The Chief Engineer and founding partner of Luminance Audio proves to be more than just a formidable engineering talent; he is also quite prolific in literary creativity if asked to do that. Including this article, DAGOGO has published five articles thus far by Steve Keiser. While three of the five articles were on his idea of a future preamplification design, this latest article promises to increase the heartbeat of many readers: the creation of a digital monoblock amplifier by Luminance Audio!
DAGOGO is grateful for the privilege to serve as a platform in presenting Steve Keiser’s latest innovative thinking.
The basic design premise supporting the conceptual idea behind the digital is developing a system design approach behind Pulse Code Modulation and implementing all the related functional circuitry around discreet componentry, rather than utilizing integrated circuit technology. My approach of using all discreet componentry allows for greater design flexibility and higher performance, albeit at higher component count and cost. It just isn’t possible to “go in” and modify an integrated circuit as it is constructed originally as a self-contained entity.
In this particular application of Pulse Coder modulation, the input signal is sampled and converted to a pulse-width signal. In this case, the area under the rectangle pulse represents a given moment in time in which the width of the rectangle is determined by the input signal amplitude. This pulse code signal modulates an output stage current, in which the pulse-width duty cycle (the time the signal isn’t at zero) as generated by the output power devices produces a train of switched on and off rectangular pulses, which corresponds to the sampled input signal. This series of rectangular pulses is then filtered as the resultant output signal constitutes a smoothed but amplified replica of the input.
Where my digital amplifier differs from other Pulse Code Modulation schemes is that there is no negative feedback employed in the input comparator circuit responsible for converting the input signal into a series of rectangular pulses (either on or off) with the width of each pulse corresponding to the amplitude of the input signal. The consequence of this lack of feedback is the potential for greater error for the rectangular pulse to not exactly match and represent the input signal level precisely.
This potential error is compensated for two ways: First, very high quality precision value components are employed as part of the sample and hold circuitry. These high quality components could not be accommodated in the form of an integrated circuit and that is why I chose an all discreet component format. Additionally, I have developed a simpler output filter network when compared with other PCM designs, thus dramatically reducing the individual filter components from interceding with the musical waveform and any associated sonic colorations and artifacts.
The resultant overall design features very high power (400 watts into 8Ω and 800 watts into 4Ω, very high current delivery into the load (30 amperes peak to peak) and high efficiency (90 percent). Another innovative nuance of this design is the output devices operate with a slight bias current so to act as a form of dither, which effectively improves the low-level signal resolving capability much in the same way dither is used in compact disc technology to “linearize” very low signal levels. I would not represent the digital circuit topology as possessing any significant advantages over an equivalent analog counterpart, other than having the potential advantage of being able to generate higher output power without the limitations imposed on analog circuits when higher power levels are strived for.
In analog amplifiers, for instance, as output power increases, the number of output devices also increases, thus generating a more taxing and distortion-producing load on the driver stage. Also, the higher power and current that analog designs must rely on corresponds to higher power and current ratings in the output devices themselves, which almost inevitably have slower transit time and speed. The accuracy of an analog amplifier is primarily determined by the circuit designers’ ability to “linearize” the internal gain components, while a PCM amplifier’s accuracy is determined by the precision by which the pulse-width is determined and maintained throughout the system. The final arbitrator in determining the potential desirability of PCM amplification can only be determined by the critical sonic assessment in reproducing music when connected to a real world application circumstance. It is here where I will allow the eminently discerning critical aural capabilities of the reviewer to expound upon this most significant parameter.
I have used “PCM” and “digital” interchangeably. Both are correct. It is theoretically possible to take a conventional digital data stream and allow it to operate in the creation of PCM without ever going through a digital-to-analog conversion process, thus allowing a digital signal to be in digital format until the very last stage of the amplifier. That concludes my outline. Again please consider my offer to update the component quality in the comparator circuit of your amplifier since I have an intuitive hunch this might produce a significantly superior amplifier.Also the monoblocks youy posses can ce converted from a basic KST-150 stereo unit since I use the pc boards of the KST-150 to incorporate all of the funtional circuitry.
- (Page 1 of 1)