To confirm the calibration of the Amplifier Test System (ATS), noise figure and gain were measured with three test systems using calibrated noise sources on the same amplifier-equalizer (AML
A412LE4011 _0904, SN 202). The motivation for this work arose from discrepancies found between noise temperatures measured with our Micronetics noise source, which was out-of-calibration, and a calibrated Agilent 346B noise source.
The three measurment systems used were:
System | | Noise Source |
NRAO Amplifier Test System (ATS) | | Agilent 346B SN SG442000102 Cal'd 2009-01-17 |
Rohde & Schwartz (R&S) FSU 046 Spectrum Analyzer | | Agilent 346B SN SG442000102 Cal'd 2009-01-17 |
Agilent 8975A Noise Figure Measurement System | | Agilent N4002A, SN SG44420128 (No calibration date, but system just a few months old) |
Results are given in the graphs below, and noise figures agree within ± 0.1 dB using the calibrated noise sources.
For all measurements, the amplifier was cooled to just above room temperature using a fan. For production acceptance measurements, it would be better to heat sink the amps to measure them all at about the same physical temperature.
R&S recommends using a preamplifier when measuring noise figures, which wasn't available, and they also default the input attenuation level to 0 dB. A graph below shows the 0.25 dB NF error most likely occuring from errors corrected the high noise figure of the R&S if the input attenuator is set to 5 dB.
Finally, the calibration of the 10 dB attenuator used between the noise source and amplifier for all tests, is given below.
Latest measurements with all three systems agree to +/- 0.2 dB. Note that earlier measurement with Micronetics NS and AML measurements differ. | |
Gain measured using noise sources (AML measured gain not included) | |
NF discrepancy when R&S input attenuator at 5 dB. | |
R&S input attenuator at 5 dB. | |
Measurement of Attenuator between noise source and amplifier under test | |
--
JohnEffland - 2009-05-08