Final Project

Final Project: PCB radio with improved RF amp

Goal

Optimization of the RF amp

For any amplifier there are 3 basic criterion on which it can be evaluated:

Optimizing for gain in this RF amplifier requires increasing the current gain of the CS amplifier. This is because the CE stage has a relatively low input impedance as it requires high biasing current. To achieve this higher current gain, we use an intermediary CC stage between the CS and the CE stages. When designed properly, this new stage adds 30 dB of gain. Choosing component values for this amplifier be covered in optimization section of this page.

Distortion in the amplifier sense is typically measured by total harmonic distortion, or THD. This is a measure of the relative amplitudes of the harmonics introduced by the amplifier. So if we are sending in a 1 kHz signal, its a measure of the 2kHz, 3kHz, ... signals introduced. These are introduced anytime the signal is made slightly off from a sinusoid. In our case this is not trivial to measure as the incoming signal is AM modulated. This means the distortion of audible frequencies has to be measured after the AM detector. Fortunately spice can measure distortion so see the next section for how this was designed

Spice driven optimization

Using spice, every component of the circuit can be optimized for the two factors we care about. Distortion can be measured on the output using the spice "fourier" command to automatically calculate the distortion at the intelligence frequency. By sweeping component values we can locate the best values for each piece of the circuit. Special care has to be taken with biasing resistors as both the ratio of the two resistors and the total magnitude matters and so they cannot be adjusted independently. After all that we end up the circuit schematic below.



The original RF amp injects 25% THD in simulation while this new amplifier is only at 2% THD. Note, any resistors that were shown to be particularly sensitive were replaced with a potentiometer to allow for on the fly tuning.

PCB layout

With the schematic designed and laid out, components have to be selected. All parts were bought from mouser. In general we limited ourselves to 0805 and above for the footprints to make soldering possible.

Layout was done in a modular fashion by laying out each stage separately and then connecting them afterwards. The full layout can be seen below and generally the signal flows from right to left. Aside from the new RF-amp, the only other piece is the lm386 audio amp and a diode detector.



Performance

The RF amp picks up substantially better in low-level environments, the exact magnitude of this increase is hard to determine but it appears to be on the order of 20-25 dB over the original design.

The measured THD appears to be 14%, which is a far cry from the theoretical 2% but still far better than the original system. The PCB was mostly without flaws aside from a single potentiometer that failed to attenuated signals. Hip-Hip-Hoorah

The presentation for this is here. It won't make much sense without Tanner or I yapping over it but no reason not to include it.