RF amp with BJT single-stage

[[email protected]]

Hi,
I'm designing a simple single-stage amp through a BJT in common-emitter config.

On several reference designs observed in the internet, I see that a resistor is used between collector and V_CC, whereas for other designs, an RF choke is used instead.

I'm getting the idea that the resistor based designs implement a class A amp, for linear (AM) amplification, whereas the RF choke is for class C, narrowband amplification.

Is this right?

Thanks
RM

 

[[email protected]]

No

Can you tell me more about? Or suggest clarification by my side?
R

 

[Piotr Wyderski]

I'm getting the idea that the resistor based designs implement a class A amp, for linear (AM) amplification, whereas the RF choke is for class C, narrowband amplification.

IMHO the choke is there only for noise isolation purposes. It prevents
your amplified RF to sneak into the power rail and vice versa, from the
power rail to the collector circuit. If it is not a tank circuit, then
only its impedance matters, i.e. no resonant amplification etc. I think
both of your circuits are class A.

BTW, why don't you use a MAR? Much easier to get a decently working
design.

Best regards, Piotr

 

[riccardo manfrin]

IMHO the choke is there only for noise isolation purposes. It prevents
your amplified RF to sneak into the power rail and vice versa, from the
power rail to the collector circuit. If it is not a tank circuit, then
only its impedance matters, i.e. no resonant amplification etc. I think
both of your circuits are class A.

Hum.. but the choke, in DC analysis, moves the collector voltage to be equivalent to V_CC. Isn't this like moving the quiescent point very close to the cut-off region? And isn't this going to mean that if I feed with a signal, only about half of it is going to be amplified, which is exactly the definition of class C?

At least this was my understanding from Figure 1.21 of this document:
http://www.badeclasses.com/Study Materials\EDC 01.pdf
(please take the burden).

BTW, why don't you use a MAR? Much easier to get a decently working
design.

If you mean using an IC, the motivation behind the choice of discrete components is exacly for the same reason why I'm posing this question: because I'm ignorant about the subject and would like to understand more about it (and to this extent discrete design is an easier starting point).

 

[Piotr Wyderski]

riccardo manfrin pisze:

Hum.. but the choke, in DC analysis, moves the collector voltage to be equivalent to V_CC

You mean ONLY a choke? Then indeed, it's not class A.
I thought it was a resistor in series with a choke.
But class C in an input amplifier? C'mon... :-/

If you mean using an IC

Yes, a 4 pin ready-made RF amplifier, e.g. MAR-6.

the motivation behind the choice of discrete components is exacly
for the same reason why I'm posing this question: because I'm ignorant
about the subject and would like to understand more about it
OK.

(and to this extent discrete design is an easier starting point).

But may be bitchy to do it properly.

Best regards, Piotr

 

[riccardo manfrin]

Il giorno venerdì 20 settembre 2013 11:30:20 UTC+2, Piotr Wyderski ha scritto:

riccardo manfrin pisze:






You mean ONLY a choke? Then indeed, it's not class A.
I thought it was a resistor in series with a choke.

This is a relief. I thought I was getting it all the way wrong.

But class C in an input amplifier? C'mon... :-/

If you are posing the doubt that such amplifier cannot amplify AM stuff (audio), well, I totally agree with you! This is going to be for strictly narrowband input signals, high efficiency amplification, as by definition for Class C.

But may be bitchy to do it properly.

I expect so

Thanks for giving me confirmation!
R

 

[Tim Williams]

You mean ONLY a choke? Then indeed, it's not class A.
I thought it was a resistor in series with a choke.
But class C in an input amplifier? C'mon... :-/

Nah, you can have whatever bias current you want. Toobz have been doing
it for about a century (which... these days, I think that's actually
true?). Emitter bias doesn't simply go away.

What it does is give you extra VCC for free, doubling your voltage swing
for no cost in power. Theoretical maximum efficiency jumps from an
embarassing 12.5% to a passable 25% (assuming continuous current, i.e.,
class A operation).

Beware the extra gain you get from a choke load; obviously, whatever load
is attached will load it down, but if you're using a light load (say,
preamp, IF stage, etc.), just something to beware of.

Tim

 

[Phil Allison]

"riccardo manfrin"

Hum.. but the choke, in DC analysis, moves the collector voltage to be
equivalent to V_CC. Isn't this like moving the quiescent point very close to
the cut-off region?

** No.

And isn't this going to mean that if I feed with a signal, only about half
of it is going to be amplified, which is exactly the definition of class C?


** No.

Using a choke allows the signal to swing ABOVE the DC rail - IOW it
doubles the available swing.

FFS - read up on class A valve stages that use coupling transformers.




.... Phil

 

[Phil Allison]

"riccardo manfrin"

Thanks for giving me confirmation!

** Classic usenet scenario.

One damn fool mis-informs others.



.... Phil

 

[Phil Allison]

"Tim Williams"

What it does is give you extra VCC for free, doubling your voltage swing
for no cost in power. Theoretical maximum efficiency jumps from an
embarassing 12.5% to a passable 25% (assuming continuous current, i.e.,
class A operation).

** A choke loaded class A stage has a max efficiency of 50%.

Same goes for a transformer coupled one, or a push-pull one.

Your pig ignorance is showing, again and again.

Fucking Google monkey.



.... Phil

 

[miso]

The CE amp needs some impedance in series with the collector to produce
gain, basically I to V conversion. Gain is proportional to the series
impedance.

With a resistor, as you get more gain by increasing the resistance, the
voltage at the collector drops. Basic Ohms law. But with an inductive
load, you can have a large impedance at the operating frequency without
the associate voltage drop of a resistor. You can get away with this
because the RF amplifier is presumably narrow band.

Ideal inductors don't have thermal noise but ideal resistors do.

Either way, the amp is class A.

 

[Anthony Stewart]

I agree with your technical comment but your personal remarks only make You look bad.

With a resonant cap, the tank circuit will have an additional voltage gain determined by loading ratio , Q or the resistive impedance to scalar value of reactive impedance of choke (which equals that of cap at resonant frequency)

Q also defines the 3 dB bandwidth of center freq, Q = f(c) / f(BW). Self resonant frequency of choke must >> desired resonant freq with load cap.

 

[Simon S Aysdie]

You can get away with this

because the RF amplifier is presumably narrow band.

It doesn't need to be "narrowband." The choke simply needs to be high in impedance relative to the load. For example, if a conical inductor is used,then it can be somewhat widebanded, where the limitations are in the transistor and other elements, but not so much the choke.