How can we understand the harmonics that are generated during the transient state of DC circuits like electronic circuits that uses DC power?
I think it should be more correct to talk about free frequencies of transition mode in DC circuit, and not about harmonics. The concept of harmonic is usually used for non-sinusoidal (more general, for non-monochrome) steady-state regimes.
Hi Yugeswar,
I feel it is a matter of periodicity: Usually, a non-periodic event would be called a transient while a periodic event has, obviously, a (fundamental) frequency, and, as Tahir Midhat Lazimov pointed out, if the temporal variation of voltage and/or current is non-monochrome, the fundamental frequency is necessarily accompanied by harmonics.
If we have measured and registered a varying quantity for a certain time interval we can always apply a transformation to the frequency domain (usually FFT) but since these transformations are based on the assumption that the time interval is a multitude of the fundamental frequency the result can be quite wrong. For example, a single event had to transform into a continuous spectrum but instead you will always get a discrete spectrum.
I'm not sure whether I understand your question correctly, but if you are asking why the spectrum of "real" transients seems to be discrete when processed using a spectrum analyzer, the answer is: These are artefacts, as I tried to explain above.
Power glitches can be approximated as a triangle pulse, usually with a steeper leading edge so these may be of use:
http://www.thefouriertransform.com/pairs/triangle.php
http://www.thefouriertransform.com/pairs/impulse.php
Respected Tahir sir, thank You for You reply. If there exists "free frequencies" as You have mentioned, how or which tool we could use to analyze them sir?
Yugeswar,
IT depends what kind of DC circuit you have in mind!
If it is just a passive circuit, with no active energy source other than the DC supply, then the "harmonic response" will depend on whether you have L & C circuit components in the "filter". The L component is an "inertial component" & the C components are "energy storage". IF you have them together in parallel then they will form a "resonant" LC combination that will oscillate at the natural LC = w2 frequency.
IF it is an active DC circuit like an "operational amplifier" with feedback, then with a resonant LC circuit in the mix it will resonate at some other frequency, depending upon the amount of feedback.
If it is an active DC-DC converter, using a phase modulated chopper to control the DC level and it has an LC filter across the output to smooth it, then it too will have a natural resonant frequency, which when you add in feedback to control the smoothed output will have a higher resonant freqency, usually as you wind up the gain.
If you have an AC - DC converter, especially a high power one driven by mains electricity supply then there may be smoothing filters on the AC side to suppress the supply harmonics, and this will result in a more complicated frequency response, involving harmonic modulation of the AC supply filter frequencies.
The simplest way to see whether you have a potential harmonic oscilllation problem is to turn on your circuit with a step change input and monitor the DC out put with an oscilloscope. If it shows just a spike, then you should have no problem with harmonic resonances. If you can detect oscillations in the output, then you have an inherent resonance, which may get worse if you turn up the loop gain too high.
Try it and see.
Best Wishes
Dr Hazell
I read the question to refer to transients in otherwise DC circuits which in a datasheet would typically be given as the response to a step change in the load. If such changes occur in a non-periodic (random) manner but with some mean rate, it is still possible to calculate a power spectrum using Fourier analysis.
Yugeswar Reddy
Dear colleague,
You know that in steady-state regime of DC circuit there aren’t harmonics (of voltage, current or magnetic flux). Magnitudes of voltage and current in this case will be equal to the zero-terms of corresponding Fourier-series.
Transitional regime (e.g. switching-on or switching-off) of DC circuit may be accompanied by free oscillations with natural frequency if your circuit has comparable inductive and capacitive reactance. At the case of relative negligibility of inductive or capacitive reactance we will have exponential change of transitional voltage and current’s functions.
You should know or calculate values of circuit’s inductance and capacitance to determine the natural frequency.
Regards
Capacitance is easy, sum the decoupling, but inductance is low mainly from the tracking and if it is a good supply plane should be negligible.
The transient response will usually be dominated by the response of the supply control loop.
- Badges
- Science topic
- Similar topics
- Filing