[hpsdr] Problem with Powersdr 2.2.3 and Hermes

[Andrew]

Hi Phil,
Thanks for confirming that the adc overload is triggered by the ADC overflow output pin, as I suspected. As you say its going to be literally 0dBFS!
Yes, Powersdr is fully calibrated and the s meter reads spot on in dBm with a sig gen applied to the input.
So are you seeing the same ~7dB difference I noted?
Really my question is I suppose, what is the meter indicating when set to ADC L or ADC R, it looks like the power in the decimated bandwidth I and Q?
How is it calculated?
KISS Konsole reads spot on also (in dBm) but does not display wideband total power i.e. ADC dBFS, might be a nice thing to add?
Thanks,
Andrew
G4XZL





________________________________
 From: Phil Harman <phil at pharman.org>
To: Andrew <irbsurfing at yahoo.co.uk>; hpsdr at lists.openhpsdr.org 
Sent: Wednesday, 17 October 2012, 11:53
Subject: Re: [hpsdr] Problem with Powersdr 2.2.3 and Hermes
 

Hi Andrew,
 
In answer to your questions: 

What triggers the overload warning, is it the actual ADC overflow 
output or triggered by the input level exceeding a certain dBFS level determined in 
software (if so, what is it -1dBFS or 0dBFS etc??)
 
There is a pin on the LTC2208 that gets set when 
the ADC overloads. It’s this pin state the gets sent to the PC via the C&C 
data bits in the USB protocol.  Not sure exactly when this gets triggered 
but I would expect at +/- full scale. 

Incidentally, I notice that there is about as far as I can estimate a 
-7dB (maybe its 6dB, can't tell) difference between ADC input level (dBFS) and the level 
indicated on the Powersdr s meter set to read ADC level.
 
Have you calibrated PowerSDR with a known signal 
level at the antenna socket? 

Is there any way to observe the actual ADC wideband level in dBFS??
 
KISS Konsole will display either the actual I and Q 
data being sent to the PC (after filtering and decimation) or an FFT of a 16k 
block of contiguous  samples taken directly out of the ADC – i.e. the 
wideband 0-55 MHz bandscope display.  At least with my Hermes and Mercury 
boards the displayed level is within 0.5dB of a signal from my HP8640B signal 
generator. 
 
73 Phil...VK6APH 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/attachments/20121017/a3c91a7d/attachment-0004.htm>