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.
The formula for upper limit to max frequency can be seen from the graph of the output when it is slew rate limited. You can take that (ideally) as a linear sawtooth with slope equal to the slew rate (the maximum rate at which the output can change.) The Voltage swing is (again ideally) from +V to -V (=2Vss) and this happens in T/2 (and from -V to +V in the other T/2).
So by definition of slope, SR = 2Vss / (T/2) which rearranges to T/2 = 2Vss / SR
That is as I said an idealised maximum. At low frequency the output appears approx square wave, but the sides always slope because switching is limited by the slew rate. As the frequency increases the flat top and bottom become shorter. Eventually (ideally) the flat top and bottom would be zero, with the sloping sides meeting to give the sawtooth at the frequency given above.
But my earlier assumption of a linear sawtooth is also an idealised fiction. The slew rate is just the maximum value the sawtooth slope can reach. Particularly at the extremities of the output the slope will be less than the slew rate, so T/2 will be longer than calculated for the linear case. Hence the > in the equation, T/2 > 2Vss / SR
The reason for the second > is, I think, to say that for practical design purposes the frequency should be *very much less* than this idealised maximum. The calculation of frequency proportional to 1/RC is based on the output being roughly a square wave, which is only a reasonable approximation when T/2 >> 2Vss / SR
Also you probably want the output to be stable and not vary with changes in Vss or SR. So the further you stay away from the limiting condition (where these factors are paramount) the better.
The maximum frequency for this circuit will also be significantly less than the gain/bandwidth product of the amplifier. Since the frequency is slew-rate limited, it follows that the frequency can increase if the aplitude is reduced (and vice versa)
The formula for upper limit to max frequency can be seen from the graph of the output when it is slew rate limited. You can take that (ideally) as a linear sawtooth with slope equal to the slew rate (the maximum rate at which the output can change.) The Voltage swing is (again ideally) from +V to -V (=2Vss) and this happens in T/2 (and from -V to +V in the other T/2).
So by definition of slope, SR = 2Vss / (T/2) which rearranges to T/2 = 2Vss / SR
That is as I said an idealised maximum. At low frequency the output appears approx square wave, but the sides always slope because switching is limited by the slew rate. As the frequency increases the flat top and bottom become shorter. Eventually (ideally) the flat top and bottom would be zero, with the sloping sides meeting to give the sawtooth at the frequency given above.
But my earlier assumption of a linear sawtooth is also an idealised fiction. The slew rate is just the maximum value the sawtooth slope can reach. Particularly at the extremities of the output the slope will be less than the slew rate, so T/2 will be longer than calculated for the linear case. Hence the > in the equation, T/2 > 2Vss / SR
The reason for the second > is, I think, to say that for practical design purposes the frequency should be *very much less* than this idealised maximum. The calculation of frequency proportional to 1/RC is based on the output being roughly a square wave, which is only a reasonable approximation when T/2 >> 2Vss / SR
Also you probably want the output to be stable and not vary with changes in Vss or SR. So the further you stay away from the limiting condition (where these factors are paramount) the better.
The maximum frequency for this circuit will also be significantly less than the gain/bandwidth product of the amplifier. Since the frequency is slew-rate limited, it follows that the frequency can increase if the aplitude is reduced (and vice versa)
I'm a new user of this forum, and this is my first post.
If you're a Capture user, you can download some pspice simulations from my personal blog. These simulations are completly free because they are made using a Capture student version and under Creative Commons license.
If you need a Capture student version it's avaliable in follow link: PSPICE - Descargar
Simulation that maybe can be useful for you: Operational Amplifier - Oscillator in square wave configuration.
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.
