lynchlab
@aoife_l22 · Joined April 2025
Engineering student · university
- Topics
- 15
- Replies
- 79
- Points
- 0
Forum Replies Created
Is it because the cylinder still holds air after the main valve closes? We had a lab demo where a cylinder kept pushing after power was cut.
Swapped to wpa-conf with priorities like above and carried it out and back twice, it picked up a lease each time within about 20 s and Hermetry showed the gap and then fresh readings. Raised the offline alarm to 30 minutes too, the walk is 15. Thanks!
Not this term, the soil sensor driver only builds on the old kernel apparently.
We did a board like this for a course project last term, here’s the list we ended with, everything else on the DevKit is nice-to-have:
- 3.3 V regulator good for 500 mA or more. The ESP32 pulls short peaks of several hundred mA when Wi-Fi transmits. We used an AP2112K-3.3; the AMS1117 everyone copies works but runs hot and needs more input headroom.
- 10 uF + 100 nF right at the module’s 3V3 pin, plus a 22 uF or bigger at the regulator output.
- EN pin: 10k pull-up to 3.3 V and 1 uF to ground, so it resets cleanly when power comes up. Without the cap ours booted randomly.
- GPIO0: 10k pull-up and a pad or button to ground. Held low at reset = download mode.
- Leave GPIO12 unconnected or pulled low. If something pulls it high at boot, the module thinks the flash is 1.8 V and won’t start.
- A 6-pin header for programming: 3V3, GND, TX, RX, EN, IO0.
For programming you don’t need USB on every board. Buy one USB-UART adapter (CP2102 or CH340), wire TX to RX and RX to TX, hold IO0 low, tap EN, and flash. Do this once per board. After that use OTA for updates so you never open the wall box again. The sensor on I2C goes on GPIO21/22 like on the DevKit, so your sketch doesn’t change.