Pico or ESP32 for a shack automation hub? Keeping CAT, relays and rotator control stable around RF
- Replies 2
- Views 12
- Created
- Last Reply
Top Posters In This Topic
-
Mod_Mike 1 post
-
Mod_Chrissy 1 post
A better way to browse. Learn more.
A full-screen app on your home screen with push notifications, badges and more.
I'm planning a small shack controller and want to sort out the architecture before I build anything. The wish list is CAT polling and logging, a remote antenna switch with 4 to 6 ports, a rotator controller using the existing potentiometer feedback, an SWR/power readout, and a Direwolf APRS box on a Pi that shares some of this data. I'd rather not build one monolithic board that needs reflashing every time I change one function.
My current thinking is a Raspberry Pi running Hamlib's rigctld and Direwolf, with one or two microcontrollers as dumb, reliable I/O nodes. That could be a Pico for relays and ADC, and maybe an ESP32 where Wi-Fi is useful. The Pi would talk to the nodes over USB serial or MQTT, and logging software would talk to rigctld over TCP, which listens on port 4532 by default.
These are the points I want to check:
First, logic levels. The Pico and the ESP32 are both 3.3 V devices and their GPIOs aren't 5 V tolerant. Some radios give TTL-level serial and some give RS-232 levels, and Icom's CI-V is a single-wire open-collector bus. So I'm treating each radio's CAT interface as its own level-shifting problem and checking the manual before wiring anything.
Second, RF getting into the electronics. Relay drivers, ribbon cables and the I2C bus to an LCD or a Si5351 can all pick up RF on HF at higher power. I'm planning a metal enclosure, ferrites on the cables, small bypass capacitors at connectors, flyback diodes on every relay coil, and a separate supply or at least a filtered branch for the 12 V relay rail.
Third, measurement. For the SWR meter I'd use a directional coupler with detector diodes and read forward and reflected voltage on the ADC. The Pico's ADC is 12-bit but noisy, and the ESP32's ADC is known to be nonlinear unless calibrated. Diode detectors are also nonlinear at low power. I expect to need a calibration table made with a dummy load and a known-good meter rather than relying on the formula alone.
Does this split sound sensible, or is it overkill? Where would you draw the line between what lives on the Pi and what lives on the microcontrollers?
Link to comment
https://www.hamradiobase.com/forums/topic/11818-pico-or-esp32-for-a-shack-automation-hub-keeping-cat-relays-and-rotator-control-stable-around-rf/Share on other sites