U N D E R S T A N D I N G T H E B B 6 0 A H A R D W A R E
One source of image-related spurious signals comes from the image from the first mixer. These are
usually far from the signal of interest and very low in amplitude, especially below 2 GHz, but may be
something like -40 dBc in small areas of the upper bands. These spurs will be rejected using the software
“spur reject.”
Another source of spurs occurs when LO1 exceeds 4.4 GHz. The LO is doubled in this case,
introducing spurs from the LO subharmonic. This is especially noticeable when you are sweeping across
5.7 GHz, as a signal injected at 4 GHz will create a false spur which may be something like -28 dBc from
the actual signal. These spurs will be rejected using the software “spur reject.”
Other spurs may occur 280 MHz below the actual signal, from the IF filter rejection. These are usually
quite low.
Finally, some spurs are introduced during the final mixing and digitizing. Anything not rejected by the
final 140 MHz may be aliased into the data, specifically signals 30-50 MHz above and below the desired
signal. These are typically about –50 dBc, but may be higher near gain compression. These will generally
be rejected by using the software “spur reject,” but there are some exceptions, especially even multiples
of 10 MHz.
For spans below 500 kHz and some streaming and zero-span modes, additional spurs from the
fractional-N local oscillator may be observed. These will usually be below -50 dBc, and will generally be
rejected by using the software “spur reject.”
Residual Signals 9.3
A residual signal appears even when there is no signal input. The BB60A has noticeable residual signals
at multiples of its 10 MHz timebase. These are guaranteed to be below -90 dBm for a reference level of
-50 dBm, attenuator 0 dB (for advanced users, 0 dB atten, gain of 2 or 3), but will typically be well below
this level. For higher reference levels or lower gain, these may be higher. If these residual signals interfere
with your signal measurements, an external RF amplifier may be needed.
Amplitude Accuracy 9.4
Some of the filters are temperature-sensitive. We have included an automatic self-calibration when a
significant temperature change is observed. If this is turned off, a temperature change may introduce
amplitude ripple in the passband, increasing measurement error by a dB or more in some cases. Because
of this, bypassing the automatic self-cal is not recommended when amplitude accuracy is important.
Scalloping Loss 9.4.1
The “native” bandwidths used in the BB60A come directly from the windowed FFT results. When a
signal falls between two “bins,” the energy is split between adjacent bins such that the reported “peak”
amplitude may be lower by as much as 0.8 dB. If frequency resolution and processing speed are more
important than absolute amplitude, use native bandwidths.
To get an accurate power reading using “Marker peak”, non-native bandwidths are recommended. They
integrate the power across several adjacent bins, eliminating scalloping loss.
To Next Page
Loading...
Need help?
General
