This is what gets logged to the SD card with the power supply problem solved. The data has continuous sections, but lags are still too frequent.
This is Client serial after transmission, both sides on USB power. Minor interruptions, but far less frequent.
Both hubs are now running on USB power, and the SD card is off. This is Serial plotter.
Client Serial print after transmission, with client on USB power. Much cleaner!
JSON packaging also doesn't lag, and the data is streaming cleanly through the JSON serialization call
Checking the Serial output from the sensor bank shows the data is streaming without interruption before transmission.
Lag rates didn't change much when moving sections of code for the JSON serialization and low-speed sensor sample rate
Zoomed in look at the timed data shows the data is coming in timed pairs from the server via websocket, about 200Hz.
The continuous sections of data seem to indicate that the websocket is streaming data fast enough to capture features
First google sheets graph of SD logger data with time stamps assigned from the sensor hub. Choppy data indicates high interrupt and lag rate At this point in the development cycle, I stopped adding components to the circuit, and started working on optimizing the code to make the datastream continuous. Power supply issues were found and accommodated (temporarily), and SD write speed problems were identified, but not solved. See the Github for full documentation of the troubleshooting cycle.
Published:March 17, 2023