Some questions about class D power amplifier

[hanhan]

Hi,
Please help me with some questions about class D power amplifier? Thanks.

 

[ronsimpson]

1) Cb is part of the power supply. It is to keep RF voltages off the supply voltage (Vcc). At RF it should be a 'short'. It is a large cap that will have near 0 ohms at RF.
2) The Vc-e is small. The transistor is not linear. Linear= increase base current and collector voltage changes about the same amount. Saturation= base current change causes a very small change in Vc-e.
3) Fig a. On the base is a square wave signal source. That is the source of RF.

 

[hanhan]

Thanks ronsimpson,
Sorry, I don't get what you meant.

1) Cb is part of the power supply. It is to keep RF voltages off the supply voltage (Vcc). At RF it should be a 'short'. It is a large cap that will have near 0 ohms at RF.

Can you explain how this capacitor is used to keep RF voltages off the supply voltage (Vcc)?
I can't see anything useful of this capacitor.
At DC bias, this capacitor is open circuit => It plays no role.
At RF signal, this capacitor is short circuit and connected to ground, but Vcc also be connected to ground. => Then it seems that the capacitor again is not useful at all.

2) The Vc-e is small. The transistor is not linear. Linear= increase base current and collector voltage changes about the same amount. Saturation= base current change causes a very small change in Vc-e.

I think I know these notions but I don't know how the transistor is in saturation mode in this case.
In the picture, the input voltage Vin is a square wave.
When Vin = 0 => Vbe = 0 => Transistor is in CUT-OFF mode.
When Vin = Vin(max) => How do you know the transistor is in SATURATION in this case?

3) Fig a. On the base is a square wave signal source. That is the source of RF.

It seems strange to me. Is this ONLY a very special case of RF signal?

 

[ronsimpson]

1) what is the purpose of a capacitor on the power supply?
a.)Most power supplies respond very slow. It can not supply energy at radio frequencies.
b.)If no capacitor. The wire going back to the power supply will act as another inductor on the collector.
c.)If no cap. RF energy will radiate from the wire going to the power supply.
d.)At DC this cap is 'open'. At RF it is like a short.
e.)This cap is where the amplifier gets its energy from. Not the power supply. The power supply charges up the cap. The cap gives power to the amplifier.
2,3)Saturation:
a.) A linear amplifier has about 1/2 the supply voltage across the transistor and about 1/2 the max current in the transistor at the same time.
power loss=VxI example 6V x 0.5A
b.)In a saturated amplifier, the transistor is open (power = 12volts X 0 amps) OR closed (0.3V X 1 amp) because 0.3 is a small number the power loss in the transistor is small.
c.) When driven into saturation the transistor is not as hot.
d.) With a oscilloscope you can see the Vc-e and increase/decrease the base current until you have good saturation.

There are class D audio amplifiers (not made like this but with the same idea). I have used this idea in both audio and RF amplifiers. And in video amplifiers with out gray-scale.

The biggest amplifier I build was a 10,000. watt FM transmitter. I have used amplifiers up to 30K watts. This is the output power. There is a great desire to not wast power in the amplifier.

 

[MrAl]

Hi,

Actually the capacitor Cb would be positioned close to the transistor so that it can shunt any high frequency AC to ground as close to the switching transistor as possible. This prevents radiation from the power supply leads as well as prevents AC from modulating the DC power line. If the DC line gets modulated it also causes AC feedback which is very bad for the circuit.

And actually this capacitor would be a small ceramic in combination with a larger cap. The larger cap is there for more slowly changing signals and the smaller cap is there for the higher frequencies because it can react better than the larger cap. The two would be in parallel, and this will bring up another question from you i would think such as, "why have a small cap (like 0.1uf) in parallel with a large cap (like 100uf) when the total combined capacitance is only a tiny bit more with the extra small cap". The answer is that the small cap can handle the higher frequencies better but we still need to handle the lower frequency components too so we use the larger cap with it.

Your question again has a lot to do with the things that are not always drawn in the schematic but are always present. For one, the power supply source voltage can not get to the inductor through a zero length lead line. The lead must have some length, and that means it has both resistance and inductance. So there is a small inductance and resistance in series with the inductor that is actually drawn. This works in combination with the Cb cap to keep noise off of the power supply, and there would be another Cb cap located close to the power supply too.

 

[rumpfy]

Anh,
Sketch a is the circuit diagram but Sketch b is the 'equivalent circuit'.
Note that in sketch b you show a capacitor Cb next to Rl but this should in fact be Co next to Rl. Cb shown in sketch a should be shown in Sketch b as being located in parallel with Vcc.
In a previous post, you asked about the use of the RFC (Radio Frequency Choke). This device was designed to be short circuit to DC power and open circuit to the signal frequencies. In sketch b, you show the power supply(Vcc), and I'm saying that Cb should be in parallel with the power supply Vcc. In sketch b, when the switch is closed, there is a current I, shown, flowing through Rsat. This current actually flows back to the power supply Vcc. The internal impedance of the power supply is generally not well defined for anything except DC current and it is good practice to ALWAYS shunt the power supply with an impedance which is open circuit for DC but short circuit for AC. Such a device which can do that, is a capacitor. That is the purpose of the capacitor Cb. As you will known from ohms law, a zero impedance will have zero volts across it, regardless of the magnitude of the current flowing through it. By use of the capacitor across Vcc, we can hold the signal voltage developed across Vcc at a zero value even thought he signal current could be large. This answers Q1.

Q2. The sketch b shows the resistance r sat. This is effectively the conduction impedance in the transistor. The value of rsat is dependent on the collector current AND the amount of base drive current. For pulse type applications (such as this), the base drive current is expressed as being either Ic/5 or Ic/10. If you look at device data sheets for switching transistors you will see what I mean. It is normally the case that in the engineering design phase, one looks closely at the power dissipation in the transistor. If the design under drives the base circuit, this can lead to excessive power dissipation in the collector circuit; and by increasing the power dissipation in the base circuit, this lead to lower total dissipation in the collector circuit. Thus, base drive arrangements are very important in establishing control on the collector dissipation.

Q3. The sketch is of a switching circuit. The switching frequency does not really matter from the point of view of understanding the DC and AC signal paths in the circuit. Such an arrangement as your sketch is used frequently at switching frequencies of 300 to 500 or 1000 kHz. A typical device could be a small welder, or a industrial eddy current heater for heating metal parts in an industrial process.

Hope this helps.