danielje
commented
3 years ago Hi Dan, I'm just looking at the schematic, it looks like GPIO 04 and 05 are used for i2c (SDA and SDL respectively), pins 13 and 14 on the ESP12S (https://www.elecrow.com/download/ESP-12S_User_Manual.pdf), so I think the layout is correct. I'm looking at the Arduino code now- maybe the default pin assignments for i2c for ESP8266 on Arduino have changed (i.e. line 342) and you'll need to initialize including explicit pin assignments(?) I do recall having difficulty getting i2c to operate- specifically I would switch off power to AD5933 when not being used partially to save battery but also under the presumption that it would just add noise onto the analog circuitry- but when powered off the communication to ADG715 (analog switch array that controls configuration of the network) wouldn't work so configuration changes wouldn't happen; turns out the i2c pins on the AD5933 act like a really big capacitor and make edge transitions really slow when the chip isn't powered (I think maybe what you are describing as "logarithmic" in shape rather than sharp pulses), and so data wasn't communicated reliably. That might explain your observation if AD5933 somehow isn't getting powered- or maybe the same might happen if ADG715 on the same bus isn't powered correctly(?) I hope it works- I've been trying to find time to update everything (to address noise issues, have more consistency in how AD5933 samples signals of different frequencies (i.e. by operating it on an adjustable clock so you consistently get same number of samples at same part of applied cycle), and changing to ESP32 (and getting rid of the extra bluetooth)- hopefully this summer (but been saying that for two years now :-/ If you need a device working I have a small pile of them here...
djs8d
I'd messaged you several months back to request some documentation (thanks for providing- it was very helpful!).
I've managed to get a PCB of the Abe-stat and all components soldered on. I've been able to connect to the bluetooth app and complete the initial calibration for the DAC and transimpedance gains. However, I'm having some issues with the i2C bus for calibration of the network analyzer- the existing calibration data when you turn on the app (or directly through the code) returns 48 NaN values. I checked with an oscilloscope on the SDA pin on the ESP 12S- and it is not working correctly but does attempt to send data when connected- (sends logarithmic form rather than pulse form).
With no obvious soldering issues present, my next step was to replace the AD5333 chip to see if it may have been fried previously during testing. Before doing this, one last possibility crossed my mind that I wanted to run by you to see your thoughts. I checked the SDA 12S pinout, and it appears that the ESP 12S's dedicated SDA pin is IO2, but the design files we used in PCB production hook the SDA bus up to GPI04. I consider myself a novice with this stuff, but I believe this would cause the bus to malfunction, although I may be mistaken. Interestingly, GPIO4 would be in the correct position for SDA on an ESP 12E controller, which I saw referenced in a PDF of the wiring in the documentation. However, the 12E has extra pins that don't align with the PCB design... I wondered if you had previously used a 12E during testing and perhaps the design files weren't updated accordingly. Grasping at straws...
In any case, any guidance you could provide would be very helpful. I'm very excited about doing some good research with this board and hope I don't need to resort to using a $10,000 potentiostat! :)
Dan