RF Levelling and Statistics

The levelling and statistics settings for an RF sweep determine how long each step goes, when the step is considered stable, and configure the feedback loop for power levelling.

Statistics

The statistics configuration tells the runner how to determine when a step is stable, minimum wait times, and statistics settings.

If the use_traditional_stats field is True, then the runner uses a statistical test to determine stability. Otherwise each step waits for the minimum amount of time then moves on to the next frequency point.

Statistics Example

key_0	key_1	value	type	comment
stats_settings	initial_wait	57600	float	time the program will spend warming up before the measurement at index 0 (units: seconds)
stats_settings	use_traditional_stats	TRUE	bool	If True: use kendal test and runs test to check stability
stats_settings	minimum_wait	1800	float	minimum waiting time at each frequency point in seconds
stats_settings	stats_window	1800	float	thermopile voltage must be stable over this time interval to proceed to next frequency point (units: seconds)
stats_settings	stats_Rcv	-3.4	float	used by runs test. higher = more stringent test
stats_settings	stats_Tcv	0.25	float	used by Kendall test. Higher = more stringent
stats_settings	measurement_interval	3	float	thermopile voltage sampling interval for computing stability stats in seconds

Levelling

The levelling configuration tells the runner what power signal to try and level to (DUT_power, monitor_power, etc), as well as configures the feedback loop. It also lets you set hard limits across all frequency points based on every power signal in the calorimeter.

Levelling Example

key_0	key_1	key_2	value	type	comment
levelling_settings	level_to		DUT_power	str	DUT_power
levelling_settings	HARD_MAX_dBm	DUT_power	11	float	# dont want the DUT getting about 11 dBm so hard max here
levelling_settings	HARD_MAX_dBm	monitor_power	0	float	# doesn’t matter when no monitor attatched but useful here
levelling_settings	HARD_MAX_dBm	calorimeter_power	1000	float	# set arbitrarily high so the DC bias doesnt trigger a shutdown
levelling_settings	HARD_MAX_dBm	RF_source_power	20	float	# sets a hard limit on the max source power across frequencies
levelling_settings	use_AM_levelling		TRUE	bool	If true, use an analog voltage source and AM input to level power
levelling_settings	use_GPIB_levelling		TRUE	bool	If true, use GPIB commands to level power
levelling_settings	AM_levelling_C		0.01	float	used for AM levelling. Should be between 0 and 1.
levelling_settings	AM_levelling_time		10000	float	how long to use AM levelling after GPIB levelling finishes (seconds)
levelling_settings	GPIB_levelling_C		0.05	float	used for signal generator levelling. Should be between 0 and 1.
levelling_settings	GPIB_levelling_time		2000	float	Duration of synthesizer levelling in seconds. After this time, AM levelling takes over
levelling_settings	max_source_power_change_dB		0.5	float	rf power will not change by more than this ammount in a levelling step
levelling_settings	AM_HARDMAX		0.2	float	in volts. Maximum amplitude modulation voltage.
levelling_settings	V_off_slow_min		2	float	in volts. If V_off slow is less than this number, assume an error has occurred and stop the experiment. Bolometers only.
levelling_settings	V_off_slow_max		4	float	in volts. If V_off slow is more than this number, assume an error has occurred and stop the experiment. Bolometers only.