High frequency power resonant amplifier, how to adjust the optimal impedance

Seven_ZGY

High frequency power resonant amplifier, how to adjust the optimal impedance [Copy link]

 

I got a circuit and applied it, but I don't know its principle. The triode works in Class C state. The requirement is to change L2, L3, L4, C4, C7, C8, or the transformer ratio, and require R10 to obtain the maximum power when the load is 500 ohms, that is, the power increases from 100 to 500 ohms, and the power decreases from 500 to 3000 ohms. The problem is that I don't know how to determine this maximum power load. For example, I want 400 ohms to have the maximum power, or 600 ohms to have the maximum power. How should I adjust the parameters? In my humble opinion, the signal is turned on by the triode half-wave or less than half-wave, and a pulse excitation current is generated in the primary of the transformer. It is coupled through the transformer and then passed through the LC resonant network to resonate the irregular signal into a full-wave sine signal. I analyzed that in addition to resonance, LC also has the function of impedance matching, but how to match the impedance and how to analyze it, please give me some advice. . Or my analysis is wrong, please correct me.

PowerAnts

The high-frequency circuit part of the university's "Linear Circuit" has impedance transformation. You can use a special impedance transformation circuit or let the filter itself add impedance transformation function.

Seven_ZGY

PowerAnts published on 2020-10-26 20:51 The high-frequency circuit part of the university's "Linear Circuit" has impedance transformation. You can use a special impedance transformation circuit, or you can add impedance transformation function to the filter itself...

Is it the linear part of "Electronic Circuits" or just called linear circuits? I can't find this book on Baidu.

PowerAnts

Seven_ZGY posted on 2020-10-27 08:57 Is it the linear part of "Electronic Circuits" or just called linear circuits? I can't find this book on Baidu

High frequency circuits are divided into linear and nonlinear. Impedance transformation is in the linear part, which is the case in the image. Aren't linear electronic circuits linear circuits? Are there also linear non-electronic circuits?

gmchen

L2, L3, L4, C4, C7, C8 are low-pass impedance matching networks, which transform the load R10 to the secondary of T2, and then T2 transforms this impedance to the transistor collector. Except for R10, the remaining resistors and capacitors can be ignored during analysis. RC impedance transformation networks are usually found in textbooks on high-frequency circuits or linear circuits. Smith charts are also often used in actual design.

gmchen

The circuit of the first building is a class C high frequency amplifier.

The characteristic of a Class C amplifier is that its output waveform is only a part of the sine wave. The subsequent LC network takes out the fundamental frequency component and sends it to the load. It also transforms the load impedance to the transistor collector to become a transistor equivalent load.

The original poster's questions can be divided into the following: 1. What is the relationship between the change of load resistance and output power?

Class C amplifiers can operate in three different regions: overvoltage region, critical region and undervoltage region.

When other conditions remain unchanged, the equivalent load impedance seen at the collector of the transistor becomes smaller, and the operating state moves toward the undervoltage state; the load impedance becomes larger, and the operating state moves toward the overvoltage state.

When the working state moves toward the undervoltage state, the peak value of the collector output voltage (that is, the fundamental voltage peak value) decreases rapidly, but the collector current changes very little, so the output power decreases rapidly.

When the working state moves toward the overvoltage state, the peak value of the collector output voltage changes very little, which is approximately a constant voltage source, but the collector current becomes smaller and a top depression appears, causing the fundamental frequency component to become smaller and the output power to decrease rapidly.

In summary, the maximum output power can be obtained in the critical area, and the output power drops rapidly when working in the other two areas.

gmchen

The second point of the original poster's question is: How to determine the critical load resistance?

First, let's look at the relationship between the load resistance of a Class C amplifier and the transistor output characteristics. The following figure is a simplified schematic diagram:

The red line in the figure is the collector equivalent load line in the critical state, and the two blue lines are the load lines in the overvoltage state (large re) and undervoltage state (small re). Obviously, the load line is related to the size of the collector equivalent load, the beita of the transistor and the base input voltage Vbm, and the static operating point of the transistor (affecting the starting point position of the lower right corner of the load line), which are the necessary conditions for determining the critical load.

gmchen

The third point of the original poster's question is: How to transform the load resistance to the collector?

I have already answered this question on the 5th floor. The specific process is a bit complicated, so I will try to introduce it below.

Since the output of the Class C amplifier has many higher harmonics in addition to the fundamental frequency, the impedance transformation network behind it is usually always a low-pass LC network. The transformation principle of this network is as follows:

The RC parallel network is transformed into an RC series network, RL is transformed into Re _, C is transformed into C' , L and C' must resonate at the operating frequency f0 , _and its resonant impedance is ₀ , and Re _is the equivalent resistance seen from the left end of the network.

The transformation formula is as follows:

Q=RL*(2*pi*f0*C)

C'=C*(1+1/Q^2)

Re=RL/(1+Q^2)

gmchen

There are three LC sections in the original poster's circuit. Starting from the rightmost section, use the formula in the previous post to calculate the resistance Re equivalent to the left side of LC. Then use this resistance as the load of the second section, and repeat the above calculation until all three sections are calculated, and you will get the equivalent load resistance seen on the secondary side of the transformer. Then use the square of the transformer's turns ratio to calculate the equivalent resistance of its primary side, which is the equivalent resistance of the load resistance equivalent to the collector of the transistor.

The above calculation does not include LC losses, and loss resistance may also need to be included depending on actual conditions.

gmchen

If the calculated resonant frequency of C' and L is not equal to the operating frequency, it means that the impedance equivalent to the left side of LC is not a pure resistor, and contains reactance components (that is, the remaining part of the resonant LC). This reactance component can be merged into the previous LC section for calculation. If there is reactance in the last section, it means that the equivalent collector load has reactance. This is usually a normal phenomenon, because the output impedance of the transistor also contains reactance, and external reactance is required to match it.

gmchen

One more thing to add. When you encounter a matching network with reactance like the one on the 10th post, it is troublesome and not intuitive to use the formula on the 8th post. In this case, it is very simple and intuitive to use the Smith chart for analysis. It is not appropriate to discuss the Smith chart in this reply. You can find a high-frequency or radio frequency textbook for reference.

gmchen

I roughly estimated the circuit of the OP, the equivalent load of the transistor collector is about 3 ohms, and the maximum output power is about 70W~80W. This is an estimated value, I wonder if the actual value of the OP is consistent with it?

Seven_ZGY

gmchen posted on 2020-11-19 13:01 After roughly estimating the original poster's circuit, the transistor collector equivalent load is about 3 ohms, and the maximum output power is about 70W~80W. This is an estimate...

Almost. I have another question. If the collector equivalent load is 3 ohms, how can we estimate the output power to be around 80W?

gmchen

Seven_ZGY published on 2020-12-2 15:20 It's about the same. I have another question. If the collector equivalent load is 3 ohms, how can I estimate that the output power is about 80w?

1. The base bias voltage of the transistor is +0.7V, so it can be judged that the working state of the transistor is approximately Class B, or its conduction angle is approximately 90 degrees.

2. The collector voltage used in the circuit is 50V. When the conduction angle is 90 degrees, the peak value of the output collector voltage (peaked cosine pulse) is approximately 50V minus the saturation voltage drop of the transistor (assuming that the input voltage amplitude is sufficient to make the transistor work in a critical state). Assuming that the saturation voltage drop of the transistor is 2V, the peak value of the output peaked cosine pulse voltage is 48V. Based on this, the peak value of the collector current (also the peaked cosine pulse) is 48/3=16A.

3. Decompose the sub-peak cosine pulse. When the conduction angle is 90 degrees, the coefficient of the fundamental frequency component is 0.5, so the current peak value of the fundamental frequency component is 8A and the effective value is 5.6A.

4. The fundamental frequency power on a 3 ohm load is P=[(5.6)^2]*3=94W, but losses must be considered, including the fact that the transistor may not be working in the best condition, the conduction angle may be less than 90 degrees, and the loss of the subsequent impedance matching network, etc. Multiplying it by a coefficient of 0.8 (this is a very subjective judgment) is about 80W.

gmchen

These estimates are made based on the circuit diagram and have a large margin of variation, but the approximate numbers should be correct. The experimental results of the original poster confirm the estimate and prove that the circuit's working state, conduction angle and other parameters are not much different from the estimate.

gmchen

The analysis on the 14th post is, in turn, the basis for designing power amplifiers. The questions raised by the original poster at the beginning of this post can also be answered from it.

Seven_ZGY

gmchen posted on 2020-12-2 21:00 These estimates are based on the circuit diagram and have a large margin of variation, but the approximate number should not be bad. The experimental results of the original poster confirm this...

Yeah, yes, thank you so much.