Impedance vs. regulation

January 25, 2023
 by Paul McGowan

In yesterday’s post, I wrote about the importance of lower impedance in the AC line voltage. I mentioned that lowering the impedance is far more important than cleaning up the AC line noise (a subject for yet another post as this goes against nearly all the conventional wisdom).

I got a lot of mail asking me to explain how regulating the voltage could result in lowering the impedance. They seem like very different processes.

Let’s give a quick look at what it means to have low impedance. It’s probably helpful to use an audio example first.

Let’s start with the output of a power amplifier. Years ago, in the days of output transformer-coupled tube amplifiers, output impedance mattered greatly. Looking at the back of one of those amps, you’d notice different speaker connections: 8Ω and 4Ω. These were necessary to match the relatively high impedance of the amplifier to the speaker. Put a 4Ω speaker on the 8Ω tap and you’re going to lose voltage/loudness.

This mismatch is because of our old friend Ohm and his law: a high impedance feeding a low impedance results in a loss of voltage. A low impedance feeding a high impedance offers little to no loss.

Ergo: we always want a low impedance feeding a high impedance.

Back to the AC line.

Suppose a power amp draws enough current to play a big loud bass note. In that case, the inherent impedance in the hundreds of feet of wire connecting your home’s AC socket to the utility pole outside your house is more than enough to result in a significant voltage drop (sometimes 3 to 4 volts or more). This modulation of the voltage is sonically unwanted in an amplifier.

If, on the other hand, an AC regenerator like a Power Plant is inserted after the AC socket, its ability to tightly regulate the voltage means there will be no voltage drop—as if there were only inches of wire as opposed to hundreds of feet.

Hundreds of feet of wire = high impedance.

Inches of wire = low impedance.

Hope that helps.

Subscribe to Paul's Posts

26 comments on “Impedance vs. regulation”

  1. Definitely helps. 🙂

    My P3 Stellar Powerplant turned 3-years old this month and is still going strong.

    Does a lot for my system. Extremely grateful.

    1. Mine turned two last fall. I traded a pretty old P500 for it, and it’s a definite improvement. A perfect fit for my mostly-Stellar system (SGCD, M700s).

  2. Thanks Paul! Couldn’t this be better done near the power supply entry point and not at the stereo?

    But I guess it would have to handle a too big load for the whole house then.

    But how about putting a regenerator into the circuit for just the dedicated line to the stereo…but still in the basement near the circuit breaker…we could then use whatever sophisticated distribution system instead of using the power plant as distributor and we could save effort for shiny housings for the power plant etc. Probably nuts, but did anyone think about it from a non marketing perspective?

    1. Good idea but not very practical and probably not as good. Consider that if it were remotely located you again add many feet of cabling between your stereo and the source of power. Plus, now you have introduced an even worse problem, multiple audio products on (essentially) the same power cord. You would have to have single point star wiring (one power cable going back to the source for each product) and still you’d have the problem of long cables.

      Don’t forget, power cables between the equipment and the regenerator matter.

      1. Surely you need appropriate mains cable, irrespective at what end you are regulating the power. If you are running a big system off two P20’s, as is sometimes the case, how much current are they drawing? You will need thick enough cable from the consumer unit to the receptacle to avoid a fire hazard.

        The point of the safety regulations we have is that if you start overheating cable it will short the consumer unit before it causes any damage, which I recall is called the earth loop impedance. All sockets have to be tested for a maximum value.

        People seem to use 10 AWG because it is about the largest gauge you can wire into a domestic power receptacle. I was recommended and used this.
        It is rated at 27A. I calculate my voltage drop figures from here:

      2. I live in Australia. I have a whole of house regenerator at the switch board. It was installed to ensure that the voltage to my house regardless of supply was 230volts. Previous to the regenerator, the supply voltage to the house often went over the solar maximum operational level of 253 volts. Problem solved. There were other reasons for the regenerator. The regenerator provides enough power to run a bore, dish washer, audio system and other devices all simultaneously without any issues.

        From the switch board to my system is a separate dedicated high quality cable circuit. The voltage to the audio is a constant 230 volts. Whereas the rest of the house there is an approx. 4volt drop. At my system I have a 12 socket Gigawatt conditioner. All system components and accessories are connected to the dedicated circuit. The major system components are connected to the conditioner. In my case it is the cables from the circuit outlets to the system that matter.

        I cannot say that the house regenerator is a benefit or not SQ wise. It is a beneficial investment though.

        1. assisi,
          Happy ‘Straya Day mate!
          A bore eh?
          So you’re out in the country?
          In the big smoke (Sydney), since 1968 anyway, I’ve never
          had an issue with the mains voltage going so far over the
          240v mark that it’s ever fried anything in my various home-
          audio rigs.
          Still, if your voltage spikes beyond 253 then better to be
          safe than sorry.

  3. I used a regenerator with one of those valve amplifiers with 8 ohm and 4 ohm connections on the back and it seemed to improve sound quality. As the amplifier manufacturer once said to me, it’s 1930s technology.

    What seems to be ignored is that impedance is a safety issue. In the UK and EU it has to be measured for safety certification (to every electrical output and appliance on the system) and hence the cabling to each output has to be evaluated for impedance. My 32A rated induction hob has a much thicker cable than the rest of the electrical circuits.

    If you run a massively powerful audio system I would expect to do these calculations and install appropriate wiring. In my system, from the wall socket I have about 4m of 2.5mm 3-core which drops 15mV/A/m and then about 10mm cable, 5m from the consumer unit to the junction box and 4m to the street cable, that drops 3.8mV/A/m. In total that’s a voltage drop of about 0.1V per Amp of current draw. As my audio system draws perhaps 0.5A and we have 230V power, the voltage drop caused by the audio system could be a maximum of about 0.05V.

    If as Paul suggests I had 3-4V drop (70 times what I calculate for my system) I would be more worried about the fire risk than the sound quality.

    To get 3-4V drop, you really must live in a forest with inappropriate cabling and a very power-hungry audio system.

    With my modern Class A/D hybrid power and sensitive digital streaming products, powerline internally generated noise is the primary issue for sound quality and what I spend effort defeating.

  4. Paul is this the reason that some amps like the Dynaco ST 400 series could be improved by adding external capacitor banks? If yes why do we not see more of those banks today?

    1. External capacitor banks would require a soldering iron, cutters, and some skill. You cannot just add caps to the input as caps have to have DC to be effective.

      Funny enough, tomorrow’s post refers to this exact question. Stay tuned.

  5. It’s not just what happens inside your house that affects the voltage, outside influences also need to be considered, especially other users. My mains supply measures around 240volts, a generally accepted figure in the U.K. but the specified figure for a power supply company is 230v with a tolerance of -6% to +10% giving a range of 216.2v – 253v. This is another issue the Power Plant is designed to overcome.

    Having said that and previously being a firm believer in regeneration I’ve recently tried some other power products. I think it’s really important to discover what problem it is you’re trying to fix, if indeed there is a problem, before parting with your hard earned. As well as regeneration I’ve recently tried filters and just DC blocking with very effective results. It all depends on your individual situation, there is no one size fits all. That £10k might be better spent on a speaker upgrade?

  6. Is it right then to say if one regulates the power amplifier’s power supply rails the effect is similar or perhaps better as the regulation is even nearer to the amplifier vs regulation at the AC inputs which still have to pass through power cord and transformer windings?

    Of course, the advantage of AC regulation is it feeds clean power with low impedance to all equipment plugged to it.

    Regulation at the equipment side means having to regulate each equipment individually which is more expensive ?

    1. It is rare that a traditional power amp has a regulated supply. The vast majority are unregulated. This is because it would take almost as much “stuff” inside the power amp to regulate as it does to amplify. (Unless you go with a modern SMPS).

      Still, it matters how low the source impedance feeding the unit is. Take for example, preamps and sources. These almost always have fully regulated power supplies and yet they too benefit from a Power Plant feeding them low distortion regulated power.

      1. Thank you for your reply. Indeed it is surprising to know that preamps and other lower power equipment with regulated supplies also benefit from a regenerative AC source with low impedance.

      2. Correct. Naim is a major hifi player and they have been making products with optional regulated supplies for 40 years. They do have as much metal as the amplifier itself and broadly similar price tags. They are less demanding, but have very high peak power (typically 5x average power rating), so they have tremendous dynamics. Due to their popularity, there is a separate source of clone power supplies as it gets expensive and the box-count can be staggering.

        All these power issues makes SMPS seem more and more sensible, but then I went down that route 13 years ago.

  7. One other thing I forgot to mention. I always understood that for maximum power transfer impedance’s should be matched, not either side higher than the other. Certainly true for transmission lines. Not less true but maybe less critical when applied to Power Plants?

  8. When I was doing research for my audio room, I found a LOT of arguments over whether it was better to reduce the AC impedance by going with a 20A breaker, 10AWG wire, quality outlet, god cabling and plugging directly into the wall vs. using a regenerator. The idea, as you describe, is that with less-impeded AC, a big bass note followed by another big bass note will see faster and more efficient amplifier “recovery.” So unless your regenerator has a bypass/filter-only mode, you’re better off plugging straight into the wall. The P3 PowerPlant does have this mode, but it would still pass along any fluctuations in voltage caused by outside effects, right? I did the line-side improvements and the recovery seems to be better, but would the overall effect of regulation in the regenerated mode be bigger than the unimpeded AC? I’m thinking that might be why there’s a bypass SWITCH – to compare? Duh?

  9. It seems like you can regulate an output and that might effectively reduce its output impedance dynamically (if it can regulate fast enough–probably not an issue), but that shouldn’t be considered the same thing as having an intrinsically low impedance.

  10. My Sanders Sound Systems Magtech amplifiers have a regulated high voltage bus which allows for higher output power from the same 120V/20A receptacle. These things sound great to my ears.

    I do have an SPP3 for all source gear and even my SSS ‘stats power supply input for HV panel bias. This power plant is also excellent at protecting electronics during electrical storms and brown out/black out events. Very pleased with the performance.

  11. Low impedance is magic stuff, for getting good SQ. The frustrating bit is that the way physics works, the higher the frequency, or transient nature of the current draw is, the harder it is to keep it low – which is why glitches are a part of life in electronics … the goal is to make the glitch as absolutely as small as possible.

    A big Perreaux amp I started my digital years with had one major flaw – the power supply caps were a poor choice by the designer, and gave me ongoing SQ issues … I ended up ripping them out entirely, and completely re-engineering how that area was done. Effectively, the voltage rails were now much, much lower impedance at high frequencies – and subjectively the SQ was in much better shape. External capacitor banks are no answer; the path is now too long, and inductance will kill the benefit, for most cases.

  12. I don’t have a power plant or a line conditioner. I cleaned the contacts on the wall wire and installed a new wall socket. My Creek integrated amplifier comes with a heavy gauge 3 prong power cord. My AC is not a problem. No noise

Leave a Reply

© 2023 PS Audio, Inc.

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