I have two OpenEVSE chargers (and a third on backorder!). The older one has an OpenEVSE v5 board and originally had wifi firmware 3.1.0.dev. The WiFi board looks like the one pictured in the store (and at https://github.com/OpenEVSE/ESP32_hardware/blob/main/OpenEVSE_WiFi/OpenEVSE_ESP32_WiFi.jpg ) and has a 8-pin JST connector at top for expansion, including I2C.
The newer charger has an OpenEVSE v5.5 board and originally had firmware 5.0.1 on it.
What's puzzling to me is that the WiFi board for the newer charger *doesn't* look like the photos of the WiFi board in the store. Instead of an 8-pin JST header at top, it has a 3-pin QWIIC I2C header.
A couple of questions:
1. First, I'm hand-rolling my own PIN pad for the charger, based on the RFID work happening upstream (see https://github.com/cscott/openevse_esp32_firmware/tree/csa-pin ), since we have a semi-public charger and don't want to have to give out physical RFID keys to our users. It uses a CAP1214 touch sensor communicating via I2C and works flawlessly on the older board -- but it's seeing the touch sensor at all on the newer charger. Did the I2C port perhaps change pins or otherwise alter in the "newer" ESP32 wifi board?
2. Is this a permanent change (a v2 of the WiFi board, perhaps?) or was it a temporary manufacturing substitution? Since I've got a third charger on backorder right now, I'm wondering whether it is going to arrive with the "old style" or "new style" ESP32 wifi board.
3. If this /is/ a new revision of the WiFi board, could we pretty-please have schematics and a pin out diagram uploaded to https://github.com/OpenEVSE/ESP32_hardware ? Thanks!
The 8 pin connector was phased out due to WAY too many cases of the 6 pin power/comms plug from the OpenEVSE controller connected into the 8 pin socket. This causes both the OpenEVSE controller and WiFi module to release all its magic smoke. We thought it was obvious the 6 pin connector goes into the 6 pin socket, however you can shove it into the 8 pin socket if you are determined - centered, offset left or offset right.
We first removed the 8 pin socket, leaving the solder pads, but later revised to the 4 pin JST plug in the Sparkfun QWIIC / Adafruit STEMMA QT format.
The 4 pin connector is the newest revision. We are standardizing on Sparkfun QWIIC 1.0mm JST SH for 3.3v I2C and the 2.0mm JST PH for 5v I2C.
Sounds good! Seems like you are moving away from the DS3231 on the LCD (5V I2C) and instead putting an MCP9808 temperature sensor on the ESP32 board (3.3V I2C)? My ESP32 board had an unpopulated footprint for the MCP9808 but a separate Adafruit breakout board for the MCP9808 hung off the LCD.
WRT my question #1 above -- there was no problem with the QWIIC I2C on my ESP32 board, looks like I just didn't seat the cable correctly the first time.
I'm attaching a photo of my capacitive keypad board, in case anyone else is interested in PIN input as an alternative to RFID tags.
Yes, we discontinued using the DS3231 several years ago. For not much more cost than the DS3231 chip and battery, we can include WiFi and get time over the network.
The MCP9808 is coming off the WiFi module. The WiFi chip runs too hot and reads about 10C higher than the actual enclosure temperature.
The external Adafruit board for the display was temporary. For about a year it was impossible to buy the MCP9808, but Adafruit had the modules in stock.
The keypad is really nice. We like the idea...
Happy to share code & design files! I put a footprint for a DS3231 on the keypad board because I have dodgy connectivity at one of my chargers and I thought it might be worth having a more reliable RTC, but I don't know if that was really necessary. Probably replacing that with a MCP9808 would make more sense.
How much extra current is available from the 3.3V rail on the WiFi board? I'm running 12 backlight LEDs at 12mA each which adds up. I had a regulator footprint on my board to drop that down from the 5V rail, but it seems like the 5V supply is even more constrained!