Is a 3-bit synchronous (3 data lines + Clock) parallel input at 16.368 MHz feasible?

RE: Is a 3-bit synchronous (3 data lines + Clock) parallel input at 16.368 MHz feasible?

haakonsh — Mon, 24 Sep 2018 12:05:28 GMT

I'll talk to some guys tomorrow, and see what they say.

You might be able to use some external shift registers to capture the bitstreams.

RE: Is a 3-bit synchronous (3 data lines + Clock) parallel input at 16.368 MHz feasible?

sadara — Mon, 24 Sep 2018 11:40:07 GMT

1) I realize I can't use the radio without a 32Mhz HSE, I should be able to switch HSE frequencies in the future (with the MCU stopped obviously)

2) The sensor is under NDA. There is not really a protocol for the sensor, the data interface is just a 3-bit ADC sampled at 16.384Mhz.

After powering up the sensor, it starts outputting 3 bit signed data over a simple parallel interface (3 data lines and a clock line). There is no control protocol, just one way raw data on this bus. After taking a small sample of the data (2ms), cut power to the sensor. It's not like SPI interface or a UART protocol, just raw data.

RE: Is a 3-bit synchronous (3 data lines + Clock) parallel input at 16.368 MHz feasible?

haakonsh — Mon, 24 Sep 2018 08:21:53 GMT

1) No, I don't think so. I think you might be able to run the the MCU off an external 32.768MHz crystal, but a radio protocol is out of the question. Either way it's outside the specifications.

2) I need more information on the protocol and the sensor device in order to answer this one.

3) I do not think this is possible. At 16MHz you're at 1/4 the CPU speed which means that you only have 4 CPU cycles to read the state of a single pin, store it, and then clear the interrupt. I'll need to find the delay from the clock interrupt fires until you can start to execute an ISR as well as the time from a GPIO state change until the registers update, etc.

We need to know what sensor and protocol you're asking for.