[hpsdr] WDSP: smeter calibration

[Simon Eatough]

Thanks Scott for the explanation. I get it. The dbm levels represented by
the smeter are not incorrect, just different. The smeter is measuring the
power in the rx passband, and the panadadaptor level represents the power
in the fft bin width. The CW reference explains why when I tested using the
hermes emulator I got the same levels on both.
Cheers
Simon
ZL2BRG

On Thu, Jul 12, 2018 at 2:55 PM Scott Traurig <scott.traurig at gmail.com>
wrote:

> I'm very sorry, Simon, but it does not seem that you understand. The dBm
> levels in the S-meter are not incorrect. Why would you think that? Please
> go back and re-read the topic on the forum.
>
> The levels shown in the S-meter represent the total power in the selected
> RX passband, either in S-units, dBm, or both, depending on the style
> S-meter selected. This is in accordance with the ITU standards for S-meter
> measurements.
>
> The levels shown in the panadapter represent the total power in the
> selected panadapter FFT bin width in dBm. There is no corresponding ITU
> standard.
>
> Since noise power is broadband and relatively monotonic, at least over
> 10's to 100's of KHz, it scales perfectly with respect to noise measurement
> bandwidth. CW signals, which generally fit within both the selected
> passband and the FFT bin width, will measure the same in either case
> because of of their narrowband characteristic. SSB phone signals can be a
> bit confusing, but you simply have to remember that ANY RF POWER
> MEASUREMENT MUST BE REFERENCED TO THE BANDWIDTH IT IS MEASURED IN. If you
> measure the power of an SSB signal in a 2 or 3 Hz bandwidth you are only
> measuring the power IN THAT 2 OR 3 HZ BANDWIDTH. If you measure the power
> of an SSB signal in a 2 or 3 KHz bandwidth then you are measuring all of
> the power IN THAT 2 OR 3KHZ BANDWIDTH. These measurements are NOT the same,
> are NOT intended to be the same, and provide different views of the same
> information, each with its own usefulness. The first measurement
> (panadapter) allows one to understand how the total power is distributed
> within the 2 or 3 KHz total bandwidth of the SSB signal. The second
> measurement merely measures the total power for the entire signal (S-meter
> measurement).
>
> 73,
>
> Scott/w--u-2-o
>
>
> On Wed, Jul 11, 2018 at 6:51 PM, Simon Eatough <simon.eatough at gmail.com>
> wrote:
>
>> Thanks, Byron, Scott. Yes I understand now. Might explain why cudasdr
>> does some math maybe in a attempt to correct  the smeter levels. Would it
>> be better to simply not display the dbm levels in the smeter at all then,
>> since they are incorrect ?
>> 73's
>> Simon ZL2BRG
>>
>> On Thu, Jul 12, 2018 at 1:25 AM Scott Traurig <scott.traurig at gmail.com>
>> wrote:
>>
>>> No calibration is required. Nothing is wrong. There are no "issues",
>>> other than one must understand that all RF power measurements must be
>>> referenced to a measurement bandwidth..
>>>
>>> See here:
>>>
>>> https://apache-labs.com/community/viewtopic.php?f=6&t=2463
>>>
>>>
>>> On Wed, Jul 11, 2018 at 5:39 AM, Simon Eatough <simon.eatough at gmail.com>
>>> wrote:
>>>
>>>> ***** High Performance Software Defined Radio Discussion List *****
>>>>
>>>> Hi Folks. I've been playing with a anan10  sdr app which the wdsp
>>>> library to do the dsp processing. I've noticed that the peak signal levels
>>>> in the pan adaptor / receiver passband are about 10-15 db lower than that
>>>> reported by the WDSP library's GetRXAMeter() function.
>>>> linhpsdr which also uses the wdsp library has the same issue.
>>>> Wonder if anybody else has noticed this.
>>>>
>>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/attachments/20180712/65b478c4/attachment.html>