Skip to content

Fix weak WiFi signal on BCM43602 (MacBook Pro 2015-2017) - #11616

Closed
andyholst wants to merge 1 commit into
omacom:quattrofrom
andyholst:fix-brcmfmac-nvram
Closed

andyholst wants to merge 1 commit into
omacom:quattrofrom
andyholst:fix-brcmfmac-nvram

Conversation

@andyholst

@andyholst andyholst commented Sep 13, 2026 •

Copy link
Copy Markdown

Problem

MacBook Pro 2015-2017 with BCM43602 (brcmfmac) gets -90 dBm signal — barely usable past 1 meter.

The root cause is the NVRAM config loaded by the brcmfmac driver is too generic. It lacks proper power amplifier settings and antenna gain parameters for the MacBook's hinge-antenna design.

What Didn't Fix It

  1. options brcmfmac feature_disable=0x82000 — fixes WPA handshake, NOT signal
  2. options cfg80211 ieee80211_regdom=US — fixes regulatory domain, NOT signal
  3. roamoff=1 — reduces scanning retries, NOT signal
  4. NVRAM config from Andy Holst (Bugzilla #193121) — helps channels, NOT signal

All help with connecting but NOT signal strength.

Fix

Install cristianmiranda's improved NVRAM config which has proper board flags, antenna gain, and power amplifier parameters specific to MacBookPro13,3:

  • boardflags=0x02000001 — proper board configuration (BT coexistence, PALDO)
  • boardflags2=0xC0000000 — LNA bypass for 2G/5G clip, spur WAR
  • boardflags3=0x40000108 — RCAL WAR, FEM table from NVRAM
  • pa2ga0, pa2ga1 — 2.4GHz power amplifier settings
  • pa5ga0, pa5ga1 — 5GHz power amplifier settings (all chains)
  • maxp2ga0=74 — max power limits
  • agbg0=2, aga0=2 — antenna gain
  • Correct MAC address substitution from the running interface

The NVRAM config file is bundled directly in the repo — no download needed at install time.

Result

Metric Before After
Signal -90 dBm -58 dBm
5GHz Unusable 360 MBit/s

32 dBm improvement — 5GHz now works reliably at full speed.

Changes

File What it does
install/hardware/apple/brcmfmac43602-pcie.txt The NVRAM config bundled in repo
install/hardware/apple/fix-brcmfmac-nvram.sh Detects BCM43602, installs config, rebuilds initramfs
install/hardware/all.sh Runs the new fix during hardware setup
migrations/1786391101.sh Applies fix to existing Omarchy installs
test/shell.d/brcmfmac-nvram-test.sh 18 tests, all passing

Testing

ok - the NVRAM fix script exists
ok - script detects BCM43602 by PCI ID
ok - script installs the correct NVRAM file
ok - script handles MAC address substitution
ok - script rebuilds initramfs
ok - the NVRAM fix runs during hardware setup
ok - the NVRAM config source file exists
ok - NVRAM config contains MAC address placeholder
ok - NVRAM config contains boardflags
ok - NVRAM config contains 2GHz PA parameters
ok - NVRAM config contains 5GHz PA parameters
ok - NVRAM config was installed
ok - MAC address was substituted into NVRAM config
ok - boardflags parameter present
ok - 2GHz PA parameters present
ok - 5GHz PA parameters present
ok - boardflags2 parameter present
ok - boardflags3 parameter present
all tests passed

Reference: #4692 (reply in thread)

Fixes #11613

- Detect BCM43602 (14e4:43ba) via lspci
- Bundle cristianmiranda NVRAM config in repo
- Install with MAC address substitution
- Wire into install/hardware/all.sh
- Add migration for existing installs
- Add 18 tests, all passing

Fixes omacom#11613
@omarchybot omarchybot added bug Something isn't working compatibility Make Omarchy work better on everything! labels Sep 27, 2026
@omarchybot

Copy link
Copy Markdown
Collaborator

Thanks for writing this up, @andyholst. I reviewed it at 955fac3c, by reading the branch against quattro and against the other open pull requests for this chip. Nothing ran on a worker, because no worker has a BCM43602 and the question here is which fix to carry, not whether this one's shell passes. The review was by Claude Opus 5.5, with Codex Medium as a second opinion, asked separately which of the competing fixes to bring forward; it chose the same one, though its independence is not guaranteed.

This pull request competes with #7671, and #7671 is the one to carry. Both install a BCM43602 calibration file for these Macs. #7671 does it only on MacBookPro14,2 and 14,3, under /usr/lib/firmware/updates/brcm/, and leaves any calibration already present alone. The draft #10141 builds on it and adds MacBookPro13,3 as a separate, model-gated file.

Why not this one:

Your measurement (-90 dBm to -58 dBm on your 13,3) is real data for the 13,3 question, and it sits next to a 13,3 report where the same file broke 2.4 GHz. If you can post your Wi-Fi card's subsystem ID (lspci -nnvd 14e4:43ba | grep -i subsystem) in #10141, it shows whether the two 13,3 results come from different boards. That is where the 13,3 work is being done.

Next: the maintainer decides between these pull requests. I am not bringing this one to ready alongside #7671.

@omarchybot

Copy link
Copy Markdown
Collaborator

Thank you for this, @andyholst. #7671 fixes the same problem in a way the review judged better, so this is closed in favour of it. If it does not cover your case, please say so there.

Closed at the maintainer's request. The review comment above has the details; it was done by Claude Opus 5.5 with Codex Medium as a second opinion, whose agreement is not independent.

@omarchybot omarchybot closed this Oct 1, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

bug Something isn't working compatibility Make Omarchy work better on everything!

Projects

None yet

Development

Successfully merging this pull request may close these issues.

BCM43602 WiFi signal extremely weak (-90 dBm) on MacBook Pro 2015-2017

2 participants