Frank says:
>Yippee - does anyone have any idea of what the theoretical maximum
>and/or the current practical maximums would be on these? Since
>I've seen the FT1000D has a 3rd order of 21 I can tell that it's
>better than the 950 - but - how much better could one yet go?
Firstly, I'm not entirely sure what is meant by 'blocking dynamic range'.
I've always measured blocking as a composite measurement - feed in a signal
at nominal sensitivity, say -107dBm - and then increase the unwanted signal
20kHz off tune until the SINAD degrades by 3dB. This is a composite
measurement, because it can be gain compression, or it can be reciprocal
mixing, or both.
The required intercept point is a function of the signal levels, and the
amount of filtering prior to the mixer. This is because the number of 3rd
order IMD products falling within the receiver front end passband is n(n-1)
+ 1/2n (n-1)(n-2), where n is the number of signals producing IMD. So the
narrower the filtering before the mixer, the lower the number of products,
and consequently, the lower the intercept point can be for any given amount
of reception degradation.
Allied to intercept point is the phase noise performance, and there is no
point in having superb intercept point and lousy phase noise. Consider a
+10dBm intercept receiver, with a 10dB noise figure and a 2.5kHz bandwidth.
The noise floor is -140dBm, and the spurious free dynamic range is 2/3of the
difference between intercept point and noise floor, or 100dB. That means
that two signals at -40dBm will produce an IMD product equal to the noise
floor. If at the offset from the tune frequency of one of those signals, the
phase noise averages -134dBc/Hz over the 2.5kHz, the phase noise will raise
the rx noise floor by a level equal to the IMD product.
Most modern transceivers seem to be around the -130 to -140dBc/Hz, so we can
see that pushing intercept points won't help too far. Obviously, these
figures are a little artificial, since we're considering very few signals,
and the separation of the unwanted from the tune frequency is another
factor.
I used to work on the basis that the use of an antenna attenuator was a sign
of inability to design a good front end. I'm now not quite so sure, feeling
that an input intercept of about +12 to + 15dBm with a simultaneous 10dB
noise figure and an attenuator is probably all you need. Add to this that if
you look at the available figures for atmospheric noise, it is very rare to
find the noise floor low enough for a very low noise figure to be required
at HF at the same time that there are really big signals present. Even on
10m, in the middle of the night, a 10dB noise figure is still low enough for
external noise to be 4dB above receiver noise where I am. that's with 2 dB
of feeder loss from a 4ele beam at 68 feet.
If you really want to push the dynamic range requirement, then a 2m SSB
contest receiver is the place!
Lower noise synthesisers are the key to being able to use much better IMD,
but the discrete spurs have to be down as well, which is why DDS has its
problems. Mayhap the old 75A4 type architecture is worth revisiting.
73
Peter G3RZP
--
FAQ on WWW: http://www.contesting.com/ampsfaq.html
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|