Tag Archives: Threshold

Asheville, Walnut Cove, Biltmore Forrest and Western North Carolina’s Audio and Home Theater specialists present Cane Creek AV and Paul McGowan – PS Audio, Intl.

Last but not least

Following the past few day’s posts about audio amplifier efficiency, Class AB biasing, and Class A biasing, let’s wrap our little mini-series up with another topology most of us have not heard of, adaptive biasing.

The promise of adaptive biasing is a best of both worlds scenario: the efficiency of a Class AB circuit with the performance of a Class A amp. Sounds too good to be true, right?

The first time I ever heard about an adaptive biasing scheme was way back in the dark ages, the late 1970s. My very dear friend and one of the all-time good guys of audio, Nelson Pass, then of Threshold Corporation, had introduced the idea of what we called a sliding bias scheme, part of what later became known as the Stasis Circuit used in Nakamichi, Threshold and if memory serves correctly, even the Mark Levinson No. 33.

The core of this circuitry is covered in Nelson’s patent from 1975 titled Active bias circuit for operating push-pull amplifiers in class A mode.

Simply put, Nelson’s design raises the level of Class A bias in cadence with the rising input signal.

Recall in our discussion of Class AB design that a small amount of always-on power keeps small signals always on. In other words, we apply Class A (always-on) bias to the first 10% of the amplifiers output signal level, then switch over to the more efficient Class B for higher signals. Compare that to Class A operation which is always-on for any level of signal—always generating a shit-ton worth of heat (recall Class A amps are at their coolest when at full signal out).

What Nelson cleverly suggested was this. Take what we do in an AB amp where the first 10% of the signal is Class A and actively monitor the signal level. When any given input signal starts to exceed our 10% Class A bias, raise the limit from 10% to, say, 20% (or whatever is greater than the signal level), and continue on the path all the way up to 100%. Then back down again tracking the signal. The heat-producing bias is only enough to accommodate the signal, then goes away when it’s not needed.

Thus, we get the benefits of both worlds. Efficient, and sweet-sounding.

Why doesn’t every amp use this even today? (Nelson’s patents ran their course years ago). Well, as with any design there are problems as well as advantages and this post is long enough already. Ain’t nuthin’ perfect. (We used this for several models of amps though their model numbers and dates escape me).

In any case, a juicy piece of history I’d thought I would share.

Asheville, Walnut Cove, Biltmore Forrest and Western North Carolina’s Audio and Home Theater specialists present Cane Creek AV and Paul McGowan – PS Audio, Intl.

More from Paul on loudspeakers.

Mission improbable

Yesterday I mentioned the reason all loudspeakers are colored and veiled. They have mass to move.

We don’t know how to move mass without affect—no matter how thin or light it is.

There have been several attempts at building massless speakers, like the Hill Plamatronics, though it included a conventional, heavy mass woofer.

One of my readers reminded me of our friend, Nelson Pass (Threshold and Pass Labs), and his attempt at a full range massless loudspeaker. Check out Nelson in a suit!

I don’t know the history of this. But it certainly never made it to store shelves and, even if it had, anything generating sound with ions also makes ozone, a gas not good to breathe in excess.

Which begs the question, if there was a massless speaker would it be without veils and added color? The answer would be no. The air itself has mass, and whatever means used to move that mass will impart a sonic thumbprint.

Colorations are part of the listening experience.

The trick is to find the ones most pleasing to you.

Climbing the ladder

Back in the dawn of time for digital audio, the 1980′s through the 1990′s, there was basically only one type of DAC architecture in popular use: the ladder DAC.  As well, the mainstay of the ADC (the recording end of things) was also a similar arrangement.

These were the classic PCM based systems we all got involved in digital audio with.  They could handle the 44.1kHz/16 bit requirements of the CD with appropriate accuracy and life was good.  The sound wasn’t so good, but life was.

It didn’t take long before the quest for better, faster and more bits started to rear its head.  One of the first indications of this came from a company called Ultra Analog out of the Bay Area.  Ultra Analog was a very technologically advanced company and they had a better, more accurate version of the classic ladder DAC.  Many of you may remember that many, many high end DACS of the day started using their proprietary DAC modules including: Audio Research, Sonic Frontiers, Mark Levinson, Threshold and yes, PS Audio.  The main selling point of this DAC is it featured true 20 bit performance.  This was no easy task given the architecture of the day.  Here’s a photo of an Ultra Analog module.

Ultra analog 300x225 Climbing the ladder

So popular was this product that PS Audio based all our DAC products around the modules (producing the Ultralink DAC) and eventually we even bought the company, Ultra Analog, so popular were these models.  That was a long time ago.

That DAC technology depended on having extremely accurate resistors.  In fact, so accurate, that Ultra Analog’s secret wasn’t in using resistor’s that were more accurate than anyone else’s (they couldn’t because we’d reached the theoretical limit on accuracy).  Instead, clever engineering and a bag full of tricks did the job.  But that job of increasing the number of bits was at its very limit – one pushed to the extreme by this very DAC module, used in so many fine DACS of the day.

Yet today 24 and even 32 bit DACS are everywhere.  These everyday DACS have performance so much better than the 20-bit Ultra Link module that the engineers of that period would have cried “impossible”!  No one seems to have noticed just how much has changed since the days of Ultra Analog’s miracle 20 bit wonder – changes we seem to take for granted.  Using the classic ladder DAC architecture, each added bit is exponentially more difficult to achieve than the last.

What changed?  Did we solve the problem of the ladder DAC or did something else happen?


Paul McGowan – PS Audio Intl.