Rabbit holes

February 10, 2018
 by Paul McGowan

The problem with rabbit holes is they are easy to go deep before you realize it’s the wrong one. I went down several in the design of the amplifier line that eventually became the BHK, as well as building the first Power Plant two decades ago.

My first idea to design a perfect AC power generator was a carbon copy of what powers our cities. A spinning, mechanical, power generator. Mine would not be spun by the fires of coal or natural gas, but instead by the very power I wanted to replace. On paper, I built an electric motor coupled to a generator like you might have on your bicycle. From it would come perfect, clean, low distortion sine waves regardless of the power quality coming in.

That rabbit hole venture was quickly abandoned as Terri told me there’s no way she’d allow a spinning noisy mechanical generator anywhere near the inside of our home. I suspected others would feel the same way.

My second rabbit hole got me closer. A tube-based HP sinewave generator connected to a stereo power amplifier. Though kludgy and impractical, it worked and proved my idea that the best AC power in the world had to be generated. Nothing else would do. Certainly not a simple power conditioner, which seemed to me at the time about as useful as the polishing of a turd. Brighter and shiner but it still stinks.

Today we officially launch the DirectStream P20, our finest expression of the art of regenerating new, perfect AC.

Go to this page and enjoy.

The P20 has once again redefined the system’s performance in Music Room One, and not by a little.

Subscribe to Paul's Posts

24 comments on “Rabbit holes”

  1. Great job done, Paul. However there are two problems with contaminated mains current: 1.: Contamination via the ground connection. 2.: Contamination from the power supply modules in the stereo chain connected to the P20. Some manufacturers of power conditioners and mains filters recommended some 30 years ago to use an isolating transformer before the power bar and filters for each component connected to the power bar for avoiding cross contamination. Contamination’s caused by the units rectifiers or motor drives. I think your balun filter was some kind of isolating device? I really would like to know how a P10 or P 20 deals with these forms of contamination or if the pioneers of purifying mains power were wrong. The balun filter I tested some years ago (PS Audio Quintessence?) caused an irritating shift of the sound stage.

    1. Thanks, Paul. What we have discovered over the two decades of research and experience we’ve invested into the art of power delivery is that the most important aspects of improving sound quality are:

      1. Low output impedance
      2. Stiff dynamic regulation
      3. Waveform purity

      Not one of these qualities is helped by baluns and certainly not isolation transformers. In fact, they are made worse.

      Every time we lower output impedance by a noticeable amount sound quality takes a dramatic jump. Same with regulation. And while these may not sound tough to engineer, think again. Peak current demands of low power factor demand products like power amplifiers consume extraordinary amounts of peak current, often as much a 70 amps! No power conditioner or isolation transformer in the real world of home audio are capable of delivering anything close. This is because they are simple passive devices without any stored energy.

      The bigger the Power Plant we build, the greater its energy storage, the lower its output impedance, the greater its ability to fill in the gaps of peak energy demands by hungry power amps and even a few preamps.

      1. This brings up an idea for an ideal world. Perhaps you should consider an all DC audio system. Amplifer cuircuits all run on DC and SMPS have direct rectifier inputs, so a filtered DC input could deliver better sound quality than AC. It also has safety value – 50/60Hz is near the peak of the fibrillation potential curve.

        Most audio is still transformer input so the sine wave output is necessary today, but a separate DC output on a power filter device or a separate dedicated device for digital amplifiers makes sense. You have the DC internally so it is a logical but non-trivial extension of the circuitry.

        I also work in lighting where LEDs are causing lots of power and noise problems. LED bulbs all have internal power supplies to convert high Voltage AC to low Voltage DC. This gets really sticky when they have to accept phase control dimming. The results are typically almost no dimming until the bottom of the range and then shutting off, flickering and blinking; and an epidemic of premature bulb failures.

        There are also LED strip and stage lights which are dimmed by DC fed devices generating PWM at many Amperes from some kind of low voltage microprocessor control. All of this wreaks havoc with current waveforms and generates loads of EMI/RFI.

        Clean energy is also DC. My solar panels come off the roof at 480VDC, fuel cells and wind generators are DC, EVs are DC. Intermittent sources and portable use require battery storage so DC is the future.

      2. Paul,

        What happens to connected equipment when power to a P5 or P10 is disrupted, such as when the power company does regular maintenance and momentarily cuts the power? Is the connected equipment protected? Should a P10 be powered from a UPS; if so what guidance can you offer?

  2. Congratulations. Always good to launch a new product. Hopefully plenty of happy bunnies out there to buy it.
    It actually weighs more than my entire audio system, including speakers and stands, combined.

  3. Paul,

    In what way(s) is a P20 better than two P10’s? I understand that you formerly used at least two P10’s in Music Room 1. Have you replaced them with one P20?

  4. In the UK there does not seem to be the same interest in power conditioners, but we have a 240v AC supply which seems to have a better reputation for being clean than is the case in the US.

    As a postgrad the old house I lived in had a cheap and nasty consumer stereo in one of the communal rooms, donated by a resident who had moved on to a better system. It periodically would give a popping noise. We traced this to the cutting in and out of the motor on the fridge in the kitchen. When we wanted more acceptable sound someone would lend the stereo from his room. The popping sound invariably disappeared. This was because our home-built amps always included a substantial regulated power supply. It left me with the feeling that a good power supply should protect the amplifier from almost all mains variations except a sustained drop-out. Perhaps if I lived in the US I would feel differently.

    1. We do generally have a good clean power supply in the UK.
      However, the two main UK premium audio brands (Linn and Naim) have for a long time taken power supply very seriously, Linn having their own Dynamik system and Naim providing a range of external power supplies.
      Companies that have developed Class D or other hybrid amplification systems (like Linn Chakra and Devialet) appear consistent in viewing the power supply as fundamental to performance. Primare, which I used for many years, now have their APRC system. Devialet’s power supply is at the core of their system.
      All these developments have occurred in the last 10 years.
      For my part, Devialet told me regeneration is pointless and mine is plugged into the wall and is on the kitchen ring. Using my P3 made no difference.

      So there would seem to be two main considerations for regeneration:
      (a) Whether your AC supply merits one
      (b) Whether your system components would benefit from one

      As all such systems are often 90% efficient and designed to deliver instant peak power, a P5 or P10 should easily deal with the reset of the system. It would seem the P20 is primarily for huge conventional A and A/B amplification, which I assume is less popular in the UK and Europe than the USA, not least because most of the Class D and hybrid developments seem to come from Europe.


    2. If it were only clean power that concerned us this would be right. Unfortunately, it is n’t the cleanliness of the power, it’s the waveform distortion due to lopping off the sine wave peaks and the resistance in the wire of your home and what feeds it from the power pole. Those have the same problem in 230 volt countries as they do here in the States. We actually sell more 230 volt Power Plants than we do 120.

      1. In broad terms, in a system of PS Audio components, in what proportion is the perceived increase in performance from regenerated power split between its application to (a) amplification and (b) source components?

  5. Caribou Studio ran on a motor-generator set that was mounted on bedrock and buried deep in a mine shaft so you couldn’t hear the vibrations. Before they put it in, lightning was taking out tens of thousands of dollars of equipment during the late Spring, Summer and early Fall. They got one with a huge flywheel that covered the frequent gap when the power went down in the mountains so they could fire up and sync the battery powered inverter – this was before always-on hot switching.

    The batteries were just a longer term temporary solution – they could only run the studio complete with incandescent lighting for five minutes. In the interim, diesel generators were started and synced to switch over. None of this required the staff to stop what they were doing or interrupt an on-going session.

    We got a tour of the facilities including the switchgear when we did a house gig on Jimmy Guercio’s turf, connecting the mobile studio to one of his rental houses to record a Country Rock demo tape out of their living room/practice hall. We had two transformers, one to run the HVAC and one for the audio, both with 240V input that we plugged into an electric clothes dryer outlet. The audio side was a high isolation shielded model with 20 Joule MOVs and 5uF, 2KV oil filled caps on the input side forming an LC/RC filter with the 100 foot 10AWG input cable.

    The transformers also had tapped inputs to adjust the output in 5% increments and compensate for over and under-voltage.

    The Ampex MM-1200 24 track deck was notorious for blowing the power supply over transients – the module was normally replaced at least once a year – but this one survived 50,000 miles of roadwork and stages with hundreds of KiloWatts of triac light dimmers and the noisy power lines of industrial districts surrounding arenas and stadia.

    If the ESL/ESR of the power connection caused audible artifacts, we were not aware of it. The main mission was delivering the electronics to the gig in working order, because a broken amp or tape machine sounds worse than one that is working.

  6. I didn’t fall into a rabbit hole, I entered a parallel universe and found out that I was entirely alone in it. But that’s another story.

    I have probably spend more of other people’s money on specifying, buying, and installing power quality improvement products than most of the people here combined. Many millions of dollars. I’ve seen all kinds including motor alternators and motor generators. Today the standard for industrial use is the double static conversion UPS. I’d never seen a power regenerator except for Paul’s. It’s an interesting idea and it has its advantages and disadvantages. As I understand it, it is a sine wave generator fed to a class AB audio amplifier whose output is adjusted to 120 volts (230 volts 50 hz for Europe) probably using a step up voltage transformer to supply power to audio equipment. The incoming power is rectified and filtered to become the power supply for the class AB amplifier. The input is therefore the equivalent of the first part of a double static conversion UPS. By adding batteries and a battery charger to the DC supply the regenerator could be converted into a UPS. Claimed output harmonic distortion is 0.5 % or less. That’s about as good as it gets in my experience. The drawback is poor efficiency. The class AB power amplifier will be around 40% efficient whereas a digital switching inverter will be at least 90% efficient and there is no practical limit to the size or number that can be paralleled. These can go into the megawatt range for large data centers. The practical limit for the regenerator may be around 4 or 5 kilowatts if that. For industrial users price per kilowatt is also a consideration and the best manufacturers are very competitive.

    Congratulations Paul on your new product. If you turn it into a UPS it might find industrial application with the most sensitive scientific laboratory equipment although its 2 KW rating can be a problem. Or it could simply be connected to the output of a standard UPS to produce a sine wave with lower harmonic distortion.

  7. Paul…

    I assume that was designed with high current power demands involved? What about us guys who listen with audiophile energy efficient D amps, while listening on audiophile desk top systems? I am talking high quality everything, but the need for high power demands has been greatly diminished compared to what the DirectStream P20 provides for.

    1. No, actually, it was designed with high performance audio in mind and anything connected to it will sound and perform better than anything else we have ever produced. I tried a pair of M700 monoblocks on the P20 and was stunned at the level of analog sound I got. It really helped a great deal. The P20 is powerful but that’s a side benefit as far as I am concerned.

      1. Could it be down scaled for a smaller audiophile system?… say needing only about six outlets? (two being needed for a PC and monitor). The way it now is like plowing snow in 20 foot driveway with a bulldozer.

        I am sure it would sound great, but very impracticable for the needs of many.

        1. You are making exactly the same point as I made above – the premium quality D amps I know of all have custom and often patented SM power supplies, add Mola Mola to the list – it’s a large part of why they work so well, and probably why they are quite expensive. There is no mention of that being the case in the M700 amplifiers. I followed my manufacturer’s advice on the use of regeneration. I moved from a Linn streamer to a PSA DAC and recall that Linn dealers, even one selling PSA, did not recommend using regenerators with Linn amps. I still bought one (a P3) for everything else. I also agree with you that if the P20 is better than the P3, P5 or P10 for reasons other than output, perhaps the smaller units should be superseded with P20 tech.

      1. Paul, thank you for your prompt reply. I have a follow up question. Would the P20 provide any further sonic advantages to justify purchasing it versus an updated P5 or P10 assuming that the system’s power requirements can be easily met by a P5 or P10? I should also mention that I like the looks of the P20 including the look of meters on the display.

        1. Yes, and for the same reasons a P10 is always preferable to a P5 regardless of power requirements.

          The size of the Power Plant improves performance because as each gets bigger we are able to add greater numbers of output transistors and energy storage caps. The more of those we have the lower the output impedance, the greater the reserve peak power capacity, the better the sound, for any load.

Leave a Reply

Stop by for a tour:
Mon-Fri, 8:30am-5pm MST

4865 Sterling Dr.
Boulder, CO 80301

Join the hi-fi family

Stop by for a tour:
4865 Sterling Dr.
Boulder, CO 80301

Join the hi-fi family

linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram