EMRFD Message Archive 56
Message Date From Subject 56 2006-08-02 11:40:46 jmilldrum Feedback Amplifier/Attenuator Pad Forgive me if I completely missed this in EMRFD, but I'm having
trouble finding the answer in the book. I have noticed a very common
circuit in the book is the pairing of feedback amplifiers with
attenuator pads on the output. The reason to use a feedback amp is
obvious, but I'm a little uncertain about the addition of the
attenuator pad. I'm assuming that they are there to ensure a good
impedance match with the next stage. Is this correct? Are there other
reasons to use this configuration?
Jason Milldrum, NT7S
58 2006-08-02 11:59:25 Jim Miller Re: Feedback Amplifier/Attenuator Pad feedback amplifiers necessarily reflect changes in the load impedence to
their inputs. the pad presents a relatively stable load impedence into which
to work to ensure that the input impedence remains stable.
in the extreme the lack of stable load impedence could also lead to unwanted
63 2006-08-02 15:46:15 Patrick McGuire Re: Feedback Amplifier/Attenuator Pad Hi Jason - The attenuator is usually specified not to match impedance but to mask impedance mismatch. It doesn't matter whether the amp is a feedback amp or not although it might be more important then. If the amplifier is driving a perfect (or very good) match there is no need for an attenuator.I use them when connecting, say, an inexpensive signal generator to a homebrew bench amplifier and then another between the amplifier and whatever piece of equipment it is feeding for testing purposes. All that stuff might have a nominal 50 input and output impedance but the match probably isn't that great.If you are building a complete piece of equipment, where say you are building an if amp then you would construct the appropriate matching circuitry so an attenuator would not be necessary or even desirable.You see the stipulation about adding an attenuator all over. But I've never seen an explanation as to how it works to mask a mismatch. The closest thing I have seen is an application note on the Mini-Circuits website (the link isn't handy) that is titled about using attenuators. Way down in the note they finally say something to the effect that, as I understand it, if you were feeding your amplifier through 50 ohm coax to another circuit, a mixer, detector, or whatever where both pieces of equipment had some mismatch to the line the two swr's would multiply. Putting a 50 ohm attenuator in the line would keep this from happening, or diminish the effect, and (I guess) both pieces of gear would see the 50 ohm line Zo. Take a look at that AN and see if that's how you interpret it.Pat----- Original Message ----- 73 2006-08-05 15:18:10 Wes Hayward Re: Feedback Amplifier/Attenuator Pad Hi all,
This subject of feedback amplifiers and pads is a recurring one that
has come up many time in emails and (in earlier times) letters. I've
been thinking of writing about it and finally got around to generating
a piece that is now on my web page. So try
I hope that this is useful.
76 2006-08-05 18:12:18 Jason Milldrum Re: Feedback Amplifier/Attenuator Pad Wes,
Thank you very much for that excellent article. Your explaination clears
up a lot of things for me, in a very descriptive way. I've been
breadboarding some BJT feedback amps, trying to characterize the circuit
with different parameters such as Ic and different feedback resistors. So
far, I've had very similar results to those shown in EMRFD. However, I can
see that I really need to build the return loss bridge in order to more
fully understand the behavior of these circuits.
I also want to thank the other gentlemen who took the time to answer my
question. Everyone provided me much insight into feedback amp/attenuator
Thanks and 73,
Jason Milldrum, NT7S
Wes Hayward wrote:
> Hi all,
> This subject of feedback amplifiers and pads is a recurring one that
> has come up many time in emails and (in earlier times) letters. I've been
> thinking of writing about it and finally got around to generating a piece
> that is now on my web page. So try
> I hope that this is useful.
> 73, Wes