This article was written in 2014 in response to what can only be described as a vitriolic debate as to what constituted a CFA on diyAudio.com. At the time, many of the participants were struggling to grasp the fundamental differences between VFA’s and CFA’s, and even how to tell them apart. So, this document is not filled with math equations (others have done it already, and done a better job of it than I could ever do) but instead focuses on how they perform differently and how they stack-up against each other in the context of audio amplification. Importantly, it goes on to show that once the designer elects to go for high open loop gains in a CFA design, the resulting behaviour of the CFA morphs into that of a high loop gain VFA, and, whatever topology is used, the output stage is the ultimate arbiter of bandwidth in any practical amplifier.
You can download the article here:
An Introduction to CFA Amplifiers for Audio
(Minor update made in May 2019 to clarify operation of H-Bridge topology input amplifiers – my thanks to Scott Wurcer for pointing out my error)
CFA topology amplifiers have been around in the IC industry for 30 years. Following a patent claim by inventor David Nelson, the earliest commercial offering was a module from Comlinear in 1982 and a few years later, IC’s from both Comlinear and Elantec. Prior to this, they were also described and analyzed in a number of papers. With regard to discrete based audio amplifiers, the topology has been used by a few esoteric brands in audio, with Accuphase, a Japanese company based in Yokohama, being a notable exponent. Cyrus, a small UK company, has also marketed CFA based power amplifiers. There are examples of Pioneer amplifiers from the early 80’s that used CFA, which apparently even pre-date the IC offerings and Mark Alexander published a design as an ADI application note in the 1980’s. CFA topology audio amplifiers continue to be somewhat upstaged by their more widely understood and deployed VFA counterparts – a situation not helped by the fact that neither Cordell nor Self touched the subject in their otherwise wide ranging audio design books. A CFA’s operation is not as intuitive as a VFA and there are some subtleties in regard to whether a transimpedance (TIS) or transadmitance (TAS) second stage is used and compnensation design in general, so designers preferred to go with something that is generally more widely documented and traditional – i.e. VFA. Although CFA audio power amplifiers have been available commercially for over 35 years (as of 2014), there is still a lot of misinformation out in the audio industry and DIY community about CFA’s, with some notable commentators dismissing them altogether. This is a pity, since they do bring very specific properties to the table that are of benefit in audio power amplifiers.
Comments