From: Jerry Avins on
rbb wrote:
>> I don't understand how increasing signal by 2x with noise
>> remaining the same ***decreases*** SNR by 3 dB.
>>
>
> Sorry. Increases the SNR by 3dB.

Coherent addition doubles the signal *voltage*, increasing S by 6 dB.
Doubling the bandwidth doubles the noise *power*, increasing N by 3 dB.
This gives a net gain in S/N of 3 dB as it should, considering that the
bandwidth (and the sideband power) has doubled. You wanted a free lunch?

Jerry
--
Engineering is the art of making what you want from things you can get.
�����������������������������������������������������������������������
From: Vladimir Vassilevsky on


Jerry Avins wrote:
> rbb wrote:
>
>>> I don't understand how increasing signal by 2x with noise
>>> remaining the same ***decreases*** SNR by 3 dB.
>>>
>>
>> Sorry. Increases the SNR by 3dB.
>
>
> Coherent addition doubles the signal *voltage*, increasing S by 6 dB.
> Doubling the bandwidth doubles the noise *power*, increasing N by 3 dB.
> This gives a net gain in S/N of 3 dB as it should, considering that the
> bandwidth (and the sideband power) has doubled. You wanted a free lunch?

It is not very simple to make good receiver for OOK because you have to
establish 0/1 decision threshold. The receiver performance depends on
that threshold dramatically, especially if incoherent detector is used.
In any realistic scenario, the amplitude of the signal is variable, and
the threshold has to track the variation. Fortunately, they use OOK
mostly for the applications where +/- 6dB of performance doesn't matter.


Vladimir Vassilevsky
DSP and Mixed Signal Design Consultant
http://www.abvolt.com