Intro
On 20apr21 we measured the xband spectra at the
input to the postamp chassis in pedestal (more
info). We saw a large increase in the spectral
power between 3 and 8 GHz. On 26apr21 we use the field fox
spectrum analyzer to measure the spectra in the hub to
see where the extra power was coming from. When
measuring the xband total power after the first amp in the
postamp chassis we found the amp output to be 2 dbm when
looking at blank sky. Most of the is power was coming from the
3 to 8 GHz extra power.TH
Measuring the spectra in the hub.
The xband chain in the hub looks like:
- schematic of the system
(.pdf).
- LNA (X1) -> 20 foot cable (X2) -> amp2 (X3)
->long cable to pedestal.
- We measured the spectra at X1, X2, and X3
- The analyzer setup was:
- 1001 points, 14GHz span , 7 GHz cfr, 5Mhz rbw, attn=0,
preAmp on, trace 1 avg10, trace pkhold
Comparing the power meter and spectrum analzyer values:
- We also measured the total power using a power meter at
some of the spots (more
info)
- When comparing the total power measured by the power meter
and computed from the spectrum analyzer spectra:
- Spectrum analzyer:
- each recorded point receives data from 14Ghz/1001 =
14Mhz span
- The rbw filter is only 5 MHz. I don't know whether the
spectrum analyzer is summing or averaging the multiple
samples that go into each display point.
- I assumed it was averaging the points.
- I computed a correction factor : 14/5 (about
4.5db) and multiplied each spectrum analzyer channel value
by this amount.
- When averaging over the spectra i limited the
integration 4Ghz to 10Ghz.
- I also computed 0 to14Ghz and the results were within
.1 db
- Power meter
- The power meter range was (100Khz to 18Ghz). So it
span was greater than the spectrum analyzer. I don't know
what the power levels looked like above 14 GHz.
Looking at the spectrum i assumed they were 0.
- The comparison results show the power meter values to be
higher by 3.2 to 5.5 db.
The measured spectra
The plots show the measured spectra (.ps)
(.pdf)
- Page 1: avg and peak hold spectra for polB.
- Black in the average 10 spectra. Red is the peak hold
spectra. The green line is the noise floor at the center
of the spectra (with a 50ohm load).
- top: (X1) At output of the lna up in the cone.
- We still see the 3-8 GHz added power in the output of
the LNA.
- Middle:(X2) after 20 foot cable from lna,before input to
amp2. This shows the cable loss.
- The power meter measured a total power of -37dbm
- Bottom:(X3) output of amp2.
- The power meter measured an total power of
-15dbm ( so the amp is about 22 db)
- The spike in the peak hold at 9.1GHz is probably rfi.
- Page 2: over plot the 3 spectra and the gain/loss.
- top,middle frame colors: Black is lna output, red
is amp2 input, blue is amp2 output.
- Top: avg 10 spectra
- The spectra at lower frequencies goes below the green
noise floor (the noise floor changes with frequency
band.
- Middle: Peak hold spectra
- Bottom: gain/loss of each set
- red: cable loss lna out to amp2 in (20foot cable)
- 0-2 GHz shows no change --> the input,output is
below the noise floor of the spectrum analzyer.
- 7.5 to 9.3 Ghz. The output hits the noise floor of
spectrum analyzer, so the actual loss will be greater
than this.
- Blue: amp2 gain.
- The input level is below the noise floor of the
spectrum analzyer for 0-4 ghz and 7.5 to 9.5 ghz. So
the actual gain will be higher than the measured
value.
- The amp gain is about 20db gain in the 8 to 9
GHz range used by the xband receiver (ignoring that
the input is below the noise floor of the spectrum
analzyer). The data sheet claims it is 18db 8-10 Ghz.
- Purple: total gain lna output to amp2 output.
- the 8-9 Ghz value is about 10db. Both the input and
output is above the noise floor of the spectrum
analzyer so the value should be accurate.