RF Power Amplifier

Depends.

Bipolar? FET? Triode? Tetrode? Pentode? TWT? Klystron? IMPATT? Class A? Class AB? Class B? Class C? Class D? Class E? Narrow band (tuned)? Broadband? Distributed?

I am looking for a qualitative difference between voltage amplifier and power amplifier.

Very broadly speaking, usually an RF amp only is designed to work around a certain center frequency, whereas a "voltage amplifier" is probably some broadband device with an overall low pass characteristic.

Hard to say, but again, very broadly speaking, I'd say first you decide if you need the amp to be linear or not (this depends on the modulation scheme you're using -- AM needs linear amps, for instance, FM doesn't). If it has to be linear, you usually concentrate of linearity, of course -- and until recently, you had usually given up power efficiency at the point you said you wanted a linear amp! For non-linear amps, power efficiency (without making the thing TOO distorted) is often the goal.

In some RF amps, phase is important, in others it isn't. It's impossible to make many generalizations without specifying how you plan to modulate that RF.

---Joel Kolstad

As people have mentioned you tend to work into lower impedance loads with power amps but also another very big point.

When you start doing power amplifiers you're much more liking to have to worry about thermal dissipation of the amp and methods for getting the heat off the chip (or discretes).

Not that it can't be a worry with voltage amplifiers as well.

Robert

I suspect you may be more interest in the difference between say an audio amplifier and an RF amplifier.

It's more relevant to compare a voltage amplifier with *any* power amplifier.

A power amplifier will amplify the voltage at its input and also supply lots of current at the load. A purely voltage amplifier is not designed to deliver much current - hence the term *voltage* amplifier, although it may actually have quite high current gain too, it isn't designed to transfer large powers to its load.

Graham

Indeed ! There are very many factors to consider.

The original question was too loosely worded. If the poster would give a closer indication of his/her area of interest we could give a better response.

e.g. is the RF amplifier intended to be a transmitter ? If so in which RF band and what modulation method.

Graham

The difference can be put as follows:

IN a "voltage amplifier" one strive to increase the voltage to a given load, and therefore the OUTUT Impedance is purposefully low.

IN an RF amplifier, one strives to send as much power to the load, and therefore the output impedance is MATCHED to the load. Further, as the frequency is high, the wavenumber gets smaller, and distances begin to be characterized as transmission lines. This further, indicates a focus on Impedance matching the amplifier to the load, transmission structures inclusive.

Need One?

Marco

Indeed. High-voltage and high-frequency voltage amplifiers often have very difficult tasks, more difficult than power amplifiers. This is because they are often working into capacitive loads, and often at far higher voltages than you'd find from out of a typical power amplifier.

For example, consider an RF power amplifier. A common design goal is to optimize it for a 50-ohm output line (an impedance-matching network may come after the amplifier). Let's say it's a 500W linear amplifier. It's output voltage won't be very high, only 158 Vrms. Furthermore, it's operating in a relaxed environment, where the output current is in phase with the voltage, so the current is highest at the same time as the voltage, which means the power dissipation across the active elements is relatively at a minimum.

By contrast a voltage amplifier working at the same frequency into a capacitive load won't be delivering measurable power (the current is not in phase with the voltage), but yet it may be working MUCH harder because its output load current is highest when the output voltage is near zero, midway between the supply rails (that's where the slew rate dV/dt is the highest), which can result in very high internal power dissipation. Furthermore, the compensation and stabilization of the amplifier is more difficult into a capacitive load. In addition, to place the scene between voltage and power amplifiers in perspective, it's common to require much higher voltage out of voltage amplifiers. The 158V mentioned above doesn't get one far when working with electric fields. Whether working with piezo-electric elements, ion deflection electrodes, particle traps, etc., it's common to need hundreds or even thousands of volts to do the job.

For example, last year I made a 10kV 600kHz amplifier. If this was a commonplace 50-ohm matched-load amplifier it would have required 1 mega-watt of raw power output. So my task was to accomplish the same thing a 1MW RF amplifier could do, but without spending 1MW to do it.

It's safe to say those of us who design voltage amplifiers have plenty to worry about. :>)

The RF amplifier has a defined input and output impedance, usually 50 Ohms. Then there are narrowband (tuned) or broadband amplifiers. To some of them, noise is also an important factor.

Rene

Sigh. As a rule, RF power amplifiers are designed to drive 50 Ohm loads. That says nothing at all about their actual output impedance. In fact, the impedance looking back into the amplifier output is all over the place.

In a nutshell, the usual way to design RF power amplifiers is to find the load resistance that would dissipate the target power level at a given combination of power supply voltage and output device. A transformer then transforms the actual load impedance to this previously determined value. That's what 'matching' means here.

There used to be a number of Motorola application notes written by Helge Granberg that spelled this out in a very clear way.

Jeroen Belleman

"Winfield Hill" a écrit dans le message de news: snipped-for-privacy@drn.newsguy.com...

Weew, that's impressive. Not for the faint heart, to say the least.

Was it really without output matching? For a linear amplifier into a 10pF capacitive load, no matching, and neglecting the circuits parasitics which obviously can't be at this level, that's already a healthy 380mA pk current and an average power dissipation for the output stages of 2.4kW !!! Certainely not manageable without an output transformer. And a transformer will require an impractical high turn ratio at this frequency and power level.

Or is it the 600kHz resonnant converter that you spoke about some time ago? Which is certainely not obvious too.

Some hard core details?

Only plenty?

ALL amplifiers are power amplifiers. More power comes out than what goes in. The name of an amplifier depends on what sort of use is made of the output. It may be a current or a voltage amplifier. Usually it's nameless.

The name "power amplifier" usually applies to a circuit specifically designed to deliver a certain power to a given load resistance. The higher power output stages of Hi-fi audio (loud speaker) amplifiers and radio tramsmitters (from 0.1watts to 500,000 watts) are always known as the PA.

In the case of higher power radio transmitters the input power to the PA is also specified and the stage which drives it is also a power amplifier known as the "Driver".

Power amplifiers usually have some sort of transformer between the output device (transistor or tube) and the external load. This is to ensure the output device works into its correct load resistance. An incorrect resistance, too high or to low, results in distortion and loss of power in what is being amplified.

Quite. There is often some reactive component to the load that needs to be compensated out by the matching network. Without that matching, there will be either inductance or capacitance 'seen' by the amp's output in addition to the purely resistive component. Eliminating the reactive component and ensuring the correct resitive should result in a happy amp, all other things being equal.

Paul Burridge (posting from sunny Turkey):-)

This tells you *only* that the voltage is 158Vrms at the point at which the (possibly complex) output impedance of the amplifier has been transformed (matched) to 50 Ohm and is loaded with 50 Ohm. It tells you nothing about the nature of the generator voltage - it could be bigger or smaller.

Again, this is ONLY true at the point at which the output impedance has been transformed to 50Ohm. If the unmatched o/p impedance of the amp has a reactive component (which will disappear *after* the o/p match), then that same match will transform the 50Ohm load into a complex load.