The main experiment
The question that we need to answer first
1) what are the various types of noise that comes out of the Power output of 5 volt mobile charger ?
2) How these noise are handled ?
1) what are the various types of noise that comes out of the Power output of 5 volt mobile charger ?
Ripple Noise:-
what is it ?
A non-ideal DC voltage waveform can be viewed as a composite of a constant DC component (offset) with an alternating (AC) voltage—the ripple voltage—overlaid. The ripple component is often small in magnitude relative to the DC component, but in absolute terms, ripple (as in the case of HVDC transmission systems) may be thousands of volts. Ripple itself is a composite (non-sinusoidal) waveform consisting of harmonics of some fundamental frequency which is usually the original AC line frequency, but in the case of switched-mode power supplies, the fundamental frequency can be tens of kilohertz to megahertz.
what causes it.?
· Switching noise. This appears at specific frequencies on the output from a switching regulator (buck-boost converter, flyback/forward converter, VRM, etc.). This noise
· arises from the PWM signal used to switch one or more FET drivers in the converter.
Ringing noise:-
what is it ?
“Transients”, a term used for simplicity, are “Transient Voltages” or more commonly referred to as “surges” or “spikes”. Transients are momentary changes in voltage or current that occur over a short period of time. In the context of power supplies, noise is any undesirable voltage impulse which appears at the output. In electronics, ringing is the oscillation of a signal, particularly in the step response to a sudden change in input. Ringing is often undesirable, but not always, as in the case of resonant inductive coupling. It is sometimes described as a “ringing transient”. This type of transients is characterized by swings above and below the normal line voltage.
what causes it.?
· Transient ringing. This type of oscillation looks like ripple from power conversion, but it is fundamentally different. This arises from the parasitic capacitance and inductance in the PDN. As an IC switches states or output levels, it draws a burst of current that produces a transient oscillation, which is sometimes called a ripple voltage. This transient response can appear on the power rails even if the supply voltage is regulated.
Switching noise jitters:-
what is it ?
Switching Noise Jitter is the aggregation of variability of noise events in the time-domain on the supply bias of an electronic system, in particular with a voltage regulated supply bias incorporated with closed-loop control, for instance, SMPS.
what causes it ?
In switching-mode power supplies (SMPS), noise present in the control loop circuitry of the supply causes dislocation in up-slope and down-slope timing of the saw-tooth ripple waveform. As a consequence, the ripple waveform exhibits jitter and noise carried on the ripple also jitters.
common mode conduction emi noise:-
what is it ?
Common-mode noise in terms of ac power is the noise signal between the neutral and the ground conductor
what causes it ?
AC line transients, such as line surges due to lightning strikes, power switching from motor controls, circuit breakers or relays actuating, can cause both differential and common mode disturbances
differential mode conduction emi noise:-
what is differential mode noise ?
The differential (normal) mode refers to signals or noise that flow in opposite directions in a pair of lines.
what causes it ?
AC line transients, such as line surges due to lightning strikes, power switching from motor controls, circuit breakers or relays actuating, can cause both differential and common mode disturbances.
2)How these noise are handled ?
ringing noise:-
We can bypass the ringing noise of the 5 volt charger by either using a large bypass capacitor for that ringing frequency or we can use Inductor to attenuate the noise.
Using capacitor to filter the ringing noise to ground.
Figure 1
Figure 1 shows the original ringing noise. frequency of ringing noise is 1.96 MHz. Maximum voltage is 460 mili volt. peak to peak is 940 mili volt. We have set the probe of the oscilloscope to 1X so
it can give 10x magnification. So the orginal Maximum voltage is 46 mili volt. peak to peak is 94 mili volt respectively.
Bench setup for the experiment:-
we have use 100nf ceramic capacitor to filter the noise.
Figure 2
Figure 2 shows attenuation of ringing noise using 0.1 uf capacitor. The noise signature on the scope is around 200 milivolt peak to peak that comes down to around 20 milivolt beacuse of 10X amplification used in the oscilloscope probe. The problem with using ceramic capacitor for attenuation of the noise is that, they provide best impedance at center frequency. if we change capacitance value to any higher the center frequency will change so as the impedance of attenuation for that frequency. you can see it in the graph bellow.
We can use multiple capacitor with same value in parellel to increase attenuation but that will not imapct it largely. We can replace single capacitor using capacitance multiplier circuit for the highest effect.
Figure 4
Figure 4 shows capcitance multiplier circuit that is used for attenuting ringing noise of 1.96 MHZ.
Bench setup for capacitance multiplier circuit
Figure 5
In Figure 5 you can see the attenuation of ringing noise that has come down from 94 milivolt peak to peak to 30 milivolt peak to peak i.e approximately equal to 3 milivolt because of using 10X amplification of probe.
Using Inductor for attenuating signal.
Inductor not only attenuates the ringing noise but also many other noises as well. The problem with inductor is that it introduces winding resistance which causes voltage drop accross the DC line.
Figure 6
Figure 6 shows attenuation of ringing noise using inductor.
The red scope line indicates the attenuation done by inductor. The noise amplitude is aproximatrely 92 milivolt which comes down to around 9.2 milivolt because of using probe at 10X amplification. We have use 100 milihenry inductor provided by newark. component RL181S-104J-RC.
Based on this analysis we have computed some crieteria that needs to be looked at while choosing the design. These are
1. board area
2. attenuation required
3. cost of manufacturing the circuit
4. voltage drop
5. maximum allowed current
6. no of other harmonics that needs to be attenuated.
7. heat generation
Filtering technique | Capacitance 0.1 uf | capacitance multiplier 0.1 uf | inductor 100 milihenrey RL181S-104J-RC |
---|---|---|---|
board area | 20 sq millimeter | 625 sq millimeter | 100 sq millimeter |
attenuation done | 6 db attenuation | 14 db attenuation | 9 db attenuation |
cost of manufacturing the circuit | Na | Na | Na |
voltage drop | 0 milivolt | 520 milivolt | 160 milivolt |
maximum allowed current | Na | Na | Na (Please note Na means not applicable) |
No of other harmonics that needs to be attenuated | Na | Na | Na |
heat generation | Na | Na | Na |
Ripple noise:-
the main wave form is shown in the bellow figure7 and the FFT spectrum of it in figure 8
Figure 7
Figure 7 shows the orginal ripple from the output of the 5 volt mobile charger. Which is a swatooth ripple along with switching noise jitter on the ripple waveform and transiant ringing noise.
Figure 8
Figure 8 shows the FFT spectrum graph of the swatooth waveform. From the FFT graph you can understand that swatooth ripple is made up of multiple low frequency sine wave harmonics from 1 hz to 5 khz range. Figure 9 shows the waveform after filtering the ripple noise and and it’s FFT graph at Figure 10.
Bench setup for filtering ripple noise
Figure 9
Figure 9 shows a straight line that is achieved after filtering all the low frequency noise from the swatooth wave. The FFT spectrum graph in Figure 10 shows that there is not low frequency ripple left to attenuate.
Figure 10
Figure 10 FFT spectrum graph. In which all the low frequency spectrum has been attenuated.
jitter noise:-
For filtering jitter noise we don't have any proper instrument. So we are leaving it there. As we are working on building some custome instruments to deal with it. As we finish designing it. We will update it on ths blog.