Time for a little DC current measurement refresher, perhaps?
Measuring DC current has always been tricky. Normal transformers do not work - for obvious reasons.
In the early Ward-Leonard systems, a shunt was all that was needed. This was possible since the WL systems were completely isolated from the grid and were usually floating. The 50, 60 or 150 mV current signal from the shunt was amplified and used in controller and protection. It could also be fed to a panel meter for current measurement.
Later on, when thyristor controlled rectifiers came along, the shunt signal wasn't isolated any more. But the shunt was a simple and accepted technique and by adding an isolation amplifier - usually with chopper and transformer isolation - an isolated signal that could be used in the control and protection circuits was obtained.
The measuring transductor (not transducer) is a magnetic amplifier with no amplification (1:1 amplification). If an iron core is wound with a primary winding and a secondary winding and the primary winding is DC, then the AC fed secondary winding will carry a current that is proportional to the DC. There is a small (and constant) excitation current that can be adjusted for). Rectifying the secondary current and feeding it to a resistor results in a simple and stable DC transducer. This transducer type was once called a Kraemer transducer. Do not see it in use any more. It provides no polarity information.
There were also early Hall effect transducers. The Hall effect devices were of a quite bad quality in the early days and could not be used for direct measurement of current. Instead, they were used as null detectors in a magnetic circuit where the primary DC current was nulled by a reverse field from a multi-turn winding (1000 - 10 000 turns) where the secondary current was adjusted by a little servo amplifier until the Hall element indicated balance. The resulting current was then a measurement of primary current (divided by turns ratio) and the output can be either the current itself (DC current transformer) or a voltage proportional to the current (DC current transducer).
Direct Hall effect transducers work directly with a Hall effect device picking up magnetic flux from the primary conductor. They are used widely in modern frequency inverters where current measurement is critical even at very low motor frequencies. LEM is the predominant manufacturer.
Rogowski coils can be used for current measurement under condition that current goes to zero periodically so that a stable zero reference can be obtained. They are particularly well suited in large DC plants for aluminium and chlorine fabrication where the 20 000 – 50 000 A currents saturate normal iron cores. The Rogowski has no iron to saturate and it works wonderfully in these applications.
Modern electronics and fiber optics are used in DC transducers for HVDC transmission. They utilise the Faraday effect. The Faraday effect causes polarized light in an optic fiber to change polarization when a magnetic field is applied to it. The effect is strictly proportional to H (the Magneto Motoric Force, MMF) and since the line integral of H around a conductor is proportional to current in the conductor (same goes for Rogowski) the result is an excellent DC current transducer. ABB has developped this technique and use it in their HVDC systems.
Flux gates could also be used, but I haven’t seen any application for that device as a DC transducer yet. They will probably be too sensitive to external fields.
Gunnar Englund
