This review page is supported in part by the sponsors whose ad banners are displayed below


EK: The EL series amplifiers are housed in very attractive and well-finished if rather small chassis. In light of the generous power rating, what design methodologies (in terms of circuit board layout etc) were applied to maintain the diminutive size?


SR: The EL series amps were actually designed from the outside in. Years ago I was in the warehouse of a local aluminium profile supplier. While there I saw, sitting at the back of a shelf, an off-cut of a particular heatsink profile that really caught my eye. It was one of those light bulb in the head moments and within a few days I had designed the concept of the EL30 chassis using 3D CAD. It wasn't long before I had purchased all the stock of that profile from the local supplier, then all the stock in Australia and then another 1.2 tons of it was bought in especially for me from overseas.


The EL series are standard equipment width (430mm) but take away half the width occupied by the heatsinks and there isn't much room left. Fitting in all the electronics including the power supply circuitry was a real challenge! Even more difficult was creating a design that could be replicated efficiently and consistently in a production environment.


First problem, the power supply. Transformers are noisy and radiate nasty electromagnetic fields which easily find their way into audio circuitry. Many manufacturers avoid this issue by either keeping the power supply in a totally separate chassis or by physically locating the power transformer as far away as possible from the sensitive amplifier circuitry. Because of the limited space in the EL chassis, I had no choice but to mount the amplifier circuitry directly on top of the transformer. I was able to do this by isolating the amp circuitry from the transformer radiation with a 4mm solid steel barrier plate. This essentially creates two sections in the chassis, basically the "clean box dirty box" technique. In addition to this steel barrier, the actual transformers used are custom designed to be very low noise and incorporate a magnetic shielding strap which dramatically reduces radiation in the first place. The transformers were also carefully designed to minimize mechanical hum and when I say minimize, I just about mean 'eliminate'. How annoying is a hifi system where transformer hum is louder than the noise floor through the speakers!


With the power supply now worked out and isolated away in its own little area, I needed to somehow get the 'power' through the steel barrier and up to the amplifier section. Now I'll be honest here, I just don't like wire looms. They're tedious to make and slow down production. They're also ugly, the connectors can be current limiting, they're a common point of degradation over time and they can be inconsistent between batches. Did I mention they're ugly? Okay, now that I have that off my chest, obviously there are no wire looms in the EL series. Instead the power rails are bought up through the steel barrier via gold plated 8mm diameter solid copper buss bars.


The rest of the design came down to the main amplifier PCB, again designed a lot from the outside in. The position of the output devices was governed by where they bolt onto the heatsinks and with the buss bar connections now set, I just had to join the dots. Much easier said than done! It's one thing to have a good amplifier design on paper (schematic), but a good PCB design is a major hurdle especially in this case without a lot of space to achieve it. With the position, orientation and proximity of each component and PCB track being extremely critical, the pursuit for ultimate performance required a multi-layer PCB design to implement the much needed field cancellation topology (FCT).


Lifting the lid on an EL series amp will reveal that the main amplifier PCB is a complete mirror image about the center line. This was one of my initial design criteria. Often a stereo amp will be constructed using the same amplifier PCB for each channel rather than mirror image. The problem with this is that the PCB layout can only be optimized for one side of the chassis, which typically means one channel only. So you get an imbalance in performance between channels. This imbalance can range from significant to almost immeasurable but nevertheless it was something that I wanted to avoid completely.

EK: Ok, so can you then expand on Field Cancellation Topology (FCT)?


SR: A wire or PCB track with an electric current passing through it generates a magnetic field that is proportional to the amount of current flowing. In a power amplifier, considerable current can flow in certain tracks, producing considerable magnetic fields. The opposite can also occur - a wire or PCB track in the presence of a magnetic field can produce a flow of current. The last thing we want are the fields generated by high current tracks to be picked up or induced into any areas of the circuit such as the sensitive input stage.


Luckily the fields generated by these high current carrying tracks are totally predictable and with careful track layout can be cancelled out. All current must flow from the power supply but it must also return there. Field Cancellation Topology uses the field generated by the return current to cancel out the forward current field. As the forward and return fields are generated in opposite direction, the net result is no (or minimal) field generation.


EK: Can you tell us more about your decision to use ThermalTrak output devices?

SR: In a typical solid-state amplifier, a certain amount of current is intentionally run through the output devices in order to reduce distortion. This current is known as 'bias current'. In general, the more bias the lower the distortion. The problem is that the current running through the output devices heats them up and as they get hotter the bias current has a tendency to increase. This increase in bias leads to more heat which leads to more bias which leads to more heat and then of course more bias. Next thing, bang, the output devices explode because they reach their current limit and burn out. So to cool the devices down we need heatsinks to dissipate the heat. But this isn't enough. We also need a control mechanism to regulate this bias current. As the devices get hotter we need to sense this and reduce the bias in order to keep it constant.


Eventually, this heat sinking and sensing arrangement will reach a thermal equilibrium and the amplifier will hopefully remain stable. The issue arises how to sense this bias increase and keep it under control. The most common method is to use a sensing device bolted to the heatsink which changes properties with temperature just as the output devices do. But with this arrangement exists something known as 'thermal lag time'. The output devices all need to transfer their heat to the heatsink, which then takes time to heat up before the sensing device can compensate for the bias increase. This can take anywhere from 10 minutes to an hour or so. ThermalTrak transistors contain a sensing diode inside the actual output device itself. So lag time is eliminated and thermal stability is essentially perfect because as the output devices heat up the sensing diode instantly heats up also.


EK: What other important design philosophies we reapplied to the Elite series amplifiers?


SR: I'm very much of the "straight wire with gain" mentality. I believe it's important to have clear design goals and objectives and I do rely heavily on measurements to achieve them. I've often said that good sound is not so much about putting the good stuff in, it’s about taking the bad stuff out. I could build the best amplifier circuit using the best available components and it would buzz, hum and probably oscillate without careful layout consideration. Implementation is very important.

EK: Tell us a little about your background.


SR:My audio understanding was built up from years of reading, experimenting, prototyping, measuring and listening. I have also been very lucky to have direct contact with one of the world’s foremost loudspeaker engineers who has proved to be a wealth of information and when things just don't seem to make sense he always has the answers. He has taught me more than any book could ever have.


I have a degree in digital systems, which is a combination of electronics, computers and robotics. I spent 9 years in the R&D department of a local electronics manufacturer working on all elements of product development including 3D CAD, schematic & PCB, software and firmware. I wrote over 1 million lines of assembly-based firmware code. I’ve also worked on contract for various local and international high-tech design companies including large scale automotive where I worked on over 60 safety critical PCB designs.

EK: How was SGR born?


SR: Audio started as a hobby and grew into a business from a shared passion to build and achieve something better than average. We started from very humble beginnings, working out of a garage with no more than a few hand tools. Today we have a large factory equipped with multiple CNC machines and state of the art facilities. My father Harry’s audio background spans some 30 odd years now, detailing elements of importing and retail sales as well as product development. I’d call him hard to please but that's a good thing because he has a very good ear and when he is pleased with the sound then we know we're on to a winner. I also need to credit my girlfriend Belinda who handles the accounts and admin side of the business and as a quite awesome web developer designed our SGR website; and a computer and software guru who always just 'knows'. His assistance behind the scenes has been essential to product development so to most around here he is known elusively as 'The Oracle'.
Enlarge!