New CODD Pick-up Amplifier

Note

• Need to measure Zin or S11 for each of the attenuator settings. Done. See: Measurements of input refection coefficients of series production amplifiers

Input impedance

Fig.1: Input impedance vs. frequency in MHz. Red: Re(Z), green: Im(Z).

Fig.2: Compensation circuit.

Fig.3: Calculated input impedance vs. frequency in MHz, with compensation. Red: Re(Z), green: Im(Z).

Fig.4: Measured input impedance vs. frequency in MHz, with compensation. Red: Re(Z), green: Im(Z).

Raw input impedance was measured with the HP8753D. Power -30dBm, stop frequency 1GHz, log frequency sweep, S₁₁ 1-port calibration with reference plane right at the input of the 1:4 impedance transformer at the input of the LNA.

Getting measured data out of the NA is a form useful for further treatment is cumbersome. The best I have found is to *plot* the list of frequency - real Z - imaginary Z triples into 7 HPGL files on a floppy. Then, on my PC, catenate the files, pass them through hpstrip to filter out the hard data, and touch up the errors using a text editor. Pay particular attention to values in μΩ, because hpstrip doesn't recognise the μ and reports a value one million times to big.

The LNA is the three-FET version. A two-FET version has been tried, with little difference. The raw measured complex impedance data is in file zin.dat. Note that the real part goes negative at about 80MHz (Fig.1).

A compensation circuit was added to flatten the impedance (Fig.2). The values and topology of the compensation circuit were determined with the aid of the Mathematica worksheet ~jeroen/afs/Mathematica/LNA_Zin.nb. Fig.3 shows the calculated input match after compensation. The measured complex impedance after the compensation is shown in Fig.4. The corresponding data file is zincmeas.dat.