Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.
That was the short and accurate answer but probably not what you wanted to hear.
The first problem is the second 40dB amplifier boosts the noise level of the first one so the output is more 'contaminated' than just 80dB of original signal.
The more practical problem is 80dB of gain is so high that it would be practically impossible to prevent some of the output being sniffed by the input and amplified in a loop causing instability.
It might be easier to think of the same gain in audio terms. It would be like standing in front of a VERY powerful loudspeaker with a VERY sensitive microphone and trying not to get howling.
It depends. What about the noise figure of a single amplifier ? Do you have the datasheet ?
Of course 80 dB is a quite high gain. In order not to have stability issues input and output must be isolated one each other of at least 80 dB.
Could you please explain a little be more on how the amplifier will be used ? I mean waveform and level in input and the load that will be connected to the output.
That isn't what I wrote.
All amplifiers introduce some noise. What comes out of the first stage is 40dB more of the signal you want plus some noise.
Put that through a second 40dB amplifier and you get the noise of the first stage amplified by that much as well.
Obviously, we try to design amplifiers to have as little noise as possible but there is always some present. In general, it is a bad idea to just daisy-chain amplifiers to get more gain. The optimum solution is usually to concentrate on low noise in the first stage, but not necessarily lots of amplification, then add more gain later where the signal is already bigger so noise is less prevalent. Exactly how the gain is distributed to get the total needed will depend on many factors including frequency and power levels and the technologies available to handle them.
Another analogy: one car, flat out may do 100 MPH, bond another identical car after it, the top speed will not be 100x100 = 10,000 MPH!
If 80 dB amplifier gain is realized in a receiver, it's a selective amplifier with filters, e.g an IF amplifier. I presume the said 40 dB amplifiers are wideband LNA with several 100 MHz or even GHz bandwidth. In this case, such a high gain isn't feasible, noise and small interfering signals will already exceed the maximal undistorted output voltage.
It depends. What about the noise figure of a single amplifier ? Do you have the datasheet ?
Of course 80 dB is a quite high gain. In order not to have stability issues input and output must be isolated one each other of at least 80 dB.
Could you please explain a little be more on how the amplifier will be used ? I mean waveform and level in input and the load that will be connected to the output.
It means that the next stages should have extremely high IP3s to achieve reasonable linearity.
So there is a trade off here. With having a high gain at the input (using 2 stages) you will get a perfect noise figure but your linearity will be awful.
The output signal will be seen by the input with the attenuation given by the input-output isolation and a phase depending from many factors, as a result at some frequencies the two stage amplifier will start to oscillate even if there is no input signal.
See it as a system with feedback that a certain frequencies is positive.
you can but your signal at output finally s always small hence it never reaches the requirement of the subsequent stages
If you wanna make it larger you have to make sure that your second stage has such an extremely huge P1dB point in order to keep the output signal not saturated
[p1db(total) = 1/p1dB,1 + G1^2/ P1dB,2)]
Hope it helps
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
