Thinking and analysis of DC transform Beginners

It is the front of many inverter circuits, which can carry out the pressure, then what is his concrete realization idea? What are the impact variables? What are the characteristics? What is the essence? These should be questions that are of interest to beginners, including me.

According to some of our previous impressions, Buck can be distributed by resistive load, while a transformer can achieve a lifting pressure. The transformer utilizes the inductor, the energy-transforming device. In fact, in the same way, the capacitance also has the function of energy conversion. Therefore, it can be assumed that the integration between them may be able to achieve the other way to achieve the lifting pressure.

Now we directly give a ready-made DC buck converter circuit, which is the circuit diagram of Buck Circuit. It is hoped that the analysis of it can deepen some understanding of DC conversion circuit.

The Q- switch tube is on and off with a fixed period and duty cycle (i.e. percentage), finally achieving a balance,and theuout also achieves buck. It is difficult to understand at first, how to achieve it?

Any one kind of circuit has corresponding mathematical model corresponds, may wish to carry on the mathematical calculation, the basic idea is simply:

Kirchhoff voltage σloop (Ui) =0, current theorem σnode (Ii) =0, the relationship between voltage and current at both ends of the inductor and capacitor :

Add some assumptions that are permissible. By the way, you can also add some intermediate results in the derivation process.

After switching on, the circuit through a period of adjustment, to achieve self-balance, here can refer to Xu De "power electronics Technology"p76 inductor current waveform. Although there, it is approximate to think that the inductor current is straight up and down is inaccurate (certainly not a straight line, because this is a second-order circuit, obviously it should be a curve), but we know that it is in a switching phase (or called cycle) first rise and fall this trend is correct. So after a period of time, can achieve a balance of state, that is, a switch phase of the first increase in how much, after the decline of how much equal.

By understanding it, it is important to note that this balance is attainable, which is necessary for the perfectionist. Because many books come up to assume that the circuit is in equilibrium, it is disappointing.

Now see what the physical quantities are, inductor l , capacitance c r uin d ts uout

1, according to a period of inductance current cycle balance, can be based on the inductor voltage and current relationship to deduce, can be exported on and off when the ∫u (t) DT and should be 0. In some books, the ∫u (t) dt=0 is called the Volt-second balance, the saying does not matter, that means. Next, either the direct quadrature, or the area (which is of course the same meaning), came to Uout=d*uin. Visible, as if with how much inductance, how much capacitance, how big cycle time does not matter oh? It seems that Uout is only determined by D , the intuitive feeling is that the result is very reasonable.

2. In the above1In the meantime, we gotilthe waveform, rising slope(Uout-uin)/L, descending slope-uout/l. Want to come, ifL, the slope is larger , so that the inductor current appears relatively less stable, and the bookildefined as1/2ilp-pPeak-to- peak, this uneven component is called Ripple. VisibleLit does affect something. This time if theilWhat do you get when you make points? Is the total charge through the inductor in a cycle, divided byTs, GetIlis the average inductance current. If the capacitance voltage is assumed to be constant, then the charge on the capacitor will not change over a period of time, that is to say, the charge through the inductor is all passed through the resistor, so that their current is the same, i.e.il=iout=uout/r=d*uin/r, so we know what the average inductor current is related to.

3, in this way, we also need to know how muchil is said just now ? Just said,il linear slope is known, thenil= slope * time can be, soIl=uin (1-d ) *d*ts/2l.

4. In the above1, we assume that the capacitance voltage is constant, so there is a linear change in the current. We said that this is not accurate, because the inductor current in the change, but also assume that the resistance current is unchanged, obviously self-contradictory. This has been avoided in almost all books, some regret. Temporarily endure this contradiction, the next consideration of a period of capacitance voltage changes, because the capacitance may have a part of the inductor current in the input, and itself in the external discharge, the specific relationship can be listed by differential equations, but in general, ∫I (t) dt=0is inevitable. In some places this is called the second balance. This is, of course, derived from the relationship between the capacitance voltage and the current. As we have just said, the assumptions we make are self-contradictory, but we sometimes have to accept a certain setting. For example, we accept that the capacitance voltage remains constant to discuss the variation of the inductor current, or that we accept that the current is linearly variable to discuss the change in the capacitance voltage. If the former, this time,Il-iout=ic, soICis the amount of a straight line lift. If the latter, we first recognizeICis the amount of a straight-line lift, this timeUCis toICof points. So,UCThere is also a process of ascending and descending, andillike, there's a ripple.u, here we can't and just calculateilso come on, because it's not straight up. However, the total charge change can be achieved by the pressure process.q=c*2△uto calculate, we have just said that the inductor current to remove the load resistor current is charged to the capacitor, soICGreater than0the part is the charging current,ICin theTThe axis is going back and forth (becauseTstotal points in time are0or byIl-iout=icfigure), there areq=1/2*△IL*TS/2 (calculating integral by area method). So we can identify theu=△il*ts/8c=uin* (1-d) *d*ts^2/16l*c. The amount needed is more visible.

5. After this, it seems to be aboutBuckall the quantities of the circuit are determined. This is not complete, because there is another way of working is called discontinuous conduction mode .DCM (discontinuous conduction mode). Because of the presence of a diode, although called a freewheeling diode, it also requires that the current cannot be reversed. Ilfirst rise and then drop, may appear to be reduced to0Critical state (discussion of criticality is due to the critical ease of discussion). Here we need toDchange it, please.D1is the conduction ratio,D2is the non-conduction ratio,D3is the rest of the time than the cutoff. byi<△ilthe relationship of discontinuous conduction can be derived:

1-d is represented as Kcrit (d), meaning a function of D , such as the 1-dhere.

The left side is defined as K, i.e. k<kcrit (D) is not continuously conduction. Of course, this relationship can also be expressed in terms of R , so that the physical meaning is more specific, that is, the change in load R caused by the conduction or not mode change.

6, intermittent time, just now we derive the physical relationship seems to be modified. However, the basic idea is the same.

For example, the relationship between Volt-SEC balance voltage,uout=uin*d1/(D1+D2)

How much is D2? We don't know, and we need an equation. Can be obtained by means of the average inductance current il=ir=uout/r .

Finally get a voltage conversion ratio M (d,k) =uout/uin=

so m=d (k>kcrit) or m= (k<kcrit) Obviously, if k ≥ 1 d If it is less than 1 d makes kcrit greater than k

1, the above about buck Span style= "Font-family:calibri" >buck What is the nature of the buck circuit.

uin*i*ts*d Span style= "Font-family:calibri" >i is the average current during conduction, and the average current of the inductor il equals, load consumes energy within one cycle is uout*iout*ts iout=il uout=d*uin ts*d Span style= "Font-family:calibri" >i ts d times the input

2, basically, for buck The analytical thinking of the circuit can be applied to boost buck-boost circuit. For example: boost The essence of the circuit is to ts energy in ts (1- D) uin*i*ts=uout*i*ts* (1-d) so get uout=uin/(1-d)

3, Dc The number of circuit schematics (or topologies) of transformations is fixed? Which means the number of topologies is limited? Consider the simplest case of an inductor, a capacitor, a switch, a diode. According to sanjaya &NBSP; maniktala proficient in switching power supply, the switch tube, inductor and freewheeling diode between different positions to determine the different functions. It can be inferred that three devices are arranged in a (3,3) =6 4 buck buck-boost each one, boost two.

4, why has the perfect buck Continuous conduction mode, but also to consider the discontinuous conduction mode? Existence is reasonable. When light load or no load, il very small (R " big ") ) il il big ( small ) d Span style= "Font-family:calibri" >r co-determined.

Thinking and analysis of DC transform Beginners