Continue to Site

# coherent and incoherent sampling ...

Status
Not open for further replies.

#### talking

##### Member level 2
coherent sampling relation is as follows:
N/fs = M/f0
where N: # of samples, fs: sampling speed in Hz, M: the integer number of cycles in sampled data, f0: sine frequency.

Here is a question. Let's assume M is changed to non-integer, and fs is accordingly calculated again to satisfy the above coherent equation (N/fs = M/f0 --> fs = f0/M*N).
Then is this relation 'coherent' or 'incoherent'?

Let's think of a specific example as below...
The following is the coherent sampling relation.
N = 8192
fs = 82MHz
M = 2503
f0 = 25.05MHz

Then, now i slightly changed M to 2502.1 (non-integer), and fs was recalculated using "N/fs = M/f0" (--> fs = f0/M*N).

N = 8192
fs = 82MHz
M = 2502.1 <------- non-integer
f0 = M/N*fs = 2502.1/8192*82M = 25.04MHz <------- results from the calculation of N/fs = M/f0.

Can we say it's incoherent sampling condition?

It is probably incoherent over the short term but over a long interval it may still be coherent with a PLL and could be harmonically related as harmonics of some lowest common denominator.

Thanks, SunnySkyguy. i assumed there is no additional circuitry to enhance incoherent condition.. basically, i intend to make the incoherent condition due to some special reason.. so i need to confirm the condition i gave can keep incoherent relation, without any additional activity (ex. pll as you commented).

Status
Not open for further replies.