Please specify what is the maximum AC injection signal voltage and current for the J2120 Line Injector? We need to measure PSRR of our DUT not only in small signal mode but also in large signal AC mode.
The Picotest J2120A Line Injector is a small signal injector, which is generally limited to less than 50mVpp-500mVpp depending on capacitance and operating current, but is designed for wideband measurement up to 10MHz in applications that require less than 5A DC current.
The measurement of large signal PSRR is a very different measurement, requiring a four quadrant wideband power amplifier. The measurement is more complex, since the large currents radiate and are easily picked up in PCB traces and output wiring, masking the measurement.
The bandwidth of the injector and the bandwidth of the measurement are two different things. The bandwidth of the injector will vary significantly depending on the loading that is presented to it. Measured into a 50 Ohm termination, the -3 dB bandwidth is approximately 5MHz. Since the measurement of PSRR is somewhat independent of the amplitude of the injection signal, we can measure PSRR above the -3dB frequency of the injector.
One of the benefits of the J2120A is that it has an open loop output stage, since almost all amplifiers will become unstable if there is any loading capacitance. This is true of even the high performance amplifiers I mentioned. One such amplifier is the NF-BA4850 , which can produce 0-14.2V from DC-50MHz. The amplifier can provide up to 1A or 2A into capacitance up to 200pF. This amplifier sells for approximately $25,000 USD from NJ in Japan.
The measurement of PSRR is one of the most difficult regulator measurements, as it is significantly dependent on the input source impedance and requires a very substantial noise floor for the new high performance regulators. We have successfully measured two of the best, one being our own high PSRR regulator with >100dB rejection and the other is the TI TPS7A8000, with more than 70dB at high frequency.
Our injector has a 3dB bandwidth of approximately 100Hz-10MHz into a resistive (50 Ohm) load. It has a useful range that does extend to 30MHz-40MHz.
The difficulty in this measurement is not with our injector, but the noise floor, interconnects and the signal level. We can measure our own high PSRR regulators, which have a low frequency rejection of approximately 130dB and a 10MHz rejection of approximately 60dB or so. This is not close to our 140dB+ noise floor, but consider that with a 200mV input signal, 120dB is a 200nV output signal, which is quite small. Printed circuit board traces, interconnecting wires and local PCB grounding all impact performance. Depending on the regulator, the source impedance can also play a large role. We use the BODE 100 for our PSRR measurements, and I believe you do as well, is this correct?