board/teensy40: Add ADC support

[ardnew]

This PR is baselined on top of the Teensy 4.0 external interrupt support PR (#1484)

This PR adds analog input support (ADC1/ADC2) for Teensy 4.0 on 14 GPIO pins (A0-A13).

I mirrored the API used in machine feather-m4, assuming it is probably a good example for TinyGo convention.

This has been tested using a cheap potentiometer with src/examples/adc/adc.go. Note you must also change the pin from ADC0 to one of this machine's analog pins, e.g. A0. All other code can be left as-is (although the call to machine.InitADC() is unnecessary - it is implemented as an empty stub for backwards compatibility with existing code such as this example).

The following describes each analog pin, its equivalent digital pin, its IOMUX pad, and finally which ADC channel(s) it is connected to. It is defined in src/machine/board_teensy40.go:

// Analog pins
const (
	//  = Pin  // Dig | [Pad]      {ADC1/ADC2}
	A0  = PA18 // D14 | [AD_B1_02] {  7 / 7  }
	A1  = PA19 // D15 | [AD_B1_03] {  8 / 8  }
	A2  = PA23 // D16 | [AD_B1_07] { 12 / 12 }
	A3  = PA22 // D17 | [AD_B1_06] { 11 / 11 }
	A4  = PA17 // D18 | [AD_B1_01] {  6 / 6  }
	A5  = PA16 // D19 | [AD_B1_00] {  5 / 5  }
	A6  = PA26 // D20 | [AD_B1_10] { 15 / 15 }
	A7  = PA27 // D21 | [AD_B1_11] {  0 / 0  }
	A8  = PA24 // D22 | [AD_B1_08] { 13 / 13 }
	A9  = PA25 // D23 | [AD_B1_09] { 14 / 14 }
	A10 = PA12 // D24 | [AD_B0_12] {  1 / -  }
	A11 = PA13 // D25 | [AD_B0_13] {  2 / -  }
	A12 = PA30 // D26 | [AD_B1_14] {  - / 3  }
	A13 = PA31 // D27 | [AD_B1_15] {  - / 4  }
)

[deadprogram]

Please rebase this branch against dev to resolve merge conflicts. Thanks.

[ardnew]

rebased against dev

[aykevl]

I think it makes sense to change the API of ADC.Configure (for all chips) to the API you're using here. The reason is that an ADC actually has a number of important configurations for different tradeoffs:

• Reference voltage (often VCC, but sometimes 1.1V or some other value).
• Accuracy vs time. I know that the nrf chips have this tradeoff (where more accurate sampling takes a lot longer but is necessary if there is a big resistor on the input) and I would guess other ADCs have something similar.

Of course, it's not necessary to implement them all right away but with this I think it makes sense to add an ADCConfig to the peripheral.

[ardnew]

Agreed. However, unless you are wanting to maintain backwards compatibility, I would think the ADC peripheral should be like all other peripherals and receive its ADCConfig as argument to Configure, and completely ditch the ConfigureMode that I'm using here.

[ardnew]

@aykevl would you rather I keep this code as is (i.e., keep it API-compatible with other machine package implementations)?

or should I introduce a new API-incompatible func (*ADC) Configure(ADCConfig), and remove my current func (*ADC) ConfigureMode(ADCConfig), under the presumption that you are planning to change those others to match this new API-incompatible function

[aykevl]

I think it would actually be better to (temporarily) break compatibility and meanwhile also update the signature of the other ADC implementations.

[ardnew]

updated to use the more conventional API: func (*ADC) Configure(ADCConfig)

[ardnew]

rebased against dev

[ardnew]

rebased against dev (3eb9dca)

[deadprogram]

Looks like this PR is ready to go @ardnew

Do you want to squash commits together so I can merge, or would you prefer that I do so, or something else entirely?

[deadprogram]

I am going to squash/merge in the interest of getting this unstuck (which is entirely my fault for not doing sooner).

Thanks @ardnew