I am currently studying wind power and the power electronics used for it. In wind power a generator is driven by wind, thus the resulting power is of widely varying frequency and amplitude. The power grid, in turn, has strict requirements for the input power in terms of frequency, phaseshift and sinusoidal form. For this reason, power converters are today used routinely in wind power.

The predominant way to get the power into the grid is to use an AC-DC converter followed by a DC-DC converter and a DC-AC converter. This seems rather complicated instead of using a single direct AC-AC converter. Why is the indirect conversion via the DC "in-between" route preferable?

(This is actually a repost from Engineering, since I only found out later that there is a more active, thematically fitting, non-beta Electrical Engineering.)

  • \$\begingroup\$ Are you thinking of a doubly fed synchronous generator for the direct AC to AC conversion? \$\endgroup\$
    – Andy aka
    Jul 27, 2019 at 12:06
  • 9
    \$\begingroup\$ Because of the frequency difference, at some point your AC output might need a peak (high), but your input AC happens to be zero. Where is that high output to come from? In DC you can store the energy in a capacitor and use a short time later. You can also convert to arbitrary polarity. \$\endgroup\$
    – Oldfart
    Jul 27, 2019 at 12:45
  • 2
    \$\begingroup\$ @Oldfart Thank you, your answer seems very sensible. In essence, you are saying that some kind of "energy storage" is required. Interestingly, the AC/AC article on wikipedia comments on so-called Matrix Converters: "In order to achieve higher power density and reliability, it makes sense to consider Matrix Converters that achieve three-phase AC-AC conversion without any intermediate energy storage element. Conventional Direct Matrix Converters (Fig. 4) perform voltage and current conversion in one single stage. " These seem to forego the storage, are you familiar with their drawbacks? \$\endgroup\$
    – ckrk
    Jul 27, 2019 at 13:01
  • 5
    \$\begingroup\$ @mkeith You are right, there is a windturbine design that employs pitching to ensure a certain rotational speed. I think this is referred to as "danish model" and was historically the first grid compatibility approach. This is however getting out-of-fashion. The reason is simply that pitching basically leads to wasting wind power, today one rather tries to avoid pitching and regulate a separate-excited generator to match its optimal RPM to the current windspeed. \$\endgroup\$
    – ckrk
    Jul 27, 2019 at 23:13
  • 1
    \$\begingroup\$ There's another approach that seems rarely used, hydraulic power conversion: artemisip.com/wp-content/uploads/2017/11/… (author was the inventor of the Salter Duck wave power system) \$\endgroup\$
    – pjc50
    Jul 28, 2019 at 10:12

5 Answers 5


There is a type of converter which can do this: the matrix converter.

In theory it can take many phases in and produce many phases out at quite a wide range of frequencies. It also has the additional benefit of not needing any power passives (in theory), or no large capacitor, no large inductors.

However, there are two golden rules with matrix converters

  1. Thou shalt not short circuit the supply
  2. Thou shalt not open-circuit the load

It is point #2 that makes the topology impractical as a simple loss of power will cause the inverter to blow up.

There is a variant of the matrix converter called the cycloconverter which uses thyristors and does not suffer the same issues as a full matrix converter. It, however, has a limitation of only being able to synthesise an output frequency around 1/10th of the input frequency. This limitation is fine for marine which typically uses 400Hz electrical supplies so generating 40Hz isn't too limiting for propulsion

So why AC-DC-AC instead of direct AC-AC ... The complications and limitations. A six switch inverter is extremely versatile.


When two routes are possible, there is rarely a good answer for why one particular one was chosen. It's often accidents of history, or advantages to one or the other depending on local industries, or common components.

There is an all electronic route directly from 3 phase AC at one frequency to another, it's called a Matrix Converter. It contains 9 switches in a 3x3 matrix, to connect any phase to any other. With suitable timing of the switch instants, and suitable input and output filters, it can create a similar output voltage to the input. They are becoming increasingly used for motor drives.

However, I can think of many advantages to using an intermediate DC link.

AC-DC and DC-AC converters are being made in large numbers, in large sizes for DC links where long distance transmission is a factor. This will lead to economies of scale. They are more mature than matrix converters, so with the long planning involved in electrical infrastructure are more likely to have been chosen. Wind turbines tend to be connected in short hops to hubs before being connected to a single long distance transmission line (very long in the case of offshore). It's easier to pool power at a nominal DC intermediate voltage, simplifying control. It's easier to stay DC for the long transmission.

  • 2
    \$\begingroup\$ The motivation for using HVDC for offshore windmills is to piss off Trump, er, to reduce dielectric losses in the cable. It’s an outgrowth of technology developed in Sweden (ASEA, ABB) for HV undersea cables between islands. \$\endgroup\$ Jul 27, 2019 at 18:38
  • \$\begingroup\$ @hacktastical Does Trump have a specific thing against HVDC? \$\endgroup\$ Jul 28, 2019 at 22:48
  • 2
    \$\begingroup\$ He opposed an offshore windmill that faced his golf course in Scotland because it would ‘ruin’ the view. More here: bbc.com/news/uk-scotland-north-east-orkney-shetland-47400641 \$\endgroup\$ Jul 28, 2019 at 23:17

The reason for direct AC-AC conversion is the size and mass of the DC choke coil (or capacitor array). You don't want to have that e.g. in a rubber-wheeled subway car or aircraft. In iron-wheeled trains it depends, because more mass means better friction.

That doesn't apply to buildings.

You cannot save on valves (transistors or thyristors). In contrary, AC-AC converters tend to have more valves (though smaller ones) than AC-DC-AC converters. The control concept is also much more complicated.


AC-DC-AC conversion wins when you have several different AC sources to combine into a single AC output, (or when you have the opposite).

each asynchronous generater produces an AC supply that is rectified and boosted to a DC bus voltage, the bus voltage then feeds a grid tied inverter.


One advantage of the AC-DC-AC conversion is that you can convert the frequency of the AC. There is also military 400 Hz power, which can result in a considerable size-reduction. In my particular case, I needed access to motors that worked within a vacuum chamber. Equipment for NASA and military use was available that met our requirements, so we opted to use 400 Hz power. I recognize that it is rather specialized, so it is probably not applicable to you.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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