It seems to me that John's argument is in complete accordance with what PRR and I stated. With the ideal transformer, the signal voltage and the noise voltage from the source increase with the same factor (the turns ratio), keeping the source's signal to noise ratio intact, while the impedance increases with the square of the turns ratio. As a result of the increased voltages, the noise voltage from the amplifier has less influence. Any amplifier noise current gets a larger influence due to the quadratically increased impedance, though.
Hence, the signal to noise ratio at the output of the amplifier is higher with the transformer than without it, if the transformer is close enough to being ideal and if the amplifier has a sufficiently low input noise current.
Hence, the signal to noise ratio at the output of the amplifier is higher with the transformer than without it, if the transformer is close enough to being ideal and if the amplifier has a sufficiently low input noise current.
Actually, the condition that the source's signal to noise ratio should stay the same is not complete. The available signal power (signal power delivered into a power-matched load) should also stay constant. (Otherwise using a series resistor with an extremely low temperature should also help, which it doesn't.)
Re: I agree with John, actually....
1 A signal source has a S/N by nature, which we can't do anything about.
2 A amplifier adds noise, differently for different signal sources, because of the design of the input stage of the amplifier.
3 NF is how many dB's reduced S/N you will get at the output compared to at the signal source output.
4 NF is valid for all frequences, including audio.
Am I wrong or do have to read a book?
5 How do we use NF..... that's the question. For audio it's better the see the diagram of the noise.
john curl said:
Jocko Homo said:
What was the real explanations here besides go reading a book?john curl said:
1 A signal source has a S/N by nature, which we can't do anything about.
2 A amplifier adds noise, differently for different signal sources, because of the design of the input stage of the amplifier.
3 NF is how many dB's reduced S/N you will get at the output compared to at the signal source output.
4 NF is valid for all frequences, including audio.
Am I wrong or do have to read a book?
5 How do we use NF..... that's the question. For audio it's better the see the diagram of the noise.
Peranders, I think you are right about points 1, 2 and 4, and about point 3 when the noise of the signal source is thermal noise and the source's temperature is about 290K.
Regarding point 5, I usually calculate and/or measure the A-weighted average noise figure for a realistic source impedance when I design an amplifier that has to work with a source that generates thermal noise. It seems to me that the A-weighted average noise figure for a realistic source impedance is the perceptually most relevant figure. You can also calculate and measure the spot noise figure as a function of frequency, of course.
Regarding point 5, I usually calculate and/or measure the A-weighted average noise figure for a realistic source impedance when I design an amplifier that has to work with a source that generates thermal noise. It seems to me that the A-weighted average noise figure for a realistic source impedance is the perceptually most relevant figure. You can also calculate and measure the spot noise figure as a function of frequency, of course.
> I have no idea what the last post was about and the only was about and the only bit of sanity in it was...
Sorry to bring up topics you have not reached in your studies
. I was describing the context where NF is useful. However in audio, it is usually not useful.
While there is pith in Curl's comments and a glimmer of truth in mine, ALW said it best of all:
> ..all you will have done is raise the noise level of the source, in order to make the JFET look better.
ALW clearly explains why Rlim's intuition is correct.
Sorry to bring up topics you have not reached in your studies
. I was describing the context where NF is useful. However in audio, it is usually not useful. While there is pith in Curl's comments and a glimmer of truth in mine, ALW said it best of all:
> ..all you will have done is raise the noise level of the source, in order to make the JFET look better.
ALW clearly explains why Rlim's intuition is correct.
And I am sorry to have brought up those outside of yours
http://www.diyaudio.com/forums/showthread.php?postid=285870#post285870
I think I will muddle by without your "knowledge" some how..........
http://www.diyaudio.com/forums/showthread.php?postid=285870#post285870
I think I will muddle by without your "knowledge" some how..........
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.