0
\$\begingroup\$

What I'd like to do is to amplify a signal coming from an analog Video Transmitter. I've created a small circuit according to the application circuit description of the "High Gain Driver Amplifier" (Avago MGA-31189). Now I plan to match the input of the amp to the output of the Transmitter and the output of the amp to the antenna in order to get the maximum power. Here are the specs:

Transmitter: Analog Video Transmitter Frequency: 850 MHz Power: approx. 10mW Impedance: 50 Ohms

Amplifier: Avago MGA-31189 Gamma locations at 850MHz: Input: mag= 0.071, ang= -113.02 Output: mag= 0.039, ang= -134.54

Antenna: Cloverleaf Antenna Impedance: 50 Ohms

Q1: How do I transform the input and output reflection coefficients of this amplifier into impedances?

Q2: Once the equivalent impedances have been found, how can an LC network be formed to achieve the required impedance matching to minimize reflections?

\$\endgroup\$
  • \$\begingroup\$ What is your question? \$\endgroup\$ – winny Jul 18 '16 at 22:19
  • \$\begingroup\$ My question is, how can calculate the load resistance (of the input of the amp) and reactance by the given gamma locations in order to for instance calculate L and C for a L-Match Topology. \$\endgroup\$ – Laurin Jul 18 '16 at 22:40
1
\$\begingroup\$

Q1: How do I transform the input and output reflection coefficients of this amplifier into impedances?

This question is fairly straightforward, however it depends on the reference impedance of the amplifier in question. The input impedance can be calculated from the input reflection coefficient as follows:

$$ Z_{in} = Z_0 \frac{1 + \Gamma}{1 - \Gamma} = 50\frac{1 + 0.071 e^{-j113.2(\pi/180)}}{1-0.071e^{-j113.2(\pi/180)}} \approx 46.9 - j6.2~\Omega. $$

Similarly, \$Z_{out}\$ can be calculated as: $$ Z_{out} = Z_0 \frac{1 + \Gamma}{1 - \Gamma} = 50\frac{1 + 0.039 e^{-j134.54(\pi/180)}}{1-0.039e^{-j134.54(\pi/180)}} \approx 47.3 - j2.6~\Omega. $$

Both of these calculations assume a reference system impedance of \$Z_0 = 50~\Omega\$.

Q2: Once the equivalent impedances have been found, how can an LC network be formed to achieve the required impedance matching to minimize reflections?

Since the input and output impedances of your amplifier are quite nicely matched on the package I don't think you should really concern yourself with impedance matching them. This is for two reasons:

1: At the frequency you're working at (850 MHz) most lumped components will at or beyond their self-resonant frequency (SRF), so impedance matching using lumped elements is pretty much out of the question.

2: Transmission-line matching networks at this frequency will be quite large unless you're going to use a high \$\epsilon_r\$ substrate like alumina (\$\epsilon_r = 9.6\$).

|improve this answer|||||
\$\endgroup\$
  • \$\begingroup\$ Thank you very much for your detailed answer. I will measure the output with the amp in series and the antenna in the next few days with a spectrum analyzer. Then I'll know how much reflection (by comparing the typical gain of 21db) I get without matching the in- and the output of the amp. \$\endgroup\$ – Laurin Jul 19 '16 at 10:02
  • \$\begingroup\$ @Laurin You would be better served by measuring the output of the amplifier on a spectrum amplifier. What you see on a spectrum analyzer will be difficult to interpret because it will depend on the placement relative to the antenna, and you will be measuring power density instead of power. \$\endgroup\$ – Captainj2001 Jul 19 '16 at 11:45
  • \$\begingroup\$ Thanks for your advice. I know that a spectrum analyzer is not the ideal instrument for the job. But unfortunately I dont have spectrum amplifier available. I have a selfmade power meter though, but it's not very accurate. \$\endgroup\$ – Laurin Jul 19 '16 at 15:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.