Driver for Ads868x

[lukh]

Implementation for ADS868x ADC

[salkinium]

Overall nice driver!

[salkinium]

I've squashed your commits and refactored the driver to use resumable functions. I also added a simple example.
However, since I don't have the hardware at hand, I cannot test it, so if you have the time, please test it and fix the driver if necessary. Perhaps also adapt the example to something more useful.

[lukh]

Thanks, I tried to do it, but I failed ;)
I two weeks I 'll be able to test it !

[lukh]

OOps
I did a mistake there. I don't know how I did it, it appears I force pushed my old branch again.

[salkinium]

Fetch your remote first, then switch branch to the new origin branch, then you can modify you branch again.

[lukh]

I'll try the code next week, I got ~2 years of modm update to import the target fw. :)

[rleh]

Thanks for contributing the driver!

I think it might make sense to name the driver differently, since there are several ADCs from TI under the name ADS868x (🤦🏽‍♀️), of which only ADS8681, ADS8685 and ADS8689 (datasheet) fit with this driver.
The chips ADS8684 and ADS8688 (datasheet) have 4 and 8 channels respectively and need a different/extended driver.

[rleh]

Probably not 😆

[rleh]

Suggested change

	def init(module):
	module.name = ":driver:ads868x"
	module.description = """
	# ADS868x Driver.
	ADS868x 16-Bit, High-Speed, Single-Supply, SAR ADC Data Acquisition System
	With Programmable, Bipolar Input Ranges
	def init(module):
	module.name = ":driver:ads8681"
	module.description = """
	# ADS8681/ADS8685/ADS8689 Driver.
	ADS8681/ADS8685/ADS8689 16-Bit, Single-Channel, High-Speed, Single-Supply, SAR ADC Data Acquisition System
	With Programmable, Bipolar Input Ranges
	This driver does not work with ADS8684 and ADS8685 quad/octa channel ADCs.

[rleh]

@salkinium (How) is it possible with lbuild to name this driver :ads8681 and have aliases :ads8685/:ads8689?

[salkinium]

Yes, lbuild has aliases already built-in. Though in this case we probably want a silent alias, not sure if that is supported currently, or if lbuild just prints an empty warning for it still.

[rleh]

A non-blocking wait would be nicer here, like for example implemented here:

modm/src/modm/driver/inertial/adis16470.hpp

Lines 315 to 316 in 2b4855b

	const modm::ShortPreciseDuration tStall{std::chrono::microseconds(16)};
	modm::ShortPreciseTimeout timeout{tStall};

modm/src/modm/driver/inertial/adis16470_impl.hpp

Lines 61 to 67 in 2b4855b

	if (this->releaseMaster()) {
	Cs::set();
	}
	timeout.restart(tStall);
	RF_WAIT_UNTIL(timeout.isExpired());
	RF_WAIT_UNTIL(this->acquireMaster());
	Cs::reset();

[rleh]

Non-blocking wait would be nicer, see above...