[Laser] minimum frequency PSK31

[Art]

>
>In a message dated 11/13/2006 9:50:33 AM Central Standard Time,
>KY1K at verizon.net writes:
>
>System 1  is a dc to 400 Hz bandwidth receiver listening to a PSK31
>signal at 300  Hz.
>
>System 2 is a dc to 800 Hz receiver listening to a PSK31 signal at  600 Hz.
>
>


OK, within the systems bandwidth, you can pick any thin slice of 
spectrum, and there will be XX amount of noise in that spectrum. Now, 
if you choose a slice that is 2 times smaller, you will hear the same 
signal strength signal 3 db better. If you reduce the bandwidth of 
the slice of spectrum you monitor by a factor of 10, than the signal 
will be 10 db stronger. This is conventional wisdom, and something 
everyone understands.

Conventional wisdom goes all to heck in a light receiver however 
PROVIDED THAT THE B_A_N_D-W_I_D_T_H OF THE RECEIVER IS CHANGED BY 
MODIFYING ITS GAIN WITH A GAIN DETERMINING RESISTOR (using feedback).

So, if I use a receiver modification to reduce the passband of a 
receiver by a factor of 2, we get a 2 X 2 enhancement. If I reduce 
the receiver passband by a factor of 10, we get 10 X 10 enhancement.

The key to this is that we are changing the receivers passband by 
limiting the G_A_I_N of the active amplifier by changing the value of 
the feedback resistor.

So, in a current to voltage converting stage, where the gain 
(therefore the system bandwidth) is determined by F_E_E_D_B_A_C_K, we 
have a cumulative reduction in noise, not a 1 to 1 reduction that we 
all understand from our radio experience.

In the current to voltage converting stage system, we get a one to 
one reduction in noise from the fact that we have reduced the 
bandwidth by a factor of 2. BUT, WE GET AN ADDITIONAL 
E_N_H_A_N_C_E_M_E_N_T BECAUSE WE USE F_E_E_D_B_A_C_K TO CONTROL THE 
GAIN (AND THEREFORE THE BANDWIDTH) OF THE RECEIVING SYSTEM.

In the 60's and 70's, Q multipliers were all the rage because they 
exhibited the same effect. In the Q multiplier, the G_A_I_N of the 
system (and therefore it bandwidth) is set by varying 
F_E_E_D_B_A_C_K, so the same concept applies.

I have to admit, it is a hard pill to swallow initially and I 
struggled with it for a long time....until I understood the key 
factors that make the issue appear to defy conventional behavior::>


>
> >From a practical standpoint, I think what you are saying is 
> that  if someone
>wanted to experiment with digital signals, then the system  should be built
>with lower rather than higher frequencies in mind, and that 
>the  receive chain
>should have low pass characteristics with the "knee" above the  expected
>signals.  This sounds like a practical suggestion.
>

YES!!!!!!!!!!!!!!!!!!!

>To me, it still "feels" like using bandpass filtering should work as  well.
>At least if you stay well within the operating range of the sound  card you
>are using to decode.  Perhaps there is a way to run 
>experiements  with a system
>that has a 400 Hz cutoff ( DC to 400 ) compared to a system with a  1600 Hz
>cutoff ( DC to 1600 ) followed by a 1200 Hz cutoff high pass  filter.
>
>

OK, go ahead and do the experiment....but it can be further 
simplified by eliminating the LPF (because the LPF is actually built 
into the receiver by appropriate selection of the gain determining resistor).

In fact, you do not even need optical input signal.

You can confirm this just by looking at the noise level over the 
passband of each receiving system. Use a POT to reduce or increase 
the receivers bandwidth and plot the noise output for each setting of 
the pot. You will see that doubling the value of the pot results in 
the noise being reduced by a factor of 4, not 2.

K3PGP did this years ago after I suggested he try it, and he verified 
the results just fine.

Hope this helps.

Art



-- 
No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.1.409 / Virus Database: 268.14.5/534 - Release Date: 11/14/2006