[PR #313] [MERGED] Keep fading slots polled and retry missed polls (#293) #283

New issue

Closed

opened 2026-10-22 05:08:41 +00:00 by roman · 0 comments

roman commented

2026-10-22 05:08:41 +00:00

Owner

📋 Pull Request Information

Original PR: https://github.com/safijari/openpuck/pull/313
Author: @safijari
Created: 10/14/2026
Status: ✅ Merged
Merged: 10/14/2026
Merged by: @safijari

Base: main ← Head: low-signal-polling

📝 Commits (1)

f88e173 Keep fading slots at full poll rate and retry missed polls

📊 Changes

1 file changed (+17 additions, -2 deletions)

View changed files

📝 OpenPuck/rf_link.cpp (+17 -2)

📄 Description

Addresses the low-signal disconnects in #293.

Cause (hypothesis, not yet confirmed from a capture): while another controller is live, a slot silent for more than 300 ms dropped to one poll every 50 ms, then to one poll every 2 s after 5 s. The controller drops the session if it hears no polls for a while, so backing off during a fade makes the fade worse. This matches the report that it mostly happens with two controllers, and the 3-5 s reconnect.

Changes

SLOT_QUIET_MS 300 -> 2500 ms: a fading slot keeps full-rate polling through short fades. A controller that sent an explicit power-off frame (0xF2) still backs off after 300 ms (new per-slot g_slotPoweredOff).
One retransmit of the E3 poll when a slot that replied within the last 300 ms gets no reply, only if it fits inside the 4 ms cycle.

Risk: a controller that vanishes without a power-off frame (dead battery, out of range) now gets up to 2.5 s of full-rate polling instead of 300 ms. With two controllers vanishing together this is the radio-duty pattern behind the #72 watchdog resets.

Testing: make check and make build pass. Flashed to a puck; not yet verified on hardware (palm over the USB-C end with two controllers; power one off while the other is live and watch for resets).

🤖 Generated with Claude Code

_{🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.}

## 📋 Pull Request Information **Original PR:** https://github.com/safijari/openpuck/pull/313 **Author:** [@safijari](https://github.com/safijari) **Created:** 10/14/2026 **Status:** ✅ Merged **Merged:** 10/14/2026 **Merged by:** [@safijari](https://github.com/safijari) **Base:** `main` ← **Head:** `low-signal-polling` --- ### 📝 Commits (1) - [`f88e173`](https://github.com/safijari/openpuck/commit/f88e17377580e1e3879ad4e7934e26815db87cca) Keep fading slots at full poll rate and retry missed polls ### 📊 Changes **1 file changed** (+17 additions, -2 deletions) <details> <summary>View changed files</summary> 📝 `OpenPuck/rf_link.cpp` (+17 -2) </details> ### 📄 Description Addresses the low-signal disconnects in #293. **Cause (hypothesis, not yet confirmed from a capture):** while another controller is live, a slot silent for more than 300 ms dropped to one poll every 50 ms, then to one poll every 2 s after 5 s. The controller drops the session if it hears no polls for a while, so backing off during a fade makes the fade worse. This matches the report that it mostly happens with two controllers, and the 3-5 s reconnect. **Changes** - `SLOT_QUIET_MS` 300 -> 2500 ms: a fading slot keeps full-rate polling through short fades. A controller that sent an explicit power-off frame (0xF2) still backs off after 300 ms (new per-slot `g_slotPoweredOff`). - One retransmit of the E3 poll when a slot that replied within the last 300 ms gets no reply, only if it fits inside the 4 ms cycle. **Risk:** a controller that vanishes without a power-off frame (dead battery, out of range) now gets up to 2.5 s of full-rate polling instead of 300 ms. With two controllers vanishing together this is the radio-duty pattern behind the #72 watchdog resets. **Testing:** `make check` and `make build` pass. Flashed to a puck; not yet verified on hardware (palm over the USB-C end with two controllers; power one off while the other is live and watch for resets). 🤖 Generated with [Claude Code](https://claude.com/claude-code) --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>

roman

2026-10-22 05:08:41 +00:00

closed this issue
added the
pull-request
label

No labels

pull-request

wontfix

No milestone

No project

No assignees

1 participant

Notifications

Due date

The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set

Reference

safijari/openpuck#283

No description provided.

Rows
Columns