RF design engineers need to fully characterize their low noise transistors and MMIC’s over a very large frequency bandwidth. The result of the characterization is a set of four noise parameters (Fmin, Rn, Gamma-opt), that allow the design of the highest sensitivity (lowest noise) amplifiers and receivers.
This is accomplished by controlling the source impedance presented to the transistor using highly precise automatic tuners and retrieving the data from a noise receiver. Traditional noise receivers and VNA’s can be used.
An ideal system should sweep extremely fast as several decades of bandwidths must be measured for optimum results. DC bias sweeps must also be included in the measurements. Advanced statistical data processing algorithms allow reliable data extraction that has been confirmed through transistor noise modeling and comparative measurements on established noise standards.
Below is a typical diagram of a noise extraction system.
Load Pull Systems and Setups
Once commercially available solutions were introduced, the typical load pull setup comprised a signal generator, a power divider, two RF power sensors, a power meter, some DC bias networks and two fundamental tuners. This diagram shows the main components in a traditional load pull setup.
Traditional passive CW load pull setup
The setup shown above could measure parameters such as input/output power, gain, voltage current, and efficiency. Designers soon developed new techniques for designing power amplifiers, therefore load pull test systems became slightly more complex as the user wanted to “see” how their devices would behave with modulated signals, more specifically multiple tone signals. Below is a diagram of a two-tone basic load pull setup.
Traditional two-tone passive load pull setup
Testing a device with an industry modulation standard, like CDMA, LTE or others can provide some helpful information. This does not affect the setup since the tuner can perform the same work as in the traditional load pull. With modulated signals the modulated signal generator and spectrum analyzer do most of the “work”. Designers are looking for information like error vector magnitude or adjacent power leakage vs impedance and frequency.
Modulated carrier passive load pull setup
This final diagram shows a vector load pull system, where a network analyzer is used to measure the A- and B-waves at both the input and output of a device. This setup is often used for hybrid load pull as the loss of the couplers limiting the gamma of the tuners can be corrected. An active injection on the output will “help” the passive tuner reach higher gammas. We don’t have the same issue on the input since we are optimizing for gain. Because we are able to measure the A/B-waves of the device we know the impedance needed to match for optimum gain at high power.
Vector load pull setup
Partner Alliance Manager – Canada
High speeds Hybrid Load pull is the main new feature
AU4850 Pulsed IV/RF Characterization System
Auriga Multiple-Band High-Power Bias Tees
Auriga PHD3100 Pulser Head Series
Cardiff Model Lite
Harmonic Load Pull
Integrated Pulsed I-V