[hpsdr] HPSDR Signal Overload ( In the digital processing???) (Update)

[Mike F]

The received noise floor has dropped several dB from this morning and 
now the problem isn't evident.  This could be an indication that if the 
total processed band energy is low enough, a digital process overload is 
not happening.

Mike Fager, K7SR

On 4/5/2011 9:40 AM, Mike F wrote:
> This morning I had a fairly strong noise floor while operating on 14m. I
> noticed that every so often the floor would take a step increase and
> then take a step decrease. As I woke up I realized that the noise floor
> increases/decreases were coincident with strong signals (transmissions)
> appearing and disappearing in the display window. These signals were not
> excessively strong. I looked at the following.
>
> 1. The adc overload light on mercury was not lighting. (I had already
> concluded that it was unlikely that a reasonable signal change in my
> small portion of the overall bandwidth would generate visible adc
> overloads.)
>
> 2. The condition was not affected by turning the preamp on/off.
>
> 3. Since I had just switched to Metis, I switched back to Ozy and
> verified that there was no change in the problem.
>
> 4. I have just switched from psdr 1.19.3.2 to 1.19.3.3. However, I tried
> 3.2 and the problem was still there.
>
> Mercury firmware version is 2.9.
>
> As can be seen from my subject, I am biased towards some problem
> somewhere in the DSP chain.
>
> I have two screen capture JPEGs with waterfalls that clearly show the
> problem described above. However, it is my understanding that we cannot
> include attachments to these messages. I will be happy to supply same to
> any interested parties.
>
> Thanks
>
> Mike Fager, K7SR
>
>

 1302031000.0