There cannot be "f" in there, because f is on the left side of the formula from the pdf (fc). You cannot calculate f from f.
So this must be some different "s".
TDA2030 from the Workbench is presenting lower bandwidth, like 20kHz instead of 100kHz of LM's. I don't know why. But doubling Cf while Rf2 twice smaller seems good idea (when observing proportions of the TDA graphs and feedback components being changed at the same time)
Last edited:
Hence my confusion earlier in the thread. The only 's' I know of in a frequency response calculation is complex frequency, s = j*2*pi*f.
Tom
If you look at the MIT videos you se that f was used in the olden days to describe the feedback transfer function. So the formula is right, but the f is not frequency, but feedback!
Yes, but the f is not T(s), see the video mentioned in an earlier post (new link)
(And remember the formula is taken from over 20 year old documentation, I think they used the MIT notation. For those guys Hz was for amatures, they was thinking in radians)
Found a series of MIT lectures:
https://ocw.mit.edu/resources/res-6...os/lecture-1-introduction-and-basic-concepts/
At 19:20 the feedback figure is displayed. There f is the anotation of the feedback transfer function.
So then fc(s) makes sense. Nowadays that is quite confusing🙂
(And remember the formula is taken from over 20 year old documentation, I think they used the MIT notation. For those guys Hz was for amatures, they was thinking in radians)
Found a series of MIT lectures:
https://ocw.mit.edu/resources/res-6...os/lecture-1-introduction-and-basic-concepts/
At 19:20 the feedback figure is displayed. There f is the anotation of the feedback transfer function.
So then fc(s) makes sense. Nowadays that is quite confusing🙂
Last edited:
Normally, using a lower case letter, as in f(t), means the time domain.
Using an upper case letter, as in F(s), means the complex frequency domain.
I agree that is the normal notation nowadays.
But this is old notation that is not updated.
They even write: At higher frequencies feedback resistance works with Cf
to provide lower AC Gain.
But this is old notation that is not updated.
They even write: At higher frequencies feedback resistance works with Cf
to provide lower AC Gain.
fc = [Rf1 Rf2 (s + 1/Rf2Cf)]/[(Rf1 + Rf2)(s + 1/Cf(Rf1 + Rf2))]
fc = [Rf1 Rf2/(Rf1 + Rf2)] [(s + 1/Rf2Cf)/(s + 1/Cf(Rf1 + Rf2))]
fc = Rf1||Rf2 * [(s + 1/Rf2Cf)/(s + 1/Cf(Rf1 + Rf2))]
fc = Rf1||Rf2 * [s/s] = Rf1||Rf2, if s is much larger than 1/Rf2Cf
fc = [Rf1 Rf2/(Rf1 + Rf2)] [(Rf1 + Rf2)/(Rf2)] = Rf1 Rf2/Rf2 = Rf1 if s is much smaller than 1/Cf(Rf1 + Rf2))
So feedback resistance (or impedace) will go from Rf1 at DC to Rf1||Rf2 at high frequency.
The shift will happen from around CfRf2 to Cf(Rf1 + Rf2) radians.
The slope will be -6dB/oct somewhere between.
That is the old world way of analysis and the formula makes it much easier to do.
fc = [Rf1 Rf2/(Rf1 + Rf2)] [(s + 1/Rf2Cf)/(s + 1/Cf(Rf1 + Rf2))]
fc = Rf1||Rf2 * [(s + 1/Rf2Cf)/(s + 1/Cf(Rf1 + Rf2))]
fc = Rf1||Rf2 * [s/s] = Rf1||Rf2, if s is much larger than 1/Rf2Cf
fc = [Rf1 Rf2/(Rf1 + Rf2)] [(Rf1 + Rf2)/(Rf2)] = Rf1 Rf2/Rf2 = Rf1 if s is much smaller than 1/Cf(Rf1 + Rf2))
So feedback resistance (or impedace) will go from Rf1 at DC to Rf1||Rf2 at high frequency.
The shift will happen from around CfRf2 to Cf(Rf1 + Rf2) radians.
The slope will be -6dB/oct somewhere between.
That is the old world way of analysis and the formula makes it much easier to do.
Some love to do the math in their own heads. (And 20 years ago spice was not that easily obtainable)
That's what we did at university in the 70s. Such notation seems much clearer, also.
The tispice lm3886 schematic file with Cf and Cc.
They certanly affect the HF responce🙂
Had not used it before. Agree with tomchr, much easier to use than pen and pencil🙂
They certanly affect the HF responce🙂
Had not used it before. Agree with tomchr, much easier to use than pen and pencil🙂
Sorry this was plain wrong.
Should be: T(s) = Vo/Vi = 1 + Zf(s)/Rg = 1 + fc(s)/Rg.
H(s) I wrongly remembered from from other feedback systems like this
https://courses.engr.illinois.edu/ece486/documents/set5.pdf
- Status
- Not open for further replies.