization switch on the side of the CBT 1000, so you can change between a music and a speech setting. With the 16 different patterns and the EQ switch, there are actually 32 possible configurations.
We have a CBT Calculator app that helps users to figure out which coverage pattern will work best in a particular environment. It allows them to try the different available coverage patterns of the speaker and shows how the speaker will perform in their space. It lets the user place imaginary microphones in a room and see what the frequency response will be at each of them. There is also an EASE GLL plugin that allows you to use any of the possible pattern configurations in EASE [a computer modeling program].
[MM] In our previous conversation, we discussed reconsidering technologies that have been around for a while and how they might be applied for a new purpose. In the case of the CBT Series, can you explain the nature of the older circuits you used, why you went back to them and how they have worked for you?
JBL CBT 1000: First passive prototype
[DB] The basic idea behind the CBT is that I got to thinking about how classic inductors and capacitors have this thing called group delay. What group delay means is that when you pass a signal through these kinds of networks, the energy stays in the network for a small period of time and then leaves. It’s not necessarily flat with frequency, but they have this nature in them, which can be referred to as phase delay or group delay and represents a certain amount of time delay at a given frequency. The real trick was to figure out how to use those components and make the group delay flat with frequency, so it was basically the same as digital [time] delay.
In a digital system, it’s very easy, because you have a little memory chip. The signal comes in, sits on the memory chip for a little while and then leaves, but analog delay in old-fashioned circuits is actually more challenging and not an obvious way to go.
The way you do the pattern steering for the CBT arrays is with group delay. Just like with a digital system, the outer transducers are pulled back in time, so to speak. That’s what bends the array. The center speakers aren’t delayed, but as you go further toward the outside, those transducers receive more delay. As you change that delay, you change the effective curvature of the front of the loudspeaker. To do that with a digital system is really easy. You just have a separate digital delay for each of those drivers.
The architecture to do this [a transmission line], at least from a modeling standpoint, has existed for a very long time but, as far as I know, hasn’t been used for loudspeakers, so it was new application and architecture for those types of components.
The outcome has been the delivery of a simple to use, first-rate loudspeaker system. Now, it’s about figuring out what’s next.
Many thanks to Doug for taking the time to speak with us again and for his insights into the CBT Series product development process. Are you an inventor or engineer? Let us know what you find interesting about the integration of older and newer technologies in the comments.