Changes

|}

[https://github.com/rsaxvc/LMS6APRS/blob/master/docs/cc1050%20frequency%20calculator.ods?raw=true CC1050 Frequency Register Calculator based on 14.7456MHz crystal]

==Test Points==

|}

The goal is to draw an approximation of a sine-wave over time in the frequency domain. This is usually done by drawing a sine-wave with a DAC, low-pass filtering it, and feeding it into an FM modulator, but we don't have those parts.

The FeatherHAB approach involves using a transmitter with an asynchronous digital modulation input(CC1050 can do this), and oversamping it like a 1-bit DAC. I think this works as long as the input signal is much faster than the charge pump bandwidth, and if so, it should act like a low-pass filter. FeatherHAB uses a hardware timer to drive their digital modulation input, but they do so by sending a variable-length pulse every 19200 Hz.

We might be able to do this using a 19200Hz periodic timer interrupt to set the modulation pin and a one-shot timer to generate another interrupt to clear it. This will use all available timers on the ST7. We cannot use a timer to generate the pulses directly as the ST7 does not wire out any timers on the right pins.

The CC1050 has a pair of frequency register sets, this allows programming a new frequency then flipping to it using a bit - this takes 4x register writes(16 bits each). We could, like a stepped frequency radar, switch between a series of tones rapidly(perhaps 8x symbol rate). This would look like an unfiltered DAC version of a sine-wave with little stair-steps.

For these to be a usable tracker we're going to need ground stations. In the Kansas City metro, there's not many APRS digipeaters on 70cm, and the few there are use 1200 baud AFSK, not 9600 baud FSK.