20
$\begingroup$

This question on Travel.SE got me thinking... Just how stable is the power that's available via the outlets in the passenger cabin on a modern, commercial aircraft?

My assumption in my answer that it's probably quite clean was based on these things, but I'm not sure if or how accurate these assumptions are:

  • Power for avionics needs to be clean to ensure there are no spikes or dips that would cause avionic failures as this could lead to otherwise avoidable emergency declarations.
  • Since the plane is already providing clean power to the avionics, the power coming out of the transformers to step it down to 110V, 60Hz for the outlets should also be pretty clean.

Are these accurate assumptions, or is the effort to provide nice, super clean power to the cabin outlets too much work/too expensive to bother with getting it any nicer than the sketchy quality stuff the power company provides to your house?

$\endgroup$
  • 3
    $\begingroup$ Avionics components are likely using 28 V DC which can be rectified to the level of defect required, while 220V or 110 AC is obtained by a static converter. There may be not relation between AC and DC default level. What do you mean by "stable"? Voltage or phase/frequency or harmonics? That said, today's mainstream equipment doesn't require a good filtered AC source like in the past. Often they can accommodate between 220 and 110 V without a switch. $\endgroup$ – mins Jun 14 '17 at 13:40
  • 4
    $\begingroup$ @mins - stable as in "I'm a layman and I don't want my laptop/charger blowing up when I plug it in." i.e. no need to worry about providing a surge suppressor and/or UPS when I fly. $\endgroup$ – FreeMan Jun 14 '17 at 13:40
  • 1
    $\begingroup$ Clean power from the power company? Have you tried putting a scope on your home outlet? Especially if you live in an industrial area... The voltage on your home outlet can swing from 110V to over 120V. If you live near a factory, the frequency can swing quite a bit too, usually 58-62hz is considered "good" power, but large factories (like smelting) starting up a piece of equipment can drag the grid frequency down as low as 55hz. $\endgroup$ – Ron Beyer Jun 14 '17 at 13:53
  • 1
    $\begingroup$ @RonBeyer - There's a reason that every PC & all my home theater equipment is behind a UPS. My last statement was intended to indicate "...any nicer than the crappy power that the power company provides...". I'll reword it a bit to make my intent more clear $\endgroup$ – FreeMan Jun 14 '17 at 14:49
  • 3
    $\begingroup$ Most switch-mode power supplies (such as used for laptops and other chargers) are very tolerant of frequency changes (the first step in sonverting to DC) and voltage changes (these days many are rated for 90-264V). Spikes and brownouts are a different matter $\endgroup$ – Chris H Jun 14 '17 at 15:38
16
$\begingroup$

There is a pretty great article on the design of the 777 electrical system here.

Power for avionics needs to be clean to ensure there are no spikes or dips that would cause avionic failures as this could lead to otherwise avoidable emergency declarations.

The avionics and passenger power systems are two completely different systems. The 28V instrument power system is pulled right from the source,

Flight instruments on the 777 are primarily powered by 28 Volts DC. Direct current power is supplied by three-phase transformer-rectifier units. There are four such units each rated at 120 Amperes.

This pulls right from the generator/alternator output of the plane. Everything ranging from your classic Cessna 152 all the way up through a 777 regulates their alternators/generators with voltage regulators and fuses to protect against current overloads. These things work in tandem to protect the avionics and internal power systems of the plane. The 28VDC system for the avionics is also connected to the battery system which prevents dips should the generators have issues outputting enough power or quit all together.

The rest of the plane is powered by 110/200V 400Hz system.

The primary electrical system is a 3-phase, 4-wire, constant frequency 400 Hz, 115/200 Volt system, which has been the industry standard on commercial jet transports since the Boeing 707

Interestingly it's 400Hz, I'm still looking to see how they get that to 60Hz.


Since the plane is already providing clean power to the avionics, the power coming out of the transformers to step it down to 110V, 60Hz for the outlets should also be pretty clean.

The power is quite "clean" but you have some incorrect assumptions here. The power is not really stepped down. That is something that occurs on the power grid since most power lines on the street carry a significantly higher voltage than what you have in your house plugs. In this case we have the luxury of choosing an alternator/generator system that is close to (most likely slightly above) our required voltage. Slide 4 from this presentation from Boeing alludes to the fact that the system outputs 115V (and is presumably conditioned) and is then split to the cabin as well as the 28V DC bus. Power does not come out of the transformers it comes out of the generator/alternators it then is conditioned and transformed as needed. Stability is achieved though numerous conditioners and regulators, see the article I linked for more details on them.

Don't worry, it's safe to charge your laptop on a jet.

$\endgroup$
  • $\begingroup$ So my assumption of clean AC/USB power for charging was correct, it was just my thoughts on how it gets there was wrong. $\endgroup$ – FreeMan Jun 14 '17 at 14:52
  • $\begingroup$ Generally yeah. Most power consumer power supplies (chargers) are capable of outputting pretty clean power from a wide range of inputs. You may want to use a small USB charger on the 120V socket in a plane to clean it up yourself if you are worried but I would not worry about aircraft power generally speaking. $\endgroup$ – Dave Jun 14 '17 at 15:15
  • $\begingroup$ There clearly must be a rectifier/inverter circuit involved somewhere. I would think that the characteristics of the output power would be almost entirely dependent on the characteristics of that rectifier/inverter circuit. $\endgroup$ – Peter Green Jun 14 '17 at 17:04
  • 3
    $\begingroup$ To underscore something that's implied here: a transformer adjusts AC voltage; it does not affect frequency. So changing that 400 Hz. power to 60 Hz. requires something other than a transformer. $\endgroup$ – Pete Becker Jun 14 '17 at 17:30
  • 1
    $\begingroup$ They likely don't get to 60Hz or 50Hz for that matter. 400Hz is actually a fairly standard industrial AC frequency - the higher frequency allows for smaller transformer components. Switch mode PSUs used for laptops, phones, etc. rectify to DC before regulation which is why they can cope with 110V/240V and 50Hz/60Hz. They should as a result also cope relatively well at 400Hz. $\endgroup$ – Tom Carpenter Jun 14 '17 at 19:18
13
$\begingroup$

AC to AC conversion with frequency change from 400 Hz to 50/60 Hz is obtained by transforming AC to DC first (usually to 28 VDC to be compatible with batteries) and then using a static power inverter to convert from DC to AC. The inverter performs the voltage transformation in the same step.

From an EASA certification perspective, electricity quality is measured at the output of static power inverters.

The requirements are described in ETSO-C73, with an exception about the frequency, as for passenger outlets the frequency must be 50 Hz or 60 Hz, depending on countries, instead of 400 Hz. This exception is described in this document:

Deviate from ETSO-C73 attached FAA Standard for Power Inverters paragraph 2.3 and provide 50 Hz +/-1 % or 60 Hz +/-1 % instead of the required 400 Hz.

ETSO-C73 itself refers to FAA requirements:

2.3 Frequency: The frequency of the inverter under all conditions of load and test environment shall be 400 cycles per second ±1 percent at the input voltages specified in 2.2a. and 2.2b.

2.4 Voltage Output: The average phase output voltage, under the conditions of input specified in 2.2a. and 2.2b. and under a ll conditions of test environment, shall be 115 volts a.c. +5 percent -7 percent

2.5 Waveform: The output waveform shall be substantially sinusoidal and contain less than 7 percent harmonic distortion under all load conditions not exceeding 110 percent rated output.

What does that mean?

  • Voltage value tolerance +5%/-7%. This is pretty nice and equivalent to what we have at home in industrialized countries. I addition, common devices are very tolerant to voltage variation, as they are simple to eliminate using a high value capacitance.

  • Frequency tolerance ±1% means ±0.5 Hz. Most, if not all, devices don't rely on a very stable grid frequency. It was the case of first digital clocks to save the price of a crystal oscillator, but now crystals and PLL are used everywhere when a stable clock is required.

  • Waveform with a maximum of 7% harmonics. This is the usually the most important tolerance. It describes how much the voltage wave can differ from a pure sine wave.
    Any periodic wave of frequency f which is not exactly sinusoidal can be broken down into a set of pure sine waves which frequencies will be f, 2f, 3f, 4f, etc, and inter-products. Breakdown is obtained by a Fourier analysis. The multiples of f are named harmonic frequencies, and in practical this is the same as if there was multiple mono-frequency generators.

Some of those frequencies can interfere with receivers or disrupt other circuits due to impedance changing with frequency. In this case the amplitude of the harmonics must be weaker than 7% of the fundamental wave.

enter image description here

For the power grid where I live:

  • Voltage tolerance is ±10%
  • Frequency tolerance is ±1%.
  • For harmonics, this is more complex as the customer's load affects harmonics, there are mutual commitments which are respectively for 2nd, 3rd, 4th, 5th, 7th orders: 2%, 5%, 1%, 6%, 0.5%, 5%.

USB sockets voltage is defined in the USB standard:

  • 5V +0.25V/−0.55V (+5%/-10%) for USB 3.

There is no difficulty to obtain it from any DC source to minimize conversion stages, losses and heat. This can be done by a push–pull converter. USB power is likely obtained from 28 VDC.

$\endgroup$
  • 1
    $\begingroup$ Missing a word in This is a very generator, (...) ? $\endgroup$ – kevin Jun 14 '17 at 17:28
  • 1
    $\begingroup$ The difference from a pure sinusoidal waveform is most likely going to be in the form of a step-wise approximation. This thread points out that desktop power supplies with active PFC (power-factor correction) can have trouble with UPSes that approximate a sine too poorly, and that having the voltage sit at zero for some time is a problem. Presumably 7% THD is small enough that no power supplies have a problem, and modern laptop PSUs are surely designed to handle airplane power. $\endgroup$ – Peter Cordes Jun 15 '17 at 3:12
  • 1
    $\begingroup$ @PeterCordes This problem seems specific to having two inverters in series (the UPS and the internal power supply) and can have two reasons: The UPS delivers current which returns to zero more than twice per period (poor design), or the internal power supply loads the UPS in non-linear way which is often the case of a pulse width modulation power supply, this can create resonance in the load under certain conditions. $\endgroup$ – mins Jun 15 '17 at 7:56
  • $\begingroup$ Surely it's not just inverter -> rectifier -> inverter that's causing the problem described in that thread. People were saying it only happens with active PFC, and that older PSU with passive PFC or no PFC were fine. (Ironically, older PSUs like that would do best with a pure square wave input. IDK what active-PFC supplies would do with a square-wave input, since I haven't totally grokked active PFC). $\endgroup$ – Peter Cordes Jun 15 '17 at 9:41
  • 1
    $\begingroup$ So... what are the chances of getting a portable oscilloscope or data-capture card through security in your carry-on to investigate this? :P I bet they love it when people show up wanting to plug home-brewed electronics into airplanes. $\endgroup$ – Peter Cordes Jun 15 '17 at 9:42

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.