PowerAnts

maychang posted on 2022-4-12 20:55 There is a source of error in RC charging and discharging, that is, charging and discharging do not start from the amplitude or zero value, but from the ripple valley value and peak value, and charge to...

From a proportional point of view, the ripple voltage is usually required to be less than 100 times the PWM level, and the capacitor is regarded as a constant voltage source. If even this 1% calculation error is not acceptable, how can we be sure that the waveform can be accurately superimposed?

PowerAnts

To minimize the error, we can only use the method on the 18th floor, rather than using the language

maychang

PowerAnts posted on 2022-4-12 21:03 To minimize the error, we can only use the method on the 18th floor, not the language

The method on the 18th floor seems to be calculated using Excel, and it requires multiple iterations. I don't know if it is calculated this way.

In addition, the method on the 18th floor is not the same as what you said on the 26th floor: "the curve segment can be regarded as a line segment, and the ripple voltage can be regarded as a triangular wave."

PowerAnts

maychang posted on 2022-4-12 21:20 The method on the 18th floor seems to be calculated using Excel, and it needs to be iterated many times. I don't know if it is calculated in this way. In addition, the method on the 18th floor is not...

The 18th floor is trying to solve it through RC charge and discharge function, which belongs to Sunshine Avenue and is in the right direction

PowerAnts

Personal summary, for RC filtering:

1. Sinusoidal circuit, capacitive reactance + Ohm's law, OK!

2. PWM smoothing filter with low ripple rate output, constant voltage charging and discharging of R to C, simple and convenient charge balance, and the accuracy of calculation results is better than the device accuracy, which meets the requirement of no adjustment!

3. In rectangular wave circuit, the RC turning point is several to dozens of times away from the PWM frequency, and the charge and discharge curve is no longer straight. Only by using the exponential function can the peak-to-peak voltage on the capacitor be correctly calculated.

PowerAnts

It's a bit exaggerated. The green line looks like this. Is this a test of Ohm's law?

maychang

PowerAnts posted on 2022-4-12 22:11 It's a bit exaggerated, the green line is like this, is Ohm's law playing a trick?

"That's a bit exaggerated. The green line is like this. Are you trying to use Ohm's law?"

The green line already includes Ohm's law.

Whether it is first-order RC or RL, the equations are based on Kirchhoff, such as the blue line in the figure below, and Kirchhoff derived it based on Ohm's law. Then the red line is simply Ohm's law.

PowerAnts

maychang posted on 2022-4-13 17:46 "It's a bit exaggerated. The green line is like this. Ohm's law is used for this?" The green line already includes Ohm's law. First-order RC is also good, ...

What about capacitance?

maychang

PowerAnts posted on 2022-4-13 18:22 What about capacitance?

The capacitive reactance is actually in the equation behind the red line: i=C*duc/dt. This equation is the definition of capacitance and also shows the capacitive reactance.

PowerAnts

You have to go to XX to find HXX to reason with him. On June 13, 2011, I said in a post:

According to Ohm's law, U=IR+Q/C

In a few minutes, HXX posted several posts like a mad dog saying that the above formula is not Ohm's law. I was so angry that I ate two bowls of braised pork to vent my anger.

PowerAnts

If you are interested, you can summarize the situation in your post and compete with HXX to see if he is willing to use Fourier to push up two or three pages.

maychang

PowerAnts posted on 2022-4-13 19:36 For this, you have to go to XX to find HXX to reason. On June 13, 2011, I said in a post: According to Ohm's law, U=IR+Q/C. In a few minutes...

"I was so angry that I had to eat two bowls of braised pork to vent my anger."

Ha! Two bowls of braised pork!

PowerAnts

maychang posted on 2022-4-13 20:01 "I was so angry that I had to eat two bowls of braised pork to get rid of my anger." Ha! Two bowls of braised pork!

Here are my two screenshots:

PowerAnts

Speaking of braised pork, half a year ago the doctor told me that I must start taking Baixintong for life, but I refused because I knew there were other ways. I only controlled my salt intake and my blood pressure dropped from 160/110 to 120/80. I can still eat braised pork, twice-cooked pork and eggs as usual.

In the past two years, I used to feel dizzy when I was having sex, but now I don't feel dizzy at all.

Few people know that if a person survives by eating only cabbage and boiled noodles in water (without adding anything else), his sodium intake will exceed the standard by twice.

PowerAnts

This is also blocked! ?

n0si9na1

I'm here to study too.

Ming8686

XC=1/(ωC), when the frequency is very low, the capacitive reactance of the capacitor is large, equivalent to an open circuit, UI is approximately equal to UO, when the frequency is very high, the capacitive reactance of the capacitor is very small, equivalent to a short circuit, UO≈0.

PowerAnts

Ming8686 Published on 2022-4-15 14:12 XC=1/(ωC), when the frequency is very low, the capacitive reactance of the capacitor is large, which is equivalent to an open circuit, and UI is approximately equal to UO. When the frequency is very high, the capacitor's...

In another place, Mr. Yao from STC gave some suggestions. They did this

Without evaluating the same second-order cascade, it is worth learning how to solve the problem of voltage floating under load by feeding back to the ADC closed-loop output.

PowerAnts

This is the benefit of interaction. Without benefit, it cannot be sustainable! A sudden enlightenment is worth 1,000 gold coins. I would rather spend my own money to reward

不亦心

I guess I am still confused by what you guys are discussing, as you haven't solved his question about Vout=VCC*Duty.

I'll give you something that LZ can understand.

Only look at the steady state, the steady-state capacitor charge balance (charging equals discharging)

(VCC-Vout)*Duty/R=Vout*(1-Duty)/R

So Vout=VCC*Duty

As for transient state, details, and trial conditions, see the discussion above.