Add heltec-v4.3 board by Quency-D · Pull Request #9753 · meshtastic/firmware

Quency-D · 2026-02-26T08:54:50Z

This is the schematic diagram for v4.3.
The main difference between v4.2 and v4.3 is the different fem and the different control pins of the fem.
You can identify whether it is v4.2 or v4.3 by the default pull-up and pull-down resistors of GPIO2.
The fem control code is centralized into a single file.
The newly added fem lna is controlled on the screen.

🤝 Attestations

I have tested that my proposed changes behave as described.
I have tested that my proposed changes do not cause any obvious regressions on the following devices:
- Heltec (Lora32) V3
- Heltec (Lora32) V4
- Heltec (Lora32) V4-TFT
- Heltec T114

Quency-D · 2026-02-26T08:58:10Z

It needs to be combined with the new variables added in this commit for it to compile successfully.

fifieldt · 2026-02-26T09:36:35Z

Exciting!

weebl2000 · 2026-02-26T10:46:20Z

Made sure heltec tracker v2 can still build: Quency-D#1

robekl · 2026-02-26T17:21:53Z

This PR moves SX126x FEM control from USE_GC1109_PA gates to HAS_LORA_FEM gates, but only heltec_v4 defines HAS_LORA_FEM=1. For GC1109 boards, CSD/CPS control is required for correct TX/RX/sleep behavior (board comments already document this in variant.h). If these controls are skipped, behavior can degrade (wrong RF path/state, higher sleep current, reduced link reliability).

Quency-D · 2026-02-27T02:30:56Z

Made sure heltec tracker v2 can still build: Quency-D#1

Thank you. Only a very small number of samples of the Heltec Tracker v2 board were released before, and it has now been redesigned. I think it's appropriate to wait until the new design is available before making any modifications. Therefore, this submission will not be merged for the time being.

Quency-D · 2026-02-27T02:35:33Z

This PR moves SX126x FEM control from USE_GC1109_PA gates to HAS_LORA_FEM gates, but only heltec_v4 defines HAS_LORA_FEM=1. For GC1109 boards, CSD/CPS control is required for correct TX/RX/sleep behavior (board comments already document this in variant.h). If these controls are skipped, behavior can degrade (wrong RF path/state, higher sleep current, reduced link reliability).

Yes, your concerns are valid. However, currently USE_GC1109_PA is only used for heltec-v4 and heltec tracker v2. The heltec-v4 part has been modified in this submission. As for heltec tracker v2, since only a small number of samples were previously released and it has been redesigned, we plan to modify and submit it again after the new design is completed.

weebl2000 · 2026-02-27T02:39:06Z

Made sure heltec tracker v2 can still build: Quency-D#1

Thank you. Only a very small number of samples of the Heltec Tracker v2 board were released before, and it has now been redesigned. I think it's appropriate to wait until the new design is available before making any modifications. Therefore, this submission will not be merged for the time being.

Should we remove the Heltec Wireless Tracker v2 from Meshtastic firmware altogether?

If we merge this PR the heltec wireless tracker v2 build will fail. That's what my PR tried to address. :)

Quency-D · 2026-02-27T02:47:02Z

Should we remove the Heltec Wireless Tracker v2 from Meshtastic firmware altogether?

If we merge this PR the heltec wireless tracker v2 build will fail. That's what my PR tried to address. :)

I tried using this pr to compile the Heltec Wireless Tracker v2, and it compiled successfully. What error did you get during compilation? Could you post it so I can see?

weebl2000 · 2026-02-27T02:54:39Z

I tried using this pr to compile the Heltec Wireless Tracker v2, and it compiled successfully. What error did you get during compilation? Could you post it so I can see?

The build itself compiles, but the issue is at runtime — the PR removes all USE_GC1109_PA guarded code and replaces it with HAS_LORA_FEM guards, but only heltec_v4 defines HAS_LORA_FEM=1. Since the tracker v2 doesn't define it, the GC1109 FEM would never be initialized, TX/RX mode switching wouldn't happen, and deep sleep RTC hold logic would be skipped. It compiles fine but the FEM would be completely uncontrolled at runtime. Essentially it breaks Heltec Wireless Tracker v2. It's better to remove it then, or merge the PR suggestion I made.

Quency-D · 2026-02-27T03:27:56Z

The build itself compiles, but the issue is at runtime — the PR removes all USE_GC1109_PA guarded code and replaces it with HAS_LORA_FEM guards, but only heltec_v4 defines HAS_LORA_FEM=1. Since the tracker v2 doesn't define it, the GC1109 FEM would never be initialized, TX/RX mode switching wouldn't happen, and deep sleep RTC hold logic would be skipped. It compiles fine but the FEM would be completely uncontrolled at runtime. Essentially it breaks Heltec Wireless Tracker v2. It's better to remove it then, or merge the PR suggestion I made.

Let me think about it. Could you please explain how you perform the format checking? What tools do you use?

weebl2000 · 2026-02-27T03:57:55Z

The repo uses Trunk for CI lint/format checks — it runs automatically on every PR. For C/C++ files it uses clang-format (config is in .trunk/configs/.clang-format).

You can run it locally with:

trunk fmt     # auto-format files
trunk check   # check for lint/format issues

Quency-D · 2026-02-27T08:09:07Z

The repo uses Trunk for CI lint/format checks — it runs automatically on every PR. For C/C++ files it uses clang-format (config is in .trunk/configs/.clang-format).

You can run it locally with:
trunk fmt     # auto-format files
trunk check   # check for lint/format issues

Thank you for your answer. I fixed the FEM control in Heltec Tracker v2. I think this macro approach is also simpler and easier to understand. What do you think?

Copilot

Pull request overview

This PR adds support for the Heltec WiFi LoRa 32 V4.3 board, which uses a different FEM (KCT8103L) compared to V4.2 (GC1109), and centralizes FEM control logic into a new LoRaFEMInterface class. The V4.3 hardware is auto-detected at runtime by reading the default pull state of GPIO2. Additionally, a new LNA toggle menu is added to the screen UI for boards that support LNA control (KCT8103L).

Changes:

New LoRaFEMInterface class (src/mesh/LoRaFEMInterface.h/.cpp) unifies FEM control for all supported boards (GC1109 and KCT8103L), replacing board-specific USE_GC1109_PA conditionals throughout the codebase
Renamed FEM pin macros from generic (LORA_PA_EN, LORA_PA_TX_EN) to chip-specific (LORA_GC1109_PA_EN, LORA_GC1109_PA_TX_EN), and added KCT8103L pin definitions to heltec_v4/variant.h
Added a new LoRa FEM LNA toggle UI menu in MenuHandler.h/.cpp controlled by a new fem_lna_mode config field

Reviewed changes

Copilot reviewed 13 out of 13 changed files in this pull request and generated 12 comments.

Show a summary per file

File	Description
`src/mesh/LoRaFEMInterface.h`	New class declaration for unified FEM control interface
`src/mesh/LoRaFEMInterface.cpp`	New implementation with runtime FEM detection and per-chip control logic
`src/mesh/SX126xInterface.cpp`	Delegates FEM init/sleep/TX/RX to `LoRaFEMInterface`
`src/mesh/RadioInterface.h`	Conditionally includes new `LoRaFEMInterface.h`
`src/mesh/RadioInterface.cpp`	Uses `loraFEMInterface.powerConversion()` when `HAS_LORA_FEM=1`
`src/sleep.cpp`	Uses `loraFEMInterface.setRxModeEnableWhenMCUSleep()` instead of inline GC1109 code
`src/graphics/draw/MenuHandler.h`	Adds new `LoraFemLnaToggleMenu` enum value and method declaration
`src/graphics/draw/MenuHandler.cpp`	Adds LNA toggle menu for boards with controllable LNA
`src/configuration.h`	Adds default definition `HAS_LORA_FEM 0`
`variants/esp32s3/heltec_v4/variant.h`	Adds KCT8103L pin definitions alongside renamed GC1109 macros
`variants/esp32s3/heltec_v4/platformio.ini`	Adds `-D HAS_LORA_FEM=1`
`variants/esp32s3/heltec_wireless_tracker_v2/variant.h`	Renames FEM pin macros to chip-specific names
`variants/esp32s3/heltec_wireless_tracker_v2/platformio.ini`	Adds `-D HAS_LORA_FEM=1`

Copilot · 2026-03-03T09:15:43Z

src/mesh/RadioInterface.cpp

+        power = loraFEMInterface.powerConversion(power);
+    }
+#else
+// todo:All entries containing "lora fem" are grouped together above.


There is a leftover in-code TODO comment on line 928. This style of embedded todo comment should not appear in production code. Either the comment should be removed entirely or the suggested cleanup should be done as part of this PR.

Suggested change

// todo:All entries containing "lora fem" are grouped together above.

src/mesh/LoRaFEMInterface.cpp

src/graphics/draw/MenuHandler.h

src/mesh/LoRaFEMInterface.cpp

src/graphics/draw/MenuHandler.cpp

src/mesh/RadioInterface.h

src/mesh/LoRaFEMInterface.h

src/graphics/draw/MenuHandler.cpp

src/mesh/LoRaFEMInterface.h

src/graphics/draw/MenuHandler.cpp

src/mesh/RadioInterface.cpp

src/sleep.cpp

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

jp-bennett · 2026-03-04T01:39:57Z

src/configuration.h

 #ifdef USE_GC1109_PA
 // Power Amps are often non-linear, so we can use an array of values for the power curve
 #define NUM_PA_POINTS 22
 #define TX_GAIN_LORA 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 10, 10, 9, 9, 8, 7


It looks like this PR would lose this define, which is probably not intentional. Or if intentional, let's remove it.

This section was also used in the Heltec Tracker v2, so it should be retained.

Quency-D added 2 commits February 26, 2026 15:10

Add heltec-v4.3 board

717c6e2

Modify LNA control display content

aa66dd1

github-actions bot added needs-review Needs human review hardware-support Hardware related: new devices or modules, problems specific to hardware labels Feb 26, 2026

weebl2000 mentioned this pull request Feb 26, 2026

Heltec v4.3 support meshcore-dev/MeshCore#1857

Open

weebl2000 mentioned this pull request Feb 26, 2026

Heltec v4.3 support weebl2000/meshcore#3

Open

Fix Heltec Tracker v2 FEM control

788725c

Quency-D and others added 4 commits February 27, 2026 16:18

Use trunk to fix formatting issues.

fc9651c

Merge branch 'develop' into heltec-v4.3

2be5225

Optimize the fem initialization control logic.

1116217

Merge branch 'develop' into heltec-v4.3

1612abe

caveman99 requested a review from Copilot March 3, 2026 09:05

Copilot started reviewing on behalf of caveman99 March 3, 2026 09:05 View session

Copilot AI reviewed Mar 3, 2026

View reviewed changes

caveman99 requested changes Mar 3, 2026

View reviewed changes

src/graphics/draw/MenuHandler.cpp Outdated Show resolved Hide resolved

src/mesh/RadioInterface.cpp Show resolved Hide resolved

src/sleep.cpp Show resolved Hide resolved

Quency-D and others added 3 commits March 3, 2026 17:33

Update src/mesh/RadioInterface.h change #ifdef to #if

609694e

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Change LORA_PA_EN to LORA_GC1109_PA_EN.

09ba527

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Remove the NodeDB.h include.

f2720fa

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Quency-D and others added 5 commits March 3, 2026 18:12

Change tx_gain to a const variable.

5285c43

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Fixed the issue where ARCH_PORTDUINO lacked the NUM_PA_POINTS macro.

81a5ef5

Remove the comment.

020e95f

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Move #pragma once to the first line.

c06afb3

Merge branch 'develop' into heltec-v4.3

984658d

jp-bennett reviewed Mar 4, 2026

View reviewed changes

Quency-D added 2 commits March 4, 2026 10:30

Remove the FEM LNA control menu.

fc467bf

Add description for KCT8103L.

5464984

Quency-D requested a review from caveman99 March 4, 2026 03:06

Quency-D and others added 2 commits March 4, 2026 14:13

Merge branch 'develop' into heltec-v4.3

f96d035

Merge branch 'develop' into heltec-v4.3

3d8a066

thebentern approved these changes Mar 4, 2026

View reviewed changes

thebentern merged commit 5ced739 into meshtastic:develop Mar 4, 2026
15 of 65 checks passed

thebentern mentioned this pull request Mar 4, 2026

Add back FEM LNA mode configuration for LoRa #9809

Merged

Uh oh!

Conversation

Quency-D commented Feb 26, 2026

🤝 Attestations

Uh oh!

Quency-D commented Feb 26, 2026

Uh oh!

fifieldt commented Feb 26, 2026

Uh oh!

weebl2000 commented Feb 26, 2026

Uh oh!

robekl commented Feb 26, 2026

Uh oh!

Quency-D commented Feb 27, 2026

Uh oh!

Quency-D commented Feb 27, 2026

Uh oh!

weebl2000 commented Feb 27, 2026

Uh oh!

Quency-D commented Feb 27, 2026

Uh oh!

weebl2000 commented Feb 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Quency-D commented Feb 27, 2026

Uh oh!

weebl2000 commented Feb 27, 2026

Uh oh!

Quency-D commented Feb 27, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Mar 3, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

jp-bennett Mar 4, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Quency-D Mar 4, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

8 participants

weebl2000 commented Feb 27, 2026 •

edited

Loading

jp-bennett Mar 4, 2026 •

edited

Loading