Current Mode View

MrCU204

Current Mode View [Copy link]

 

 

In electrical engineering and physics classes, there are two types of power sources: current sources and voltage sources.

In the field of electronic technology, there is an additional thing called a controlled source. The property of the current loop is actually a controlled source. If there is no supply mechanism, it will only be an ordinary current source!

MrCU204

This common circuit can actually be understood in this way.

R2 and R3 form a feedback chain, R3 is a sampling resistor, and this circuit is a voltage stabilizer. If R2 is used as the mounting point, then R3 is the current sensing resistor, and this circuit becomes a constant current device.

Supplementary content (2023-9-6 23:10): The technology is advanced, but the principle is not profound. The negative leg remains unchanged while the positive leg changes, and it can also be adjusted. The positive leg is the one that is controlled in the current loop.

MrCU204

In a linear system, the voltage divider of R3 is dynamically balanced.

In the switching circuit, R2 becomes an inductor and the voltage drop of R3 is a gradual value. The operation process of the circuit is actually like a multi-harmonic oscillator.

MrCU204

MrCU204 posted on 2023-8-16 00:48 This circuit, which is very common, can actually be understood in this way. R2 and R3 are the feedback chain, R3 is the sampling resistor, and this circuit is a voltage regulator. If R2 is used as the...

By replacing R2R3 with an adjustable resistor, Vout can also be adjusted.

If it is a constant current device, it should be mounted at the R2 position, and R3 is used to detect the current. If R3 is adjustable, the current can be adjusted.

For an adjustable constant current device, if the current sensing resistor changes with the load impedance and is in a fixed ratio, then Vout will not change.

MrCU204

For pure resistive voltage division negative feedback, the most stable point is at the sampling point, without exception!

The voltage drops of R3 of the voltage regulator and Rg of the op amp are only controlled by the reference or signal and are not affected by the voltage divider ratio or their own resistance.

MrCU204

As a constant current device, how much current R2 consumes depends on R3.

If R3 changes in proportion to R2, the effect of this circuit will be like voltage stabilization.

This circuitous voltage stabilization method is not applicable in the linear field, but in the switching system, it can avoid the shortcomings of the voltage loop and improve dynamic indicators. Of course, the current loop of the switching circuit does not work this way.

MrCU204

The purpose of a voltage stabilizer is to imitate a constant voltage source, but the nature of the current loop is constant current.

If this "current source" cannot be controlled, how can this switching power supply maintain a constant voltage? !

But the problem is that the current-sensing resistor in the current loop is an ordinary resistor, which is not variable, let alone controllable.

MrCU204

Whether it is voltage regulation or constant current, if it is to be adjustable, the usual practice is to make R3 adjustable.

However, there is another way, which is to replace the reference. The nature of the current loop is that it is a controlled source, and the control scheme adopted is to change the reference. This "reference" is actually the error information provided by the voltage loop.

MrCU204

There are three constant current devices, 1A, 2A, and 4A respectively.

1A is mounted at 12Ω, 2A is mounted at 6Ω, 4A is mounted at 3Ω,

The result is that the output voltage of the three constant current devices is 12V. If an adjustable constant current device is adjusted according to this example, the output voltage can also be maintained at 12V.

MrCU204

As a voltage regulator, the current through the inductor must be able to change according to the law of the constant voltage source; if Vref and R3 are not adjustable, the function of the current loop cannot be performed.

If R3 is adjustable, it doesn't matter if Vref is fixed. However, in the current loop, the current-sense resistor is an ordinary resistor. To control the current, the voltage drop of the current-sense resistor must be changed. If Vref is not adjusted, there is no room for the voltage drop of the current-sense resistor to change.

MrCU204

MrCU204 posted on 2023-8-23 09:48 As a voltage regulator, the current through the inductor must be able to change according to the law of the constant voltage source; if Vref and R3 are not adjustable, the function of the current loop will be...

The Vref required by the current loop is the error information from the voltage loop.

Assume that the maximum output of this regulator is 10A, and the current loop current sense resistor is 0.1Ω.

Then, the maximum voltage drop of the current-sense resistor is 1V, which means that the error information from the voltage loop must also be as large as 1V in order for the current loop to control the maximum duty cycle.

MrCU204

The only switch in the voltage regulator is firmly occupied by the current loop.

If the current loop is not adjustable, the entire voltage regulator will be blocked by the current loop. How can that work? !

The benchmark is not fixed. It is not written in books or taught by teachers. Therefore, my understanding of the current loop principle is confused at this step.

According to the usual routine, there is one thing that cannot be adjusted in an adjustable power supply, which is the potential of the sampling point. As long as there is no problem with the circuit, when you adjust R3, the only thing that will change is Vout and the current passing through R2, no matter what you do with this topology!

The switch tube of the current mode regulator belongs to the current loop. If the current loop cannot be adjusted, the whole circuit is not a regulator. However, the current sensing resistor of the current loop is just an ordinary resistor, and I didn’t know that changing Vref can also make the circuit adjustable and controllable. This dilemma has finally been circumvented.

MrCU204

It's true that there is a manual adjustment interface placed on the reference side. If the adjustable resistor is replaced with voltage deviation information, this circuit is equivalent to a current loop.

MrCU204

Switching power supplies, except charge pumps, generally have inductors. Even though they are inductors, they behave differently in different modes.

In ripple mode, the system itself is a multi-harmonic oscillator, and the inductor and load form a time-based network as the main shock trough.

In voltage mode, the current is directly dependent on the load, and the inductance is not monitored. The only thing it does is to flatten the PWM wave.

In current mode, the inductor is a component of the current loop, and the current loop is a voltage-controlled constant current device. Without the intervention of the voltage loop, the load will not get a constant current but a voltage regulation.

The current loop can be regarded as a monostable trigger with a variable threshold. Rι is replaced by an inductor and Rf is fixed.

The time base of the components is fixed, and the duty cycle is the duration of the temporary steady state. The higher the "base", the greater the duty cycle.

The current loop's current-sense resistor is only in the Ω level or even the mΩ level, so the time constant obtained by combining it with the inductor is equivalent to the inductor's own charging and demagnetization time.

The load and smoothing capacitor are not in the current loop, and will not create any time constant with the inductor. RC, LC, and LR are all time bases. Without a time base, there will be no zeros and poles. This is probably what the so-called order reduction is all about.

MrCU204

In terms of system principle, the switch tube covering the current loop is equivalent to the adjustment tube of the linear regulator .

The transistor has its own constant current characteristic. The linear voltage regulator uses the dynamic resistance of the transistor.

The switching mode shields the dynamic resistance. When the inductance is controlled due to the intervention of the current loop, it can be equivalent to the dynamic resistance of the transistor to a certain extent.

MrCU204

The so-called order reduction is probably to crush the passband or stopband characteristics into monotonically changing characteristics.

MrCU204

MrCU204

The circuit above, if mounted on R2, is a constant current device.

Remove R4, D1 and R1, change R2 to an inductor, add U3A, and it becomes like this.

MrCU204

The circuit on the 17th floor is applicable to both voltage-stabilized constant-current switches and linear circuits.

This architecture, no matter how you play it, has no distinction between voltage loop and current loop.

Moreover, when used as a switching circuit, it only has a hysteresis mode, which is actually a multi-harmonic oscillator.

The current loop is actually a multi-vibrator converted into a monostable trigger. The circuit on the 18th floor is like this.

Glue it to Q1 in the 17F circuit and the linear regulator becomes a current mode switching regulator.

MrCU204

The hysteresis mode is actually a multivibrator, while the current mode can be regarded as a monostable trigger. It is also an inductor, but in voltage mode it is completely a choke, only focusing on filtering and not caring about the duty cycle. In the ripple (hysteresis) and current modes, the inductor is actually a timing element and has a dominant influence on the current because the current is a function of the duty cycle. The current mode regulator is a "trigger", not an oscillator. The subharmonic oscillation is not self-excited or parasitic oscillation, but the timing element is inaccurate. But this is not the fault of the components or the circuit design problem, but the inherent defect of the peak current mode itself. There was a professional term called the chasing phenomenon in the last century. I think that subharmonic oscillation is actually the chasing phenomenon caused by inherent defects!