Tuning Sweep Settings
Depending on the various tinySA settings the time to do one scan can vary between below 20 milliseconds and above 100 seconds. Here are some measurements with different settings to demonstrate their impact on scan time.
First a measurement of the fundamental of the 30MHz cal output with a span of 4MHz and all other settings on default.
The important number to notice is the indicated scan time. In this case it is 339 milliseconds.
Sometimes small signals are visible and enabling the SPUR REMOVAL will help identifying if these are real signals.
As the small signals are no longer visible these where internally generated artifacts. Notice the SPUR REMOVAL doubles the scan time.
In case the noise level is too high it is possible to manually select an RBW, for example selecting 3kHz will reduce the noise level almost 10dB.
Notice the scan time increases to 16.8 seconds as a small RBW requires more frequencies to be measured and the time per measurement is longer.
One of the ways to reduce scan time is by selecting the FAST sweep mode. This gives some more noise.
This fast scan is only taking 63 milliseconds, roughly 4 times faster compared to default settings. Another disadvantage of the FAST scan is the reduce accuracy of the level measurement as you can see in all FAST measurements on this page.
Especially when selecting a small RBW the scan time reduction is relevant.
Again about 4 times faster as a normal scan with 3kHz RBW
Another way to reduce the measurement time on a small frequency span is by reducing the number of measurement points. As an example a measurement with only 51 point.
Notice reducing the number of points increases the automatic RBW selection from 28kHz to 168kHz but reduced the scan time to only 21ms. Ideal for tuning some circuit or filter.
Reducing noise is possible by averaging. This can be done on a per sweep basis showing both the actual sweep and the averaged sweep using the AVER 4 option.
This setting does not impact scan time but it may take some sweeps before a change in level is reflected in the calculated yellow trace.
Another way to reduce noise is by setting the SAMPLE REPEAT to above 1. Setting the sample repeat to 10 give almost the same improvement compared to the AVER 4.
As the sample repeat measures each point multiple times with some in between delay the scan time will increases.
The selected signal at -25dBm is far below the maximum input level and the automatic attenuation has selected 0dB attenuation. For much stronger signal the level of attenuation will be automatically increased. The impact of this increase in attenuation can be seen by manually setting the attenuation to a higher level such as 20dB.
As a consequence the noise floor has moved up with 20dB but the signal level measurement has not changed.
As there is sufficient headroom it is possible add a low noise amplifier (LNA) before the tinySA. Using a 22dB LNA the noise floor should drop considerably but due to the phase noise of the PLL and the noise generated by the CAL OUTPUT this noise reduction does not happen.