5
dc2078af
DEMO MANUAL DC2078A
Setup Signal Sources
Figure 4 shows a proposed IP3 test setup. This setup has
low phase noise, good reverse isolation, high dynamic
range, sufficient harmonic filtering and wideband imped-
ance matching. The setup is outlined below:
a. High performance signal generators 1 and 2 (HP8644A)
are used. These suggested generators have low har-
monic distortion and very low phase noise.
b. High linearity amplifiers are used to improve the reverse
isolation. This prevents cross talk between the two
signal generators and provides higher output power.
c. A low pass filter is used to suppress the harmonic content
from interfering with the test signal. Note that second
order inputs can “mix” with the fundamental frequency
to form intermodulation (IM) products of their own. We
suggest filtering the harmonics to
–
50dBc or better.
d. The signal combiner from Mini-Circuits (ADP-2-9)
combines the two isolated input signals. This combiner
has a typical isolation of 27dB. For improved VSWR and
isolation, the H-9 signal combiner from MA/COM is an
alternative which features >40dB isolation and a wider
frequency range. Passive devices (e.g. combiners) with
magnetic elements can contribute nonlinearity to the
signal chain and should be used cautiously.
e. The attenuator pads on all three ports of the signal
combiner will further support isolation of the two input
signal sources. They also reduce reflections and promote
maximum power transfer with wideband impedance
matching.
Setup the Spectrum Analyzer
a. Adjust the spectrum analyzer for maximum possible
resolution of the intermodulation products’ amplitude
in dBc. A narrower resolution bandwidth will take a
longer time to sweep.
b. Optimize the dynamic range of the spectrum analyzer
by adjusting the input attenuation. First increase the
spectrum analyzer’s input attenuation (normally in
steps of 5dB or 10dB). If the IM product levels decrease
when the input attenuation is increased, then the input
power level is too high for the spectrum analyzer to
make a valid measurement. Most likely, the spectrum
analyzer’s 1st mixer was overloaded and producing its
own IM products. If the IM reading holds constant with
increased input attenuation, then a sufficient amount
of attenuation was present. Adding too much attenua-
tion will bury the intended IM signal in the noise floor.
Therefore, select just enough attenuation to achieve a
stable and valid measurement.
c. In order to achieve this valid measurement result, the
test system must have lower total distortion than the
DUT’s intermodulation. For example, to measure a
47dBm OIP3, the measured intermodulation products
will be
–
92dBc below an
–
18dBm/tone input level and
the test system must have intermodulation products
approximately
–
98dBc or better. For best results,
the IM products and noise floor should measure at
least
–
102dBc before connecting the DUT.
OPERATION
