[hpsdr] Bogus ARRL Test method for IMD

[Graham / KE9H]

Frank:

Measuring IP3 on Mercury and receivers of similar architecture
is a particularly interesting problem.  The reason is that IP3 is not
a (somewhat) constant value as on a traditional receiver.  It continuously
"floats" and changes as a function of the strongest signal in the
passband of the A->D converter.  Phil H knows a lot more than
I do about the mechanism, but strong signals inside the passband (unless 
they are
so strong as to overload the A->D) actually improve the linearity
of Mercury by forcing it to move the weak signal (embedded in a composite
input signal) across many more steps in the A->D converter, thus
averaging out conversion errors.

If the only signals you are looking at are the signal generators being
used to measure IP3, you will get one (lower) value.  If you insert
another off frequency strong signal. The IP3 value you measure
will rise.

I think this is very counterintuitive for what people expect from a
receiver, that strong off frequency signals improve, rather than
degrade Mercury's receiver performance.

So, all the existing IP3 measurement standards will give you a value,
probably all a different value, and none of them will really tell you
how Mercury will work in the real world.  And, none of them will
show Mercury in its best light.

--- Graham

==


FRANCIS CARCIA wrote:
> ***** High Performance Software Defined Radio Discussion List *****
>
>   
>
> ------------------------------------------------------------------------
>
> The new method was devised to hide the synthesizer close in phase 
> noise floor wich is usually more of a problem in a good receiver 
> design. This is a good system if you never work anyone below S5.
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> HPSDR Discussion List
> To post msg: hpsdr at openhpsdr.org
> Subscription help: http://lists.openhpsdr.org/listinfo.cgi/hpsdr-openhpsdr.org
> HPSDR web page: http://openhpsdr.org
> Archives: http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/


 1256955665.0