In the technical documentation on the chip of the controller of the pulse converters you can be found entire sections dedicated load mode “ligth” (light load), in which the output current is considerably lower than the nominal or non-existent. The modern controller with a light load can operate in continuous mode, Burst mode, discontinuous conduction mode, Forced continuous conduction mode and other less common mode. Some chips support different operating modes, that allow to optimize the characteristics of the converter according to the specific situation.
The presence of many options shows that the work under light load has its own characteristics. In the literature on the subject, it is seen that all the authors distinguish two modes of operation of the power unit, depending on the nature of the current (or stream) inductor power (figure 1). A CCM (continuous conduction mode) the inductor current flowing to the entire conversion period, in contrast to the discontinuous mode DCM (discontinuous conduction mode), in which the inductor current is absent for some time. And if there is no problem with the analysis of the electrical processes in a continuous mode, you can not say the same thing in discontinuous mode, the processes in this regime are much more complicated and, so most people do not consider this problem or provide complex and obscure explanations of what is happening in the part of the circuit power.
Figure 1 Magnetic flux in continuous mode and discontinuous.
So what is the characteristic of the light load?
After all, if there are no problems when you convert your 100% power, So why appear when you convert 1%?
In the real world, this usually does not happen. If the bag does not tear when it contains 10 kg, It will not break if it contains only 100 g Similarly, If a truck with a load capacity of 3 tons is loaded with a lot of weight 30 kg, there should be no trouble on the trip.
But this does not apply to pulse converters. For these schemes, processes to convert the 1% of the power are significantly different from full-load processes. E, perhaps, This problem does not require attention detailed as if there were no applications recently bidirectional energy conversion, whose realization is impossible without a detailed understanding of the characteristics of the light load regime. Among these applications there are voltage converters CA (AC converters / AC), Network inverters for solar and wind power plants, power regulators for battery-powered devices and other.
In this article I will examine the characteristics of the operation of the pulse converters in all possible modes, from the point of view of the direction of energy transfer:
transmission – when the energy is transferred from the entrance of the converter at its output,
to a minimum – when the energy is not transmitted,
recovery – when the energy is transmitted in the opposite direction – the input of the.
Consider the work of a flyback converter idealized (assembled with elements with ideal characteristics lossless). In the most common mode, let's call transfer mode, the electrical energy is transmitted in one direction: from the entrance of the converter at its output.
In this mode, the energy from the power source to the load arrives in portions (pulses) the value WIMP. Every second the converter passes NIMP energy. Each portion is converted into two phases (Figure 2). In the first phase the energy WIMP via the closed switch S1 is transmitted from the electric field in the capacitor C1 to the magnetic field of the inductor L1, in the second phase the L1 magnetic field through the closed switch S2 is poured into the electric field of the capacitor C2.
Figure 2. Transmission Mode.
The load connected to the output of the converter consumes energy from the capacitor C2 at a speed PH. The amount of energy in the capacitor WC2 and the voltage on its plates are related:
TheO – voltage on the capacitor C2, equal to the output voltage of the converter;
FROMC2 – capacitance of the capacitor C2.
Most converters stabilize the output voltage, supportandola, and then the amount of energy in the capacitor C2, it is constant. obviously, for this, the following condition must be satisfied:
If the condition (2) It is not satisfied, then the amount of energy in the capacitor C2 will continue to increase (The FH<WIMPNIMP) or decrease (The FH> WIMPNIMP) until it establishes a new balance ( PH= WIMPNIMP= 0), or something does not break, because a constant increase in voltage on the capacitor will lead sooner or later to breakage of the dielectric.
Under light load, the power PH tends to zero, then in order to avoid the increase of the output voltage, the controller must reduce NIMP oIMP , or both.
It is easier for the controller to change the’IMP . In this case, under normal load conditions, the number of pulses corresponds to the frequency conversion (NIMP= fPR), and for light loads, a part of the cycles is skipped (NIMF<fPR). This mode is defined mode pulse jump. However, The WIM is diminished, the output of the pulse moves into low frequency (Figure 3), which it is unacceptable in some cases. Therefore, with the light load it is also desirable to reduce the amount of energy WIMP converted per cycle.
Figure 3. jump pulse mode.
Consider what influence the value WIMP . The energy in L1 transformer input occurs when it is connected to the capacitor C1 via its S1 button. We know that at the time of closure of the S1 key after the previous conversion cycle there is a certain amount of energy WHF:
FMIN – the magnetic flux of the magnetic circuit at the time of closure of S1;
AL – It is the design parameter of the magnetic circuit, Commonly used to calculate the inductance L of its windings (L = N2AL, where N is the number of revolutions).
Figure 4.Parameters of the magnetic flux of the transformer.
The key S1 is closed for a time t1 in which winding W1 L1 of the input voltage U is appliedBX , under the action of which, according to the Faraday's law, the magnetic flux in the inductor is changed by an amount ΔФ
where n1 is the number of winding turns W1.
Then, when it is opened S1 button, the magnetic flux in the transformer will reach the value ФMAX :
that corresponds to the energy WMAX:
Subtracting (6) the magnetic energy at the time of closing of the S1 key (3), You are obtained in the amount of energy transmitted by the capacitor C1 transformer:
where ФСР = average value of the magnetic flux in the range t1 :
The formula (7) It shows that the power absorbed by the load influence the parameters of the magnetic flux in the transformer. To reduce the WIMP under light load, you need to reduce the constant ΦCP Df or the variables of the magnetic flux components.
The value is determined by ΔΦ (4). Of all the components of this formula, the controller can only change the duration t1 , because in almost all the key S1 is controlled converters. However, the output voltage also depends on t1 , which is defined for the flyback converter by the formula:
N2– number of turns of the winding W2
t2 – the duration of the closed state S2 button
Therefore, in order to ensure that the output voltage does not change, along with t1 you need to change t2 in such a way that the ratio t1/ t2 remains the same (Figure 5). But when using a semiconductor diode uncontrolled as S2 key, the ability to control the duration t2 it's impossible. And even if t2 It decreases automatically and the controller may provide the desired ratio t1/ t2, all equal, to reduce t1, even with t2, It is possible only within certain limits. Sooner or later there will come a time when t1 o t2 It will be less than the time required to change keys, because in practice they are not ideal.
Figure 5. FMagnetic luxury in discontinuous mode.
It turns out that in a discontinuous manner, which it is used in this method of dimming, the inverter can not, in line of principle stabilized no-load voltage when PN= 0, t1= 0 e t2= 0. If the controller does not support additional pulse discontinuous mode , then the power section must necessarily have a minimum load, to which it is still possible to somehow keep the output voltage within the required limits, otherwise it will not be controllable. To do this, the output is usually loaded with a loss resistor to have the minimum load.
Mode to a minimum
In agreement with (7) e (9), the support of the light load regime can be guaranteed by decreasing the average value of the magnetic flux ΦCP , preferably without changing t1 e, Consequently, DF. Second (5), the movable element DФ determined by considering the sign of the magnetic flux, however, if you have FMIN = -FMAX, then, by the formula (8), You are obtained FCP= 0 for arbitrary values FMIN and ФMAX .
What gives us this?
The variable component of the magnetic flux ΔΦ depends on the ratio between the voltages at the entrance and exit of the converter; In thatIN/ TheOUT , second (9), It depends on t1 , and on it, second (4), DF. Therefore, Af during operation of the converter is actually determined by the loop voltage stabilization. If the 100% of the power converter operates in mode- Break (FDo I> 0), decreasing the load current value PSTART and PWITH They are reduced by the same amount without Af changes. These processes take place as long as FDo I will not reach the zero value (Figure 6).
From this moment, the drive shifts to the light load mode and its further operation depends already from the base element of the power section.
Figure 6 Magnetic flux when the load current decreases.
If the S2 key is a semiconductor diode is not controlled, the inverter enters discontinuous mode in which Af and FSR They are reduced simultaneously reducing t1. But if the S1 and S2 keys are able to pass current in either direction, for example, when constituted by MOSFET, the inverter enters a forced continuous conduction mode. In this mode, the variable component of Af does not change and the decrease of the converted power occurs only due to a decrease of PhiCP.
Figure 7. Operating the Converter inactive.
A further decrease of the load current will lead to a shift even larger than the magnetic flux in the negative region. When the load is completely disconnected, the inverter will go into idle mode, whose characteristic is the observance of equality ФMAX= -FMIN. In this mode, the exchange of energy between the capacitors C1 and C2 occurs in the W valueXX (Figure 7):
At the closure of the switch S1 energy WXX the first inductor L1 is transferred to the capacitor C1 until the flow reaches zero, the inductor L1 and discharged. After that, under the action of the voltage on the capacitor C1, the energy will flow again in the accelerator, but already with a different polarity of the magnetic flux. In the moment in which opens the switch S1, the L1 transformer will contain the’ In the energyXX that, after switching the switches S1 and S2, It will be transmitted to the capacitor C2. In the middle of the second stage of the transformation, after that the magnetic core has been completely downloaded, under the influence of the voltage on the capacitor C2, the magnetic flux will change again the sign and the coil will begin to absorb energy from the capacitor C2.
An obvious advantage of the forced continuous conductivity under light load is the complete controllability of the converter. In this mode, the duration t1 e t2 not dependent on the load current, that provides more effective regulation of the output voltage, in contrast with the burst mode and the skip mode pulse. Disadvantages include higher energy losses due to the forced conversion WXX, that for some applications it can be a serious problem.
The forced continuous conductivity mode is only possible in cases where the switches S1 and S2 provide current flow in both directions, because with variable magnetic flux, in accordance with the law of the total current, also the current in the windings will be variable. For the flyback converter considered, in which the current always flows through one winding, the relationship of the currents I1 I e2 of the windings W1 e W2 to the magnetic flux F it will be determined by the formulas:
Only the MOSFET can be driven for passing the current in both directions, then the continuous forced conductivity mode is only possible in the synchronous converters based on this type of semiconductor devices (Figure 8). If at least one of S1 and S2 is based on bipolar transistors, IGBT, diodes or other elements, in which the current can flow in only one direction to implement a forced continuous conduction mode is necessary to adopt additional measures.
Figure 8. synchronous and non-synchronous converters
Furthermore, the role of C1 and C2 becomes obvious, acting not only as filters, but also as energy storage devices, fundamentally necessary for operation under light load.
And what if the average value of the magnetic flux FCPwill have an opposite sign ΔФ , for example, The ФMAX< 0 and ФMIN< 0 , when the condition ФMAX< FMIN ? In this case, second (7), WIMP<0, and the energy through the converter will go in the opposite direction from the entrance (Figure 9).
Figure 9. Methods of regeneration.
When such a scheme is necessary?
Eg, if the input of the converter is connected to the system bus and the output to the battery that contains the emergency power reserve (Figure 10). In normal mode, the system is powered from the main source and the converter performs the function of the charger, while energy is transferred from input to output of the converter, that corresponds to the transmission mode. If the battery is charged, the energy is not transmitted anywhere and the inverter is at a minimum. In the event of a power outage, the battery power is sent to the power bus through the converter that operates in the regenerative mode,, providing power to the load.
Figure 10. Example of operation of the converter in three modes.
It should be noted that the transition from one mode to another is automatic, without any participation by the controller, whose main task in this case is only to maintain the desired ratio t1/ t2 so that, in base a (9), provide both the U-value requestIN/ TheOUT or the load current request.
To ensure that the light load mode does not create problems, the magnetic flux should be able to change its polarity.
With the DC conversion / DC, This is easier to provide converters with synchronous MOSFET-based. Furthermore, this is automatically executed in AC converters / AC, because in them the flow of alternating current through the power switches, however, as the work on the reactive load, It is a prerequisite. In other cases, you need to carefully study the light load mode to ensure the required accuracy, reliability, the efficiency and other characteristics of the converter.
List of sources
In the past I have pointed out that not quote my sources, now for the future, as far as possible, I will try to cite sources.
Kadatsky OF, Rusu? D. Analysis of the principles of construction and mode of operation of pulsed electrical energy converters // Power electronics practice.- 2016. – No. 2 (62).- P.10 – 24.
Kadatsky OF, Rusu? D. Analysis of electric and magnetic processes in bottlenecks of electricity converters pulse // Technology and design of electronic equipment (TCEEA) – 2016. – № 6.- P.17 – 29.
Rusu? D. Transformation of pulsed alternating current / / Radio Lotsman – 2017. – № 6.- P.24 – 32.