Re: ADC saturation rate

[Robin48gx]

On Mon, 15 Sep 2003 15:17:03 +0000, Yu Xiaodong wrote:

> Hi
>
> I am doing a AGC project. Does anyone know how to decide the ADC
> saturation rate. In my simulation, it seems to have a little saturation
> is better that there are no saturation. It is difficult to decide which
> level the saturation rate the system should maintain. I appreciate if
> anyone give me some hint of it.
>
>
>
> Regards,
> Rxd


Do you mean the time for the holding capacitor to charge ?

What ADC / microcontroller are you using ?

[Oliver Faust]

Robin48gx <[email protected]> wrote in message news:<[email protected]>...
> On Mon, 15 Sep 2003 15:17:03 +0000, Yu Xiaodong wrote:
>
> > Hi
> >
> > I am doing a AGC project. Does anyone know how to decide the ADC
> > saturation rate. In my simulation, it seems to have a little saturation
> > is better that there are no saturation. It is difficult to decide which
> > level the saturation rate the system should maintain. I appreciate if
> > anyone give me some hint of it.
> >


Hi,
To determine the "saturation rate" you need to state the probability
density function of your input signal. Then the number of quantization
levels (N=2^B) is required. The next task is to partition the
probability density function into N portions having the same area
under the curve. The voltage levels for these partitions are the
borders between the quantisation steps.
Saturation only occurs when you want to implement the conversion
system. Off the shelf ADCs have fixed quantization steps and a fixed
operation voltage range. The job of the gain control is to minimize
the difference between the optimal system, described above, and the
actual ADC hardware. That means for a hardware ADC we have a fixed
scheme of partitioning the probability density function. The only
thing we can do to make the area for all partitions equal is to alter
the gain of the input signal. The result of this optimization process
is a gain factor. This gain factor stretches or compresses the
original probability density function of your input signal. After
obtaining the gain factor, there might be a condition where the
resulting probability density function is wider then your input
voltage range. The rate of the overflow is then determined by the area
under the curve at the right starting from the maximum input voltage.
The rate of the underflow is the area under the curve starting form
the minimum voltage level.

Oliver Faust