[Hackrf-dev] HackRF One receiver sensitivity vs temperature vs host computer resources

[Andrew M.A. Cater]

On Thu, Sep 18, 2014 at 08:07:22PM -0400, Robert Morales wrote:
> Hello,
> I just received my HackRF One.  It is running git-44df9d1 firmware.  I am
> using a repurposed old laptop with a Centrino Core2 Duo 2.4GHz processor
> and 2Gb of RAM on which I've loaded the latest Kali Linux distro and the
> latest build of the GNU Radio packages.
> 
> My question is this, doing some initial validation testing that the board
> is working, I'm using gqrx to tune to strong local FM stations with the
> appropriate settings (WFM, etc).  The audio quality is decent but not
> "stereo FM" quality.  I'm not as concerned by that as the apparent
> degradation in receiver performance over time.  Monitoring the same station
> say 30 mins after the computer has been on, I can't seem to find as many
> stations on spectrum as before and the one being monitored appears somewhat
> attenuated.  By comparison, if I use the same antenna but with the cheap
> RTL-SDR on a newer PC running Win8.1 and SDR#, the quality of the audio is
> "commercial receiver quality" plus the receiver appears to be more
> sensitive than the HackRF One under GNU Radio.  I have also noticed that
> the GNU Radio/Kali Linux PC gets hot quickly once it is doing signal
> processing (fan kicks in and hot air definitely cycled out of the laptop).
>   Is my problem more likely to be due to the Kali Linux laptop overheating
> or thermal issues with the HackRF One itself.  My HackRF One is stock in
> its original plastic enclosure.
> 

SDR takes processing power. Software processing of radio signals as done
here needs high I/O, significant post-processing - most people seem to
recommend a quad core machine with a large-ish amount of memory.

The nearest comparison I can think of is the early days of computer graphics and
fractals where a 286 or 386 could labour all night to produce a good image.

The highest spec highest price SDRs have significant amounts of FPGA processing
and/or are for much lower bandwidths. Ultra wide band SDR with a range similar
to that of the HackRF is difficult - if you just want a test receiver, you add a signal
output from a very, very expensive spectrum analyser.

It's also worth thinking that the TV dongles are intended for FM radio reception
so may also have minimal tuned circuits and filtering that will operate in the 
broadcast band.

> I have not yet tried using the HackRF One under SDR# for comparison.
> 
> Thanks in advance for any guidance!
> Bob.
> wp4bqv at gmail.com
> 

All the above is purely my opinion,

All the very best,

Andy

[amacater at galactic.demon.co.uk / G0EVX]