|
Before investigating pushing the limits with regards current, it does well (as we did with voltage) to ascertain what is the proper way things should be done. Putting aside the inaccuracies introduced through resolution when using an input at the bottom of its range, a problem also exists when using current clamps at the bottom of their range. Current clamps, effectively being transformers, also suffer the loss known as "magnetising current". This is the energy required to magnetise the core to the flux level needed to generate the correct current in the secondary. If a clamp is operated between the recommended 10-100% of its current range, the output is usually sufficiently linear to ignore the inaccuracies in the reading. Below 10% the magnetising current starts playing a significant part in the reading - or should we say the lack thereof. The non-linear response at the lower extreme can be seen in the accompanying graph. Operating a clamp above the designed range will also create a problem as the magnetic material of the clamp saturates (known as hysterisis), significantly changing the waveform and therefore the RMS value and resulting in the output reading low. In steps the Rogowski Coil. This beast has two superb features that make it very suitable for certain issues encountered in power quality measurements. The first is it uses no magnetic material and therefore cannot be overloaded like the magnetically cored clamp. This makes it ideal for capturing high fault currents, as long as the integrator amplifier can process the signal input and not clip the waveform. It is, owing to the typical design of an integrator, pretty close to impossible to damage the integrating amplifier. There is no also no issue with regards linearity vs frequency because of the coil not being coupled with the aid of magnetic material (and can therefore closely exhibit the coupling by "square of the frequency" rule). The second (and probably the most attractive) feature of the Rogowski is its flexibility. It is ideal for sensing currents where standard magnetic clamps cannot or are dangerous to fit. Places such as busbars, switch panels, and tightly wired distributions boards are just some of the places where such a clamp is perfect for use. Ok! If they were so good there has to be a catch, especially as magnetically based clamps are still in production. And you are right. Rogowskis are not as accurate as magnetic clamps. No one seems to state their Rogowski coils at anything better than 2%, and then that's really pushing it. The problem is they never sit "square" to the conductor. I've tested many and have found them to drift as much as 3% by simply moving them on the conductor. Put it this way, I personally wouldn't use them for accurate consumption and/or billing purposes! They are also a little tricky when subjected to large fault currents as the integrator will start biasing the feedback capacitor and could force the measuring device to see this as a large RMS current (but this is more a fault of the integrator design than the coil). If it is suspected that there is a large fault current that needs to be captured then having a clamp capable of just a little more than the operating current is asking for bad readings. We cover this type of problem later. That's enough slating of the Rogowski coil. They do have their well deserved place in power monitoring and as long as we accept their limitations (as if the magnetic clamps don't have their fair share!) then we're ok. With current clamps, frequency response is affected by the magnetic material. However, I have personally found they become more accurate as the frequency goes up, or at least till 5kHz. Based on this the clamp is good to the 100th harmonic so I doubt there will be any issues here. When the Current is Too Small >>
© 06.04.02 |
MEASUREMENT TECHNIQUES: