[time-nuts] OT: Spectrum Analyzer

[Lux, Jim (337C)]

On 10/31/09 5:26 AM, "John Ackermann N8UR" <jra at febo.com> wrote:

> The HP8569 is a good moderate-price choice that goes up to 18GHz.  It's
> new enough to have digital display, so you can do trace math like
> normalization, and dump plots via GPIB.  However, you cannot set
> commands via GPIB.
> 
> 
> John
> ----
> Brent Gordon said the following on 10/30/2009 11:06 PM:
>> I'm thinking of buying a spectrum analyzer and would like to know what
>> Time Nuts recommend.  My requirements are fairly simple:
>> 
>> 3GHz Max frequency or higher
>> Either GPIB or Ethernet interface for control and data capture
>> Not much larger than an average desktop computer.  Portable is nice but
>> not necessary.
>> Preferably under $3000.
>> 
>> I thought about building Scotty's Spectrum Analyzer or Poor Man's
>> Spectrum Analyzer, but decided I would rather buy one then build one.
>> 
>> I have an HP 141T but I am looking for something more modern.  One of my
>> uses will be looking at C and Ku band satellite signals (down converted
>> to 950-2050 MHz).  I'll also be using it to look at various RF data
>> links from 433 MHz to 2.4 GHz.
>> 

Since you're thinking about building something, how about somewhere in
between a piece of lab gear and total homebrew.  The PC controlled receivers
like the Icom PCR1000 work pretty well as a spectrum analyzer and tune DC to
1 GHz-ish.  So a good LO and a mixer in front of this might be a way to go.
Yeah, you won't get the whole 3 GHz in one span.  For smaller resolution BW
than the narrowest filter in the PCR1000, you can run the audio output with
the receiver in SSB mode into a spectrogram program.

Instead of the PCR1000, you could use one of the SDR widgets out now (like a
softrock), but that's going to be a LOT more homebrew.

The LO for mixing down from microwaves could be done a bunch of ways.
There's PLL eval boards which have a USB interface, for instance.