Could older AVR-based development boards be compatible with CRSF for Arduino?

[ZZ-Cat]

In my Readme, I assert that legacy AVR boards (such as the Arduino UNO R3) aren't compatible with CRSF for Arduino.

Excerpt:

Development boards of yesteryear that were built around ATmega AVR microcontrollers such as the 328, 32u4, 2560 & 4809 are incompatible with CRSF for Arduino. Their processing capabilities (such as their limited 16 MHz CPU speed) simply are not enough to meet the requirements of the CRSF protocol.

Over time, I have reconsidered my original position on this claim & thought "Is there a possibility that older development boards such as the Arduino UNO R3, Arduino Mega 2560, & Arduino Micro could be compatible?" & "If there is the possibility of these older boards being compatible, what limiting factors could there be that could significantly impact the performance of CRSF for Arduino?"

I would like to investigate my original claim to see whether-or-not there is any truth behind this or I am unnecessarily gate-keeping.
If my investigation disproves my original claim, that could allow me to diversify CRSF for Arduino's compatibility even further than what I already have available to you.

My hypothesis is if there are any limiting factors, it would be the following:

• Flash memory space.
  CRSF for Arduino occupies 14824 bytes of flash memory when compiled for the Adafruit Metro M4.
  The ATmega328p (the microcontroller inside the Arduino UNO R3) has a flash memory of 32 kilobytes.
• Core clock speed is 16 MHz, which could limit the baud rate.
  CRSF requires a baud rate of 420000. If the clock speed is not high enough, this could reduce the baud rate to a value lower than 420000 - EG 115200.
• If the baud rate is severely limited, this might not provide sufficient bandwidth for the entire CRSF protocol.
• Due to the 16 MHz clock speed, this means that hardware UART baud rates are typically set to 115200.
  This means that I may not be able to achieve the required 420000 baud rate in order to satisfy the requirements of the CRSF protocol.

I already have a couple of legacy development boards on-hand:

• Arduino Micro (Microcontroller: ATmega32u4)
• Arduino UNO R3 (Microcontroller: ATmega328P)

These are the boards that I am using to test my hypothesis on.

Additional hardware:

• Controller: RadioMaster TX16S Max Edition
  • Transmitter Module: RadioMaster Ranger
    • ExpressLRS Version: 3.3.0
    • Packet Rate: 333 Hz
    • Power Settings: 25 mW
    • Switch Mode: 8 Channels, Full Resolution
    • Telemetry Ratio: 1:128
• Receiver: RadioMaster RP3 Diversity
  • Output: CRSF
  • Power Settings: 100 mW
  • To err on the side of caution, I am adding 5 V to 3.3 V level shifting on the UART data lines between my development boards & my RP3 receiver.

The version of CRSF for Arduino that I am using is Version 0.4.0, which enables telemetry feedback.
For the firmware, I am using the GPS Telemetry example with CRSF_DEBUG & VIEW_RC_CHANNELS enabled.
My RadioMaster TX16S is used to monitor GPS telemetry as well as sending RC data to my development board.
My Ranger & RP3 serve as the over-the-air pipe for the CRSF protocol.

I am conducting my investigation during my hiatus, over the month of August.
I will report my findings in the comments section below.
When my investigation is complete, I will announce my conclusion in the main discussions channel & put links to this investigation there & in my main Readme.

This should give me plenty of homework to do in my absence.
"Until next time..." --Kyle Hill.

[ZZ-Cat]

I have made some minor changes to CRSF for Arduino for the purposes of this investigation.
These changes are as follows:

• Adding the Arduino Mega 2560, Arduino Micro, & Arduino Uno to the Compatibility Table
  This allows me to flash CRSF for Arduino onto older AVR boards.
• Temporarily re-factor how Serial & Serial1 are allocated.
  Initially, I had forgotten that some AVR boards use a USB-to-Serial bridge because they lack native USB hardware.
  This is one implication against CRSF for Arduino, especially when the target board only has one hardware serial port available.
  With this being said, I have made it so that no matter what board I am using, CRSF for Arduino is always on pin 0 & pin 1.
• Static assert if the hardware serial buffers' size is smaller than 64 bytes.
  The CRSF protocol has a maximum size of 64 bytes. Anything smaller than this is immediately excluded from compatibility.
• Add Frames Passed & Frames Failed counters.
  These two counters keep score of how many CRSF packets have passed the CRC checksum versus the ones that have failed.
  This is only enabled when CRSF_DEBUG is enabled, & this is one of the things that I am recording throughout this investigation.
  In a perfect world, the number of frames that have failed should be zero, & the number of frames that have passed should be literally every CRSF frame that is received. However, the world isn't perfect. So, what I am looking for is a percentage of the number of failed frames versus the number of passed frames. The lower the percentage, the better.
• Increase wait time to 3 seconds.
  This is an aesthetic thing more than anything. This governs how long the wait time between the end of setup() to running loop().
  Setting the wait time to 3 seconds allows me to catch a glimpse of the start-up process before the main function blitzes all of that information away & starts hitting me with debugging data.

Absolutely none of these changes will make it into the Main-Trunk, as none of this is intended to be a part of CRSF for Arduino. I have done this only for the purposes of this investigation. Once my investigation is complete, all of these changes will be discarded.

Now, over to my control.
For my control, I am using my Adafruit Metro M4 Express. This is the main devboard that I have been developing CRSF for Arduino on, so I know everything works admirably. The version of CRSF for Arduino I am using, from today onwards & until my investigation is complete, it includes the aforementioned changes & it remains unchanged throughout the course of my investigation.
If I am to make any changes to CRSF for Arduino (most likely in my attempts to get it running on my Arduino Micro), I will post updates in this comments section about those changes & why I made those changes. Other than that, I will do my best to keep the firmware as it is from today until my investigation is complete.

Now, I said that I will be using both my Arduino Uno & my Arduino Micro. In hindsight, I had forgotten just how outdated my Uno was & I no longer have a USB cable that will fit it. All but one of my USB cables are all USB-C, with only one of my cables being USB Micro B, which I keep around for my other development boards that have a USB Micro B connector. Absolutely none of my USB cables have the right connector for my Arduino Uno. To give you a fair idea of how old my Uno is: It's an Arduino SRL (IE the old Arduino.org) board & I bought it about a year or two before the whole "Arduino SRL vs Arduino LLC" spat went down.
Because of this, my Arduino Uno is sitting this one out.

So, my Arduino Micro serves as the only AVR representative in my investigation. It's not ideal, because my palette is nowhere near as diverse as I would like it to be, & my own circumstances right now prevent me from expanding on this for the time being.
This will not stop me from revisiting this investigation in the future - I can always redo it later, if the need to do so arises.

Tomorrow, I am recording my baseline values with my Metro M4 Express, & I will post my results as a reply to this thread.

[ZZ-Cat]

When I did my initial Control above, I realised that it's my RC Channels firmware & not my GPS Telemetry example. So, I have decided to include both firmwares for testing my hypothesis.

So, here is the results of the GPS Telemetry on my Control...

Control:

• Branch: Investigating-AVR-compatibility
• Commit: 4d34b83
• Environment:
  • Editor: VSCode, version 1.80.2
  • IDE: PlatformIO, version 3.3.1
• Devboard: Adafruit Metro M4 Express
  • Core Clock: 120 MHz
  • UART: Serial1
    • Rx: Pin 0.
    • Tx: Pin 1.
• Firmware: GPS Telemetry example
  • Flags:
    • CRSF_DEBUG enabled
    • CRSF_DEBUG_FRAMES_CRC_PASSED enabled
    • GENERATE_RANDOM_GPS_DATA disabled
    • USE_DMA disabled, because it's decided to be problematic for some reason.
    • VIEW_RC_CHANNELS disabled
  • Initial Telemetry Values:
    • GPS
      • Latitude: -41.18219376°
      • Longitude: 174.949707°
      • Altitude: 100 cm
      • Ground Speed: 500 cm/s
      • Ground Course: 275.7999878°
      • Satellites in View: 4
• Transmitter: RadioMaster TX16S Max Edition Mk1.
  • Firmware: EdgeTX, version 2.10.0-nightly
  • Internal Module: Disabled (permanently =^/.~=)
  • External Module: RadioMaster Ranger
    • Dynamic Power: Off
    • Firmware: ExpressLRS, version 3.3.0
    • Packet Rate: 333 Hz
    • Switch Mode: 8 Channel Full Resolution
    • Telemetry Ratio: 1:128
    • Transmit Power: 25 mW
• Receiver: RadioMaster RP3 Diversity
  • Firmware: ExpressLRS, version 3.3.0
  • Serial output: CRSF
  • Telemetry transmit power: 100 mW

GPS Telemetry (from Control, as displayed on my RadioMaster):

• Location (GPS): 41°10'55.9"S, 174°56'58.9"E
• Ground Speed (GSpd): 18.0km/h
• Ground Course (Hdg): 275.79°
• Altitude (Alt): 1m
• Satellites in View (Sats): 4

Number of times my RadioMaster called out "Sensor Lost": 0

Serial Monitor Sample (from Control):

GPS Telemetry Example
[CRSF for Arduino | INFO] Devboard Name: Adafruit Metro M4 Express
Ready
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 13 | CRSF Frames Failed: 35
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 22 | CRSF Frames Failed: 72
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 33 | CRSF Frames Failed: 103
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 45 | CRSF Frames Failed: 137
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 55 | CRSF Frames Failed: 172
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 66 | CRSF Frames Failed: 204
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 77 | CRSF Frames Failed: 239
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 87 | CRSF Frames Failed: 274
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 98 | CRSF Frames Failed: 309
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 109 | CRSF Frames Failed: 341

Again, this is a really good baseline to go with.
For shits & giggles, I am throwing my Adafruit Metro M0 Express into the mix. This is the bare minimum that CRSF for Arduino is compatible with & adding my Metro M0 diversifies my baseline a little bit.

[ZZ-Cat]

Here are the results from my Adafruit Metro M0 Express...

Control:

• Branch: Investigating-AVR-compatibility
• Commit: 4d34b83
• Environment:
  • Editor: VSCode, version 1.80.2
  • IDE: PlatformIO, version 3.3.1
• Devboard: Adafruit Metro M0 Express
  • Core Clock: 48 MHz
  • UART: Serial1
    • Rx: Pin 0.
    • Tx: Pin 1.
• Firmware: RC Channels example
  • Flags:
    • CRSF_DEBUG enabled
    • CRSF_DEBUG_FRAMES_CRC_PASSED enabled
    • USE_DMA disabled, because it's decided to be problematic for some reason.
    • VIEW_RC_CHANNELS disabled
  • Initial Telemetry Values:
    • GPS
      • Latitude: 0.0°
      • Longitude: 0.0°
      • Altitude: 0 cm
      • Ground Speed: 0 cm/s
      • Ground Course: 0.0°
      • Satellites in View: 0
• Transmitter: RadioMaster TX16S Max Edition Mk1.
  • Firmware: EdgeTX, version 2.10.0-nightly
  • Internal Module: Disabled (permanently =^/.~=)
  • External Module: RadioMaster Ranger
    • Dynamic Power: Off
    • Firmware: ExpressLRS, version 3.3.0
    • Packet Rate: 333 Hz
    • Switch Mode: 8 Channel Full Resolution
    • Telemetry Ratio: 1:128
    • Transmit Power: 25 mW
• Receiver: RadioMaster RP3 Diversity
  • Firmware: ExpressLRS, version 3.3.0
  • Serial output: CRSF
  • Telemetry transmit power: 100 mW

GPS Telemetry (from Control, as displayed on my RadioMaster):

• Location (GPS): 0°00'0.0"N, 0°00'0.0"E
• Ground Speed (GSpd): 0.0km/h
• Ground Course (Hdg): 0.00°
• Altitude (Alt): -1000m
• Satellites in View (Sats): 0

Number of times my RadioMaster called out "Sensor Lost": 0

Serial Monitor Sample (from Control):

RC Channels Example
[CRSF for Arduino | INFO] Devboard Name: Adafruit Metro M0 Express
Ready
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 1 | CRSF Frames Failed: 28
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 1 | CRSF Frames Failed: 52
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 12 | CRSF Frames Failed: 83
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 24 | CRSF Frames Failed: 119
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 34 | CRSF Frames Failed: 154
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 44 | CRSF Frames Failed: 190
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 56 | CRSF Frames Failed: 222
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 68 | CRSF Frames Failed: 256
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 78 | CRSF Frames Failed: 290
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 90 | CRSF Frames Failed: 324

Remember, my Adafruit Metro M0 Express is a part of my Control - alongside my Adafruit Metro M4 Express.

[ZZ-Cat]

These are the results from my Metro M0 Express...

Control:

• Branch: Investigating-AVR-compatibility
• Commit: 4d34b83
• Environment:
  • Editor: VSCode, version 1.80.2
  • IDE: PlatformIO, version 3.3.1
• Devboard: Adafruit Metro M0 Express
  • Core Clock: 48 MHz
  • UART: Serial1
    • Rx: Pin 0.
    • Tx: Pin 1.
• Firmware: GPS Telemetry example
  • Flags:
    • CRSF_DEBUG enabled
    • CRSF_DEBUG_FRAMES_CRC_PASSED enabled
    • GENERATE_RANDOM_GPS_DATA disabled
    • USE_DMA disabled, because it's decided to be problematic for some reason.
    • VIEW_RC_CHANNELS disabled
  • Initial Telemetry Values:
    • GPS
      • Latitude: -41.18219376°
      • Longitude: 174.949707°
      • Altitude: 100 cm
      • Ground Speed: 500 cm/s
      • Ground Course: 275.7999878°
      • Satellites in View: 4
• Transmitter: RadioMaster TX16S Max Edition Mk1.
  • Firmware: EdgeTX, version 2.10.0-nightly
  • Internal Module: Disabled (permanently =^/.~=)
  • External Module: RadioMaster Ranger
    • Dynamic Power: Off
    • Firmware: ExpressLRS, version 3.3.0
    • Packet Rate: 333 Hz
    • Switch Mode: 8 Channel Full Resolution
    • Telemetry Ratio: 1:128
    • Transmit Power: 25 mW
• Receiver: RadioMaster RP3 Diversity
  • Firmware: ExpressLRS, version 3.3.0
  • Serial output: CRSF
  • Telemetry transmit power: 100 mW

GPS Telemetry (from Control, as displayed on my RadioMaster):

• Location (GPS): 41°10'55.9"S, 174°56'58.9"E
• Ground Speed (GSpd): 18.0km/h
• Ground Course (Hdg): 275.79°
• Altitude (Alt): 1m
• Satellites in View (Sats): 4

Number of times my RadioMaster called out "Sensor Lost": 0

Serial Monitor Sample (from Control):

GPS Telemetry Example
[CRSF for Arduino | INFO] Devboard Name: Adafruit Metro M0 Express
Ready
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 1 | CRSF Frames Failed: 28
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 3 | CRSF Frames Failed: 50
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 14 | CRSF Frames Failed: 83
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 24 | CRSF Frames Failed: 118
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 35 | CRSF Frames Failed: 151
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 46 | CRSF Frames Failed: 184
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 56 | CRSF Frames Failed: 222
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 67 | CRSF Frames Failed: 255
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 77 | CRSF Frames Failed: 293
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 88 | CRSF Frames Failed: 329

Remember, my Adafruit Metro M0 Express & my Adafruit Metro M4 Express are my Control.

Now that I have this done & out of the way, it is time to test my hypothesis on my Arduino Micro...

[ZZ-Cat]

Here are the results from my Arduino Micro...

Arduino Micro:

• Branch: Investigating-AVR-compatibility
• Commit: 4d34b83
• Environment:
  • Editor: VSCode, version 1.80.2
  • IDE: PlatformIO, version 3.3.1
• Devboard: Arduino Micro
  • Core Clock: 16 MHz
  • UART: Serial1
    • Rx: Pin 0.
    • Tx: Pin 1.
• Firmware: RC Channels example
  • Flags:
    • CRSF_DEBUG enabled
    • CRSF_DEBUG_FRAMES_CRC_PASSED enabled
    • USE_DMA disabled, because it's decided to be problematic for some reason.
    • VIEW_RC_CHANNELS disabled
  • Initial Telemetry Values:
    • GPS
      • Latitude: 0.0°
      • Longitude: 0.0°
      • Altitude: 0 cm
      • Ground Speed: 0 cm/s
      • Ground Course: 0.0°
      • Satellites in View: 0
• Transmitter: RadioMaster TX16S Max Edition Mk1.
  • Firmware: EdgeTX, version 2.10.0-nightly
  • Internal Module: Disabled (permanently =^/.~=)
  • External Module: RadioMaster Ranger
    • Dynamic Power: Off
    • Firmware: ExpressLRS, version 3.3.0
    • Packet Rate: 333 Hz
    • Switch Mode: 8 Channel Full Resolution
    • Telemetry Ratio: 1:128
    • Transmit Power: 25 mW
• Receiver: RadioMaster RP3 Diversity
  • Firmware: ExpressLRS, version 3.3.0
  • Serial output: CRSF
  • Telemetry transmit power: 100 mW

GPS Telemetry (from Arduino Micro, as displayed on my RadioMaster):

• Location (GPS): 0°00'0.0"N, 0°00'0.0"E
• Ground Speed (GSpd): 0.0km/h
• Ground Course (Hdg): 0.00°
• Altitude (Alt): -1000m
• Satellites in View (Sats): 0

Number of times my RadioMaster called out "Sensor Lost": 0

Serial Monitor Sample (from Arduino Micro):

RC Channels Example
[CRSF for Arduino | INFO] Devboard Name: Arduino Micro
Ready
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 0 | CRSF Frames Failed: 27
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 0 | CRSF Frames Failed: 54
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 0 | CRSF Frames Failed: 80
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 0 | CRSF Frames Failed: 104
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 0 | CRSF Frames Failed: 131
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 0 | CRSF Frames Failed: 160
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 0 | CRSF Frames Failed: 188
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 0 | CRSF Frames Failed: 215
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 0 | CRSF Frames Failed: 240
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 0 | CRSF Frames Failed: 266

Straight away, this is looking like my hypothesis is being confirmed. What's interesting is, my RadioMaster is receiving telemetry perfectly fine. I say "fine" in the context of the RC Channels firmware literally sending a bunch of zeroes back to the handset as telemetry. 🙃

Now, I am not one for jumping to conclusions. All this tells me is "I need more evidence". So, onwards & upwards. GPS Telemetry is up next...

[ZZ-Cat]

GPS Telemetry results are up...

Arduino Micro:

• Branch: Investigating-AVR-compatibility
• Commit: 4d34b83
• Environment:
  • Editor: VSCode, version 1.80.2
  • IDE: PlatformIO, version 3.3.1
• Devboard: Arduino Micro
  • Core Clock: 16 MHz
  • UART: Serial1
    • Rx: Pin 0.
    • Tx: Pin 1.
• Firmware: GPS Telemetry example
  • Flags:
    • CRSF_DEBUG enabled
    • CRSF_DEBUG_FRAMES_CRC_PASSED enabled
    • GENERATE_RANDOM_GPS_DATA disabled
    • USE_DMA disabled, because it's decided to be problematic for some reason.
    • VIEW_RC_CHANNELS disabled
  • Initial Telemetry Values:
    • GPS
      • Latitude: -41.18219376°
      • Longitude: 174.949707°
      • Altitude: 100 cm
      • Ground Speed: 500 cm/s
      • Ground Course: 275.7999878°
      • Satellites in View: 4
• Transmitter: RadioMaster TX16S Max Edition Mk1.
  • Firmware: EdgeTX, version 2.10.0-nightly
  • Internal Module: Disabled (permanently =^/.~=)
  • External Module: RadioMaster Ranger
    • Dynamic Power: Off
    • Firmware: ExpressLRS, version 3.3.0
    • Packet Rate: 333 Hz
    • Switch Mode: 8 Channel Full Resolution
    • Telemetry Ratio: 1:128
    • Transmit Power: 25 mW
• Receiver: RadioMaster RP3 Diversity
  • Firmware: ExpressLRS, version 3.3.0
  • Serial output: CRSF
  • Telemetry transmit power: 100 mW

GPS Telemetry (from Arduino Micro, as displayed on my RadioMaster):

• Location (GPS): 41°10'55.9"S, 174°56'58.9"E
• Ground Speed (GSpd): 18.0km/h
• Ground Course (Hdg): 275.79°
• Altitude (Alt): 1m
• Satellites in View (Sats): 4

Number of times my RadioMaster called out "Sensor Lost": 0

Serial Monitor Sample (from Arduino Micro):

GPS Telemetry Example
[CRSF for Arduino | INFO] Devboard Name: Arduino Micro
Ready
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 0 | CRSF Frames Failed: 23
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 0 | CRSF Frames Failed: 42
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 1 | CRSF Frames Failed: 67
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 1 | CRSF Frames Failed: 89
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 1 | CRSF Frames Failed: 113
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 1 | CRSF Frames Failed: 139
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 1 | CRSF Frames Failed: 163
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 1 | CRSF Frames Failed: 185
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 1 | CRSF Frames Failed: 210
[CRSF for Arduino | DEBUG] CRSF Frames Passed: 1 | CRSF Frames Failed: 231

See, now this is promising. I even have one CRSF frame pass the CRC check, here.
Also, my RadioMaster is receiving GPS telemetry perfectly fine. No "Sensor Lost", no zeroes, just exactly what I wrote in the GPS Telemetry example firmware.

So, where do I go from here?
My Arduino Micro is setup on a janky "solderless" breadboard, & I am only using a resistor divider on the Micro's Tx pin so it doesn't cook the Rx pin of the ESP8285 in my RadioMaster RP3 receiver.

When I am using my Adafruit Metro M4 Express or Adafruit Metro M0 Express, I am using my prototyping shield. This shield allows me to connect my RadioMaster RP3 Diversity receiver directly to my development board's UART Rx & Tx pins as well as the 5V & GND pins. These are solid connections & is more reliable than my old breadboard.

All this is telling me right now is, because I am using my old breadboard (which really needs to be binned because I have had numerous reliability issues with it in the past & I had totally forgotten about this), there is a high likelihood that it's skewing my results in favour of the status quo.

Later on in the year (around about November/December), I will put in an order for an Adafruit Metro 328.
This shoots two stones with one bird:

1. Adafruit Metro 328 has a selectable logic level between 5V & 3.3V. My RadioMaster RP3 Diversity uses 3.3V logic levels.
2. Adafruit Metro 328 has the Arduino R3 form factor, which makes it compatible with my prototyping shield. Thus, providing a stable connection between my RadioMaster RP3 Diversity & my development boards that I am testing.

So, for the time being, until November/December, I am putting my investigation on hold because I have done everything that I can do here with the limited resources that I have available to me.
When I pick this investigation up again, I will have more suitable hardware to test my hypothesis on.

"Until next time..." --Kyle Hill.

[ZZ-Cat]

I have finally reached my conclusion with this experiment.
There simply isn't enough dynamic memory available on all targets that use the AVR architecture.

Over on the Main-Trunk, somewhere between ce0fc7c and 86af1e1, I started using dynamic memory allocation.
Barring all of the other limitations that come with the AVR architecture, the tiny RAM size that comes with most AVR devices is now the deal-breaker.

This provides me with a single definitive answer to "Why does your library not work on my Arduino UNO?"
...and that's not enough RAM (AKA dynamic memory).
Both the Arduino IDE and PlatformIO will not compile CRSF for Arduino, because the size of its dynamic memory is larger than the maximum dynamic memory of every AVR-based development board in existence... including your Arduino UNO.