Chris Brunhaver is the head loudspeaker designer at PS Audio. Chris will be talking about speakers and much more in our new “Speaker Builder” Column. His early background was covered in Copper Issue 112 in “Attack of the 20-Foot Tweeter” and we interview him here.
Frank Doris: What do you do at PS Audio?
Chris Brunhaver: My official title is senior speaker design engineer. I’ve had some really big shoes to fill because my predecessor was Arnie Nudell (co-founder of Infinity Systems and Genesis Technologies along with Paul McGowan), and this is kind of the gig of a lifetime for me, doing design work once more.
FD: Had you ever met Arnie?
CB: I grew up in the Seattle area and one of my first loudspeaker jobs was at Adire Audio. Among other things they did research in magnetics. They got hooked up with Gary Koh (current president of Genesis) and Arnie was doing some work for him at the time. Adire was working on amplifiers and woofers for Gary and I met Arnie at a CES. Quite a character!
FD: One of those larger than life people.
CB: He was one of those industry giants and was a lot of fun to meet. He left me with big shoes to fill but I’m very excited about it.
FD: When did you get into audio, and speaker design in particular? Did you start taking radios apart when you were five years old or something like that?
CB: Oh yeah; my parents called me Mr. Break-It because I would take things apart and they wouldn’t quite fit back together (laughter). But my dad had a speaker company, Speakerlab – he bought it in the early 1980s. They had factory stores and were offering kit speakers, hobbyist stuff, and dad took those and turned them into a finished speaker line for sale at retail. At one point they were doing cabinets for JBL in addition to their own stuff. Later, they became purely a retail shop, so I had some exposure to both the manufacturing and the retail sides of hi-fi.
I was a music student in college [Chris is an accomplished acoustic and electric bass player – Ed.] but went right back into loudspeaker stuff after college.
I always had a mindset of making things. I was around a place that manufactured loudspeaker drivers, and that was a big influence.
The engineer from my dad’s company was a guy named David Graebener, and he was one of the founders of Speakerlab before my dad bought it. They couldn’t really keep David busy, though, so David also did consulting work. One of his projects was with Carver, where he developed this planar magnetic speaker that became The Amazing Loudspeaker. A big surfboard with two sections and four 12-inch woofers!
David and Jim Croft did The Amazing Loudspeaker, and David then came back to Speakerlab and said, “I’ve got some ideas on how to make the speaker better,” (in particular, the method for tensioning the drivers). My dad said OK and spent $300,000 trying to refine that and make their own version of the speaker. Ultimately it still had some issues, so my dad gave David some of the tooling and David got some investment money and started Bohlender Graebener. I later went to work there as an engineer.
Genesis used both the Carver and the BG ribbons in one of their speakers, so even though I wasn’t working for Genesis directly, I had been quite close to it for years. BG wound up moving into custom installation and home theater stuff and designed a number of higher sensitivity/higher output speaker solutions, and I have carried those innovations forward but into the high-end audio context.
FD: So the upcoming PS Audio FR-30 speaker you’re working on is in a sense an evolution of this background.
CB: Everything that someone does is the cumulative result of their previous efforts. Arnie and Infinity did the EMIM and EMIT ribbon midrange and tweeter drivers and all that stuff in the 1980s and my dad’s stuff and the Carver stuff was in the late 1980s and 1990s, and I didn’t start to do my stuff until the 2000s, but it’s been refined a lot and now I’m taking it further. It’s a case of standing on the shoulders of giants and has been a life’s work. I guess all of us are lifers in this business in some way!
Without giving away any trade secrets, what’s different about the FR-30 than what’s come before?
CB: In concept, a lot of speaker technology has been in place for decades, but a lot of what makes speakers better today are advances in computer simulation, better analytical tools and measurement gear, and in materials science - better materials and better understanding about how they interact.
When working for Adire Audio, they were doing a lot of work on linear motor structures in woofers and had developed some proprietary FEA (finite element analysis) tools and motor topologies and methods. They created a split magnetic gap technology they call XBL (for eXtreme BL Linearity, a method of flattening the loudspeaker driver’s BL – motor strength –curve to yield lower distortion) I’ve leveraged a lot of those techniques along with things like Faraday (magnetic control) rings and custom motor assembly suspensions, for significantly lower distortion and increased dynamic range and output.
When I look at a modern-day planar magnetic driver compared to what was used in the old Infinity EMIMs and EMITs, we now have driver units that require less than one-tenth the power for a given output, and offer far lower distortion.
The system configuration of the FR-30 is different from previous designs. Over the course of 30 years or so, a lot of research has been done about perception and the way listeners hear sound, as well as things like, what’s the ideal target response of a speaker and its ideal dispersion, from guys like Floyd Toole at the NRC (the National Research Council in Canada) and the work he and a number of others did at Harman.
When you compare prior speakers to modern designs, there are multiple valid approaches but one of the key concepts is to look at the entire sound field of the speaker, not just its on-axis response. Speakers are super-challenging to design because they deviate so far from the ideal as compared to things like a DAC or an amplifier. They’re just really imperfect devices.
FD: A speaker designer once told me he wants to cry every time he thinks about how much sound is getting lost in the enclosure.
CB: That’s why I’m excited about speaker design in general as there’s lots of room for optimization and doing it better.
Having really low-distortion drivers and having them in a configuration that gives well-behaved polar response – not just having the direct sound from the speakers but also the off-axis sound and the total sound power coming into the room being of the same character – is a lot of what makes a speaker sound cohesive and have stable stereo imaging, and a smooth tonality. It’s nothing revolutionary, just good practice, but not everyone does that.
You’d think that as speakers get more expensive, everything would converge, but it’s just the opposite. And some approaches are radically different than others. I’m not saying that the approach that I’m doing is any more valid than anyone else’s, but at the same time I’m very carefully trying to optimize compromises. Every engineer has their own thoughts about where to put the money to achieve something.
But I’m sort of a transducer-first guy. It’s like painting with pure pigments – if you can start out with something that is inherently cleaner, better and has more performance, everything leads from there. So I nail the transducer performance first and take it from there.
I’m using a combination of planar magnetic drivers, which isn’t too common an approach, and some new low-distortion woofers. Igor Levitsky was vice president of engineering at Bohlender Graebener when I was there. Igor also designed planar drivers for HiVI, SLS (later Dolby) and others. He, and a guy named Dragoslav Colich, the chief technology officer of Audeze, are kind of the preeminent experts on planar magnetic drivers. There are only a couple of people in the world at their level and I picked up a ton of knowledge from Levitsky.
We also believe in giving customers a bit more adjustment capability for tailoring the tonality of their speakers with regards to the low frequency balance.
FD: Because not everyone has a room where they can set the speakers up optimally.
CB: You can’t break the laws of physics in dealing with things like early reflections, and there’s the so-called Allison Effect, where the reflections off the floor and front wall will cause interference with the direct sound from the speaker. So, where you place the speakers still dictates the notch in the [frequency] response from boundary interactions, but you can still correct the tonality of a speaker in some ways.
When designing speakers you have to make some assumptions regarding how near they will be to boundaries, both acoustically and physically. The sound from a typical speaker changes from being omnidirectional at low frequencies to being sort of hemispherical at higher frequencies, where the sound starts bouncing off the front baffle. At that transition point it’s an advantage to be able to make adjustments for typical distances from room boundaries, and we provide that kind of adjustment in our speakers. I’m surprised more speaker manufacturers don’t do this...having some kind of adjustment capability in this regard is a good thing.
But you want to give the right amount of adjustment capability to everyone without confusing them. Sometimes when you put a knob on something, people will think, “do I need this?”
We’re trying to do some things to let people know about setup. Some of the features of the FR-30 require a small amount of explanation, but we’re hoping it’ll be a net positive.
FD: How about the aesthetic design? It’s very different.
CB: Speakers are a big piece of furniture in peoples’ rooms and product design is a skill unto itself. Making something aesthetically pleasing and well-crafted isn’t necessarily the same thing as making a functional loudspeaker. Speakers are science in the service of art and music, but if a speaker won’t be accepted into the home, no one’s going to know how good a speaker it is! So, job number one is to make it through the front door! It’s challenging. It’s hard to come up with a different design language for a loudspeaker.
FD: Is there a reason why the front baffle is straight instead of angled back, like, say, a Thiel?
CB: I spent some time with Jim Thiel and the people there – a week in Lexington in horse country. They had very specific design goals about time alignment and phase coherence. But that created a whole set of nested compromises, that they were able to engineer out over time. In our case we’re using a fairly large 5-inch x 10-inch planar midrange driver which is highly dynamic with high output and sensitivity, but fairly directional at high frequencies because of its physical size versus the wavelengths that it’s playing. It’s a flat baffle because if we were to rake it back, it would be a detriment to the driver’s controlled vertical coverage. The listening axis would get more distance-sensitive, but with our design, the speaker is a lot more forgiving about how far away you’re sitting. The speaker becomes more room- and placement-friendly.
FD: Are you going to be doing the speaker adjustments via digital signal processing?
CB: At first we looked at DSP and room EQ but we actually moved away from that. Now we’re looking at potentially using FIR (fixed impulse response) and convolution filters, which corrects the time and frequency response of the speakers at the listening position. The problem with DSP is that if you use it on only some of the drivers, it has latency and that can cause timing issues of its own. To work successfully, DSP needs to be correcting all the drivers in the system. So we moved away from that for now and are doing a simple passive design.
FD: Vinyl lovers will be happy.
CB: Yes, they’re not going to have the vinyl playback go through A to D conversion. No digitizing your vinyl! But DSP is great for the process of speaker development. You can type in a bunch of “what ifs” and hear what they sound like as you’re playing around interactively. But once you settle on a design, it’s generally pretty easy to do [a circuit] in analog.
I like being a bit of a Luddite too. Sometimes it’s hard to beat doing things the old ways.
Probably the most controversial thing that I’ve done since I’ve been here: Arnie was a huge champion of servo-feedback, motional-feedback woofers. I’m not doing that. Sometimes when you go into developing a correction system for a driver, the money could be better spent just making a better drive unit.
FD: But you’re not opposed to servo correction per se.
CB: No. I’ve seen some fantastic stuff, but there are inherent issues in servo systems with noise at low signal levels (where it is audible) and their overload/clipping behavior. Our approach is to get the woofer as linear as possible. There are good and bad examples of every technology. The implementation is what’s important. The proof of the pudding is in the eating. I’m looking for a speaker that’s musically satisfying, correct and natural-sounding.