I have read about a shielding system to reduce or eliminate harmonics by polarization: to design a system for your needs have a look at the document linked here: if you find it useful please discuss it here, I would appreciate learning more about these applications.
On DC rectifier systems (DC drives and standard AC drives) for industrial systems there are significant harmonics generated.
Good practice would be to have these drive systems on there own drive type shielded isolation transformer, preferably with delta wye windings and a High (or medium) voltage primary.
Other (sensitive) loads should not be on a drive supply.
Having said all that, the drive systems can be supplied with harmonic filters to eliminate the 5th, 7th, 11th harmonics. It is costly and not required if the above practice is followed.
With a UPS, you have to manufacture a sine wave from DC... and typically at a much higher voltage. The simple way to do this is to chop the DC at line frequency and send it through a transformer. Of course the chopped signal is going to be rich in harmonics. (won't go into Fourier here)? The concept of filtering out the odd harmonics, seems like a waste of energy. All the energy in those harmonics has to be dissipated in heat.
OR IS the harmonic problem generated by the loads? I can imagine many multi-polar motor giving varying loads as per the phase and windings of the motors, and thus providing harmonic distortion back to the source.
Actually, the power dissipation is not that great in the filters. It is essentially an LC tuned tank circuit that blocks the harmonics from the being propagated down the line. However, it has the net effect of lowering the apparent impedance of the system and can severely stress the scr's or diodes. I don't recommend them.
The harmonics are generated in DC drives and rectifiers by the non linear drawing of power from the line. The load voltage waveform from the rectifier represents "camel humps" or "saw teeth" depending on the phase control, but the load current can mirror that for resistive loads, or be dead flat from inductive loads, capacitive loads with a small inductive series impedance rides the crests. No matter, you end up with the odd harmonics, diminishing with 1/harmonic number. So first 3 harmonics catch the bulk of content.The 11th harmonic is much smaller and is usually ignored.
For the harmonic content I reference the IEEE book "Power Semiconductor Applications" or Dewan and Straughen "Power Semiconductor Circuits"
I am sure there are many more texts on the subject, my references are getting old.
Yes there are. Here is a list in order of increasing cost and performances.
LC-R filters. (Complicated to design with risks but the least expensive. Improve power factor)
Mitigation transformers. (Must be size properly, Don't reduce all harmonics. Don't improve power factor)
Active harmonics filters. (Simple to size and use as they adapt to the load automatically, reduce most harmonics, improve power factor, most expensive per ampere)
Send me a mail if you need details. I can recommennd suppliers.
__________________
Experienced is earned, common sense is taught, both are rare essentials of life.
Harmonic Distortion in Supply Systems
Re: harmonics distortions
