Hi,
if the buffer is used to following 20k sin wave and the thd of the output must be limited to 0.1%, how can i design the buffer?
Example the amplitude of 20k sin wave is 0.5v,which paramater of the opamp is important? And why?
Can someone help me?
3x
I think the parameter you should check first is the slew rate of the opamp. It should be more than the maximum slope of the input signal i.e. slope of the sin wave at 0 crossover.
The slew rate of opamp must be larger than vo*W0 discribed in razavi's book.But i don't konw why.
And somebody told me the bandwith is the crucial parameter in this case.Is it really?
In standard catalogue the seller presents the following op-amp parameters:
Supply voltage,
Bandwidth,
Slew-Rate,
V offset Min,
I bias,
Temperature range.
Manufacturer will have many more...
For first selection you will look at supply voltage and bandwidth.
The other parameters are also important but in your case, for 20kHz sine wave of 0.5Vpp in buffer configuration, they play almost now role...
thank u for your help.
we can use clear supply for simulation so the noise in the supply will not be calculated.
but i want to know which order of the bandwidth is needed in the case.
In the application this buffer must drive out a big capacitance so i want to design the Ibias ,slew rate,bandwidth ... of the opamp.
hi,
about the gain-bandwidth vs Slew-rate relationship, you can simply see it as
the dc gain of the buffer Av=Gm*Rout and the dominate pole=1/(Rout*Cload)
the gain-bandwidth product equals Av times pole, such GBW=(Gm*Rout)*1/(Rout*Cload) = Gm/Cload = Gm*SR/I
then as you mentioned, the higher the SR, the bigger the GBW
In my opinion, if the THD is your main concern, use fully diff architecture.
But in your case, i do not think SR and GBW is critical. Because the output swing is small and the signal bandwidth is not so high.
In my opinion, if the THD is your main concern, use fully diff architecture.
But in your case, i do not think SR and GBW is critical. Because the output swing is small and the signal bandwidth is not so high.
Well.
Since u asked, i think several MHz will be enough, but you should make sure the SR is enough. Like u said, SR should be larger than vo*Wo.
The reason is the SR is limited by the current and the capacitor. Consider the output stage(large signal), the charge and discharge capability should be enough.
U may do some simulation by yourself to find which parameter is crucial.