> thread. I summarized some of their comments and included a
> method that
> Rich suggested for measuring AVERAGE IMD with another
> receiver and a
> calibrated step attenuator at <http://gs35b.com/imd.html>.
> Of course
> this method can't be used to quote specific numbers for
> the 3rd and 5th
> order products but produces an overall look at your system
> performance.
I'm pretty sure that does not give you the average IM level
Tony.
I'll have to think about that a little bit more, but off the
top it seems wrong because the meter on our receivers
detects peaks, not average.
I'm pretty sure it actually would measure the highest value
of peak power contained in the bandwidth of the receiver
filter, assuming the receiver is perfect.
Average makes no sense to me, since a receiver S meter is
not an averaging measurement device.
I would add a correction that the level is not the average,
but rather the peak power contained in the filter bandwidth.
I'd add a second caution that the S meter should be set as
low as possible to give a reading, not half scale. Many
receivers fall totally apart at 30-40 over nine. If you had
40 dB of attenuation to make S-9 at the main signal, that
would place the main signal at 9 +40 when you are on the
adjacent channel. That would not be very smart. On the other
hand if you set the reading on the unwanted trash to a
minimal but reliable S meter movement of S-2, you would
have much less fundamental signal to deal with.
Let's look at my ICOM 751A data ( I happen to have it on my
desk).
S1= -130dBm
S2= -110dBm
S5= -100dBm
S7= -90dBm
S9= -80dBm
S9+10= -70dBm
S9+20= -60dBm
S9+30= -50dBm
S9+40= -40dBm
S9+50= -25dBm
S9+60= -17dBm
>From S2 to S3 is about 1dB, so resolution would be good. If
we were measuring peak power 40 dB down, S2 would place the
main signal 40 dB above -110dBm at S9 +10 and we would have
good confidence the 751A would not be non-linear.
On the other hand if we placed the signal as suggested at
S-9, the main signal would be at 40 dB over nine and the 751
is wiping out. Using mid-scale is not a good idea.
The whole thing really needs looked at carefully. Obviously
there already are a few problems.
73 Tom
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|