NACK encountered while communicating with fan controllers

[ryan-summers]

During normal operation (no RF input, enabled RF output on CH0), Booster was observed to undergo a reset. Querying the service information retrieved the following data:

> service
Version             : v0.1.0-61-g7e78b6b
Git revision        : 7e78b6b8840e2328adc520b9a72dffb2fb8ac902
Features            : 
Panic Info          : panicked at 'called `Result::unwrap()` on an `Err` value: Interface(NACK)', src/chassis_fans.rs:109:64

Watchdog Detected   : false

booster/src/chassis_fans.rs

Lines 107 to 117 in 7e78b6b

	fn set_duty_cycles(&mut self, duty_cycle: f32) {
	for fan in self.fans.iter_mut() {
	fan.set_duty_cycle(max6639::Fan::Fan1, duty_cycle).unwrap();
	fan.set_duty_cycle(max6639::Fan::Fan2, duty_cycle).unwrap();
	}
	// Deem the fans to be spinning if the duty cycle is greater than 5%. This is to avoid
	// floating-point comparison errors at near-zero duty cycle. The driver should not be
	// configuring fans to a non-zero duty cycle below 20%.
	self.fans_spinning = duty_cycle > 0.05;
	}

Background:

I ran Booster over the weekend with an enabled RF channel and no RF input. The reset was observed very shortly after plugging in Booster's USB to my computer. It is unknown if these are related. Service data was retrieved immediately after the fault.

[hartytp]

@ryan-summers once there is a workaround for this, I'll reflash and report back on reliability

[ryan-summers]

I'm not entirely convinced this is related to #128. This seems much more to me like a grounding/ESD-related fault. The NACK occurred pretty much right as I plugged in the USB port, which doesn't seem to coincide with the fault described in #128

[ryan-summers]

Correction - I failed to notice that my ST-LINK programmer was plugged in to the same USB HUB as the booster user USB. I do not believe the fault occurred immediately as I plugged in the device. The reset I observed appears to actually be caused by the ST-LINK resetting Booster. It is unknown when the panic occurred.

[hartytp]

sorry, I posted on the wrong issue. Are you satisfied this was the ST-Link programmer? If not we should look at the ESD protection on Booster again.

[ryan-summers]

While I do not believe the ST-LINK / USB insertion caused this fault, Booster still registered a NACK on the I2C communication with the fans. I do not know if this occurred simultaneously to USB/ST-link insertion or not. I'll keep this issue around in case we see similar issues.

[hartytp]

Sounds like a good plan

[hartytp]

@ryan-summers now greg has access to a Booster, hopefully he can figure out what's going on. Thanks for the debugging

[ryan-summers]

I have temporary workarounds for this and #128 that I'll be testing over the weekend as well, so hopefully we can get you guys some usable firmware soon. Sorry for the difficulty in bringing this up, and thanks for spending the time to help out! :)

[hartytp]

Ace! Thanks again