It looks like audiophiles go as far as installing separate powerlines to power their high quality audio systems. The motivation is that the amplifier and other hardware would run cleaner when the powersource provides cleaner waveform and with a separate powerline the amplifier will not be affected by various electrical devices run by other households.

Let's not assess whether any audible difference can exist because that's highly subjective. I'm interested in the purest waveform of the powersource part.

That separate powerline

  • can be hit by lightning
  • is also parallel to many other wires around it and so can be affected by processes in those wires
  • is connected to a large grid which has tons of transient processes happening all the times

so it doesn't really sound like a reliable best purest waveform source.

Would a motor-generator set run on the original powerline be better? It looks like it doesn't care much about all those lightnings, parallel wires and the grid.

How good is a motor-generator set for generating the purest waveform electricity?

  • 1
    \$\begingroup\$ What about batteries and a Full Sine wave inverter? \$\endgroup\$ – Solar Mike Aug 29 '18 at 11:49
  • 2
    \$\begingroup\$ @Andyaka: with acdc, there is always noise... \$\endgroup\$ – PlasmaHH Aug 29 '18 at 11:52
  • 2
    \$\begingroup\$ @Plasma AC/DC always made a lot of noise, good though :) \$\endgroup\$ – Solar Mike Aug 29 '18 at 11:57
  • 2
    \$\begingroup\$ What you really need is battery power for the audio system. You can charge the batteries during hours that the system is not in use. Better yet, build the system with two sets of batteries and use one set while the other is charging. Install the battery system remotely from the audio system and run the DC power lines in a shielded cable to the audio room. \$\endgroup\$ – Charles Cowie Aug 29 '18 at 12:08
  • 3
    \$\begingroup\$ Note that you used the word "reasonable" in the question title and the word "audiophile" in the first sentence. Those are mutually exclusive terms. \$\endgroup\$ – Charles Cowie Aug 29 '18 at 12:35

The "purest waveform" you want for noise-sensitive equipment is flat DC.

The both simplest and best way to get it is a battery. Why?

Any kind of AC needs rectification of some kind. Even a motor-generator with a standard DC generator produces a rectified sinusoidal current. You could use an unipolar DC generator, as it was done for ultra-low-voltage ultra-high-current DC applications as galvanic plating in the past.

But even then, you have a second concern, and that's AC currents on ground. In your motor-generator the two machines are connected with a steel rod. It's not possible not to have at least some AC noise on the DC ground without replacing that one by glass fibre or something like that.

Use a battery. Charge it if the device is not used. Medical equipment is sometimes designed that way. (It makes also the compliance testing simpler.)

  • \$\begingroup\$ Is it a requirement that motor and generator are both electrically connected to "ground"? \$\endgroup\$ – sharptooth Aug 29 '18 at 12:24
  • \$\begingroup\$ The AC motor has a steel casing, and the DC motor has also a steel casing. That's your common ground. If you don't connect it to an earth contact, the AC noise on the casing is actually worse. It's created by the capacitance to the conductors and by induction. \$\endgroup\$ – Janka Aug 29 '18 at 12:25
  • \$\begingroup\$ Technically it would be no problem to have no shared casing. Both would stand on ceramic stands and be connected with some dielectric part to transfer the torque. \$\endgroup\$ – sharptooth Aug 29 '18 at 12:27
  • \$\begingroup\$ I find it still simpler to buy a $40 car stereo, a $30 car battery and a $20 car battery charger. And a pair of $400 20kg chipboard tower speakers. \$\endgroup\$ – Janka Aug 29 '18 at 12:31
  • 1
    \$\begingroup\$ I’m very skeptical that a simple generator would have a better distortion performance than mains \$\endgroup\$ – PDuarte Aug 29 '18 at 12:32

Would a motor-generator set run on the original powerline be better?

The smaller the generator, the more distorted the waveform will be. This is simple physics since smaller generators will have less and less iron, and will produce more triangle-like waveforms. Especially when loaded.
You'd need at least a few dozen kVA, preferably even higher, to get near a "clean" cosine. The grid is (for households) "infinite" kVA due to the low impedance, thus in theory more ideal than any generator.

If your goal is to purify your input to filter anything beyond 50 Hz, transformers do typically perform a really good job at this. Combine these with line reactors, and you'll have a pretty decent filter.
But high energy loss.

However, when a large customer in your area generates a very high energy, low frequency harmonics (up to a few kilohertz). This will be audible in some systems, and very hard to filter out.
A dual conversion UPS or generator might be necessary.

Or batteries, that's always the best.


If you want to sell clean power solutions to audiophiles, you might consider two products.

One product should be a dual battery DC supply with a charger and automatic micro-computer controlled battery alternation system.

The second product should be a motor-generator system with two stators in the same housing and two rotors on the same shaft. That would eliminate any coupling backlash due to torque variation caused by power level variations. Design the motor and generator as short, large diameter machines. That would allow space for a shaped interior-magnet rotor design and optimized distributed windings on the stators. A proprietary design of that type would eliminate competition from cheaper, two standard machines bolted to a base, products.


Consider a 100 amp peak waveform with 10uS turnon, in the rectifier diodes of the Audio Power Amplifier. The dI/dT is 10^7 amps/second or 10,000,000 amps/second.

Assume the HOT lead of the 117VAC power is 10cm from a twisted pair running between the phono cartridge and the phone preamp/de-emphasis circuits. We'll first assume the twisting of the phono wires does nothing to attenuate the power line interference, then we'll come back and modify various assumptions. Assume the twisted-pair wires are 1mm apart and of length 1meter.

The voltage induced into the wiring from the phono cartridge is the standard

Vinduce = 2e-7 * Area/Distance * dI/dT

Vinduce = 2e-7Henry/meter * (1mm * 1meter/ 0.1 meter) * 10^+7 amp/second

Vinduce = 2e-7 * 0.001/0.1 * 1e+7

Vinduce = 2e-7 * 1e-2 * 1e+7 = 2e-2 = 20 milliVolts

Given the strongest output voltage from a MM moving-magnet cartridge is 5milliVolts????, the SNR is -12dB, with the interference an evil-sounding singing.

And if the signal source is MC Moving Coil with 500 microvolt output (or even smaller), the preamp is hugely overloading and clipping.

What to do? (1) use DC power (2) place the 117VAC power blocks several meters away (3) be glad the usual phono twisted pair should give at least 20dB cancellation, depending on the quality/consistency of the twisting (4) be glad the standard power line wiring has the HOT wire adjacent to the RTN wire, thus another 20dB attenuation is yours for free. (5) pay $1,000 for a special woven HOT/RTN/Safety power cable (6) use steel chassis on the power block (7) use steel chassis on the RIAA preamp (8) keep the power cable far away from the phono-cartridge cable (9) install heavy power-line filters in the Audio Power Amplifier rectifier circuits, to avoid conducted emission of the rectifier fast turnon


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.