23.04 - First Batch Testing
LDO idea doesn't work.
For those in the same situation I was before starting this project, a linear regulator keeps current the same on the input and output so any excess power is disipated as heat. This makes them unsuitable for large voltage drops at anything more than a few milliamps. Below are notes of my experience of utter failure thinking an LDO would work for this application.
For the initial design of the FLIP-C3, an LDO from Seward Electronics was chosen due to its max suggested input voltage of 60v. The SE8650X2-HF has a current limit of 200ma so it was determined that four(4) of them be used in parallel considering the potential needs of the ESP32-C3, a WS2812B LED, a red LED plus some overhead.
The initial design duplicates the suggested circuitry for each LDO, so each input has a 1uF ceramic capacitor to ground while each output has a 10uF capacitor to ground.
The first batch of 10 arrived in early April and testing began with USB connectivity and the ESP32-C3's ability to run code. Aside from a missing pull-up resistor on the ENABLE pin, everything seems ok here.
Upon testing the DC input component, it was placed on a bench power supply (Riden RD6018) at about 9v which is near the low end of acceptable input voltages for the array. It behaved normally for several hours.
When raising the voltage I noticed extreme heat from both the ESP32 but especially the array of LDOs... in-fact too hot to touch. Going further seemed to elicit a thermal shutdown effect.
TODO: Connect RD6018 to Home Assistant for testing automation.
With an enabled ESP32-C3 running ESPHome and sending the following values to Home Assistant over API:
- 5v rail voltage
- 3v3 rail voltage
- Internal ESP32-C3 Temperature
- Normal temperature range on USB is around 40-50°C
- Red LED was on during testing
- 12v input ~ 60°C
- 14v input ~ 70°C
- 16v input ~ 78°C
- 18v input ~78°C
- 20v input ~ 78°C
- 0.39 - 0.99w
- Starts rebooting on its own, temps drop
- 19v input stable until attempted OTA update
Thanks to Arya for talking me through a good portion of this troubleshooting.
One board with the ESP32-C3 disabled was able to go up to 60v with minimal heat... what I would describe as "not cold." The 3v3 LDO was still connected and both put out their respective 5v and 3.3v at arbitrary input voltages up to the max.
A 100ohm 1/4w resistor was placed on the 5v rail. It should be a 50ma load, but it was shown as 0.02a on the Riden bench supply. This current is consistent through changing voltage, by the time we reach 40v, we're dumping almost 2 watts into the LDO to dissipate as heat. This is the cutoff where the output voltage started to breakdown.