Hi there, just got a Teensy 4. I have a friend who's using the TI ADS7279 at 1MS/s sampling rate to gather sensor data. SPI serial communication is used for triggering and data transfer. At least 11bit resolution is needed.
=> My goal: Control the ADC via the Teensy's SPI interface, get the data, and send the data out via the USB interface to a computer. On that computer there''s a node.js serial port that displays the ADC output on a SVG canvas.
Data collection will work in fast bursts of ~20ms sampling time each, triggered by software according to the experimental setup. So ~20ms sampling, 5 ms preparation, ~20ms sampling and so on.
After 3-4 seconds the measurement could pause for a moment if that's necessary.
=> Question: Is the Teensy 4 powerful enough to run the ADS7279 at >=11bit resolution and >=1MS/s sample rate, and at the same time churn out that data over USB? It's ok if some individual samples get lost.
No further computation required. Although it would be awesome if there was enough performance left to compute rolling averages across 30-100 samples, or do some hashing.
Also, would this performance require low level development, or could I still use the default SPI libraries etc.?
Thanks for your help!
=> My goal: Control the ADC via the Teensy's SPI interface, get the data, and send the data out via the USB interface to a computer. On that computer there''s a node.js serial port that displays the ADC output on a SVG canvas.
Data collection will work in fast bursts of ~20ms sampling time each, triggered by software according to the experimental setup. So ~20ms sampling, 5 ms preparation, ~20ms sampling and so on.
After 3-4 seconds the measurement could pause for a moment if that's necessary.
=> Question: Is the Teensy 4 powerful enough to run the ADS7279 at >=11bit resolution and >=1MS/s sample rate, and at the same time churn out that data over USB? It's ok if some individual samples get lost.
No further computation required. Although it would be awesome if there was enough performance left to compute rolling averages across 30-100 samples, or do some hashing.
Also, would this performance require low level development, or could I still use the default SPI libraries etc.?
Thanks for your help!