USB-A Expansion Card Stops Working Until Unplugged
Summary
The USB-A Expansion Card stops functioning — devices are not recognized and the port provides no power — until the expansion card is physically removed and reinserted12. The issue affects multiple Framework laptop generations and occurs on both Linux and Windows34. On FW13 AMD models, the left-side ports (slots 1 and 2) are most commonly affected and the card appears completely powered off when the issue strikes3. On FW16, ports 5 and 6 lose connection after sleep/resume or repeated device plug/unplug cycles56. Some users traced the problem to a hardware defect on their mainboard and received replacement boards under warranty that resolved the issue completely78.
Symptoms
- USB-A devices not detected — plugging a device into the USB-A expansion card produces no response; no new device appears in
lsusbordmesg13 - No power on the port — devices with LEDs (flash drives, mice) do not light up; a USB power meter shows 0 V39
- Persists across reboots — powering the laptop off and on again does not restore the port; a full power-off with AC disconnected for 2+ minutes may help110
- Occurs after device disconnect — unplugging a USB-A device (flash drive, YubiKey, mouse dongle) from the card may leave the port in a dead state311
- Occurs after sleep/resume — waking from suspend may leave USB-A cards unresponsive56
- Occurs after shutdown — on some AI 300 systems, the USB-A card is dead after a fresh boot and must be reseated1213
dmesgerrors may appear alongside the issue314:ucsi_acpi USBC000:00: ucsi_handle_connector_change: ACK failed (-110)usb 2-2: USB disconnect, device number N
Affected Models
| Model | Evidence | Ports Affected |
|---|---|---|
| Framework Laptop 13 (11th Gen Intel) | Multiple reports101516 | Any slot |
| Framework Laptop 13 (12th Gen Intel) | Original tracking thread1; also after BIOS 3.061 | Any slot |
| Framework Laptop 13 (AMD Ryzen 7040) | Dedicated tracking thread with 30+ replies3; mainboard replacements confirmed78 | Left side (slots 1, 2) most affected |
| Framework Laptop 13 (AMD Ryzen AI 300) | Multiple reports on fresh units12131718 | Various slots |
| Framework Laptop 16 (AMD Ryzen 7040) | GitHub issues #154 and #5956; community reports1920 | Ports 5, 6 most affected |
| Framework Laptop 16 (AMD Ryzen AI 300) | GitHub issue #154 comments5 | Port 5 |
Root Cause
The issue appears to have multiple contributing causes depending on the model:
Hardware defect (FW13 AMD 7040)
Multiple users on FW13 AMD Ryzen 7040 confirmed the issue was caused by a defective mainboard. User 0xc0ffee received a warranty replacement board and the problem disappeared completely after 9+ months of testing7. User Markus12 and SuperTux88 reported identical resolutions with replacement boards821. The defect appears to affect the USB PD (Power Delivery) controller or retimer circuitry on the left side of the mainboard3.
USB power management state machine (FW13 AMD)
User SuperTux88 discovered that the USB-A port power-down behavior depends on what else is connected on the same side of the laptop21:
- USB-A cards stop working on the left side when on battery or when power is connected on the right side
- USB-A cards keep working when power is connected on the left side (front or back)
- Active expansion cards (HDMI, network, MicroSD, or even a second USB-A card) in the other left-side slot keep the USB-A card alive
- Passive cards (empty USB-C pass-through, empty SD card reader) do not prevent the power-down
This suggests the PD controller or USB retimer enters a low-power state when no active device is detected on the side, and fails to wake when a USB-A device is plugged in.
PD controller / retimer state corruption (FW16)
On FW16, GitHub issue #154 shows Cypress PD controller (CYPD) logs with HARD_RESET_SENT and TYPE_C_ERROR_RECOVERY messages when the port enters the dead state5. The port fails after repeated connect/disconnect cycles or after sleep/resume, suggesting the PD controller state machine gets stuck and cannot recover without a physical reseat that forces a full power cycle of the card.
Firmware interaction (FW13 12th Gen Intel)
The issue appeared after BIOS 3.06 on 12th Gen Intel, where some Type-C ports would not function in DP alt mode after the update1. Framework's Kieran_Levin noted that a full power-off with AC disconnected for 2 minutes could restore the ports, indicating EC/PD controller state that persists across normal reboots1.
Diagnosis
-
Verify the card works in another slot — move the USB-A card to a different expansion slot. If it works in the new slot, the original slot or its controller may be affected13.
-
Test with another USB-A card — swap the expansion card. If the problem follows the slot rather than the card, the issue is on the mainboard side1.
-
Check
dmesgfor errors when plugging in a device to the dead port3:sudo dmesg -wIf the port is completely dead, no new messages will appear when plugging in a device.
-
Test on a live USB distro (Fedora or Ubuntu) to rule out OS-level configuration issues3.
-
Check if TLP is installed — TLP's USB auto-suspend can cause similar behavior. Verify with
tlp-stator check if thetlpservice is running1. -
Test power-side dependency (FW13 AMD) — connect power to the same side as the USB-A card and check if the port stays alive21.
-
BIOS check — test whether the card is responsive in the BIOS setup screen. If it works in BIOS but fails after OS boot, this points to an OS/firmware interaction rather than pure hardware21.
Workarounds & Fixes
| Solution | Effectiveness | Notes |
|---|---|---|
| Reseat the expansion card13 | Always works temporarily | Remove and reinsert the USB-A card. Fastest immediate fix. Problem will recur. |
| Mainboard replacement7821 | Permanent fix for hardware defect | Open a support ticket with Framework. Multiple FW13 AMD 7040 users received replacement boards under warranty that fully resolved the issue. |
| Connect power on the same side (FW13 AMD)21 | Prevents occurrence | Keeping a charger connected on the left side (same side as the USB-A card) prevents the port from powering down. |
| Place an active card in the adjacent slot (FW13 AMD)21 | Prevents occurrence | HDMI, network, or MicroSD card in the other left-side slot keeps the USB-A port alive. A second USB-A card also works. |
| Use USB-C slot instead | Workaround | Move the USB-A card to a right-side slot (FW13 AMD), or a known-working port on FW16 (ports 2, 3, 4). |
| Full power cycle1 | May restore port | Disconnect AC, power off, wait 2+ minutes, then power on. Clears EC/PD controller state. |
| Disable TLP USB auto-suspend1 | May help if TLP is the cause | Add the USB device ID to TLP's deny list, or uninstall TLP entirely. Only applies to Linux systems with TLP installed. |
| Powered USB hub15 | Workaround | Use a self-powered USB hub connected to a USB-C port instead of relying on the USB-A expansion card. |
Resolution
Framework support (Matt_Hartley) acknowledged the issue in January 2026 and reverted the tracking thread's "Solved" status, noting that effective triage requires individual support tickets22. Users experiencing this issue should open a support ticket, link to the tracking thread, and request escalation to the Linux support team22.
For FW13 AMD 7040 users, mainboard replacement under warranty has consistently resolved the issue7821. Framework has not publicly commented on whether a firmware fix is planned for the non-hardware-defect cases.
GitHub issues remain open:
- Framework 16 Port 5 loses connection until reseated · Issue #154 — assigned to quinchou77 (Framework engineer)5
- USB 5 and 6 Port Reset on Framework 16 · Issue #596
Related
- [TRACKING] USB-A Expansion Card stops working until unplugged — Framework Community (main tracking thread, 29 replies, Jan 2023–present)
- [TRACKING] FW13 AMD USB-A expansion card powers down and does not wake up — Framework Community (32 replies, Feb 2024–present)
- [TRACKING] USB-A an ongoing problem? — Framework Community (68 replies, Jul 2023–present)
- USB-A Extension stops working after shutdown — Framework Community (FW13 AI 300, Jan 2026)
- USB-A expansion card disconnects after Linux shutdown — Framework Community (FW13 AI 300, Jun 2025)
- Expansion cards (USB-A) require unplug-replug before working on boot — Framework Community (FW13 11th Gen, Apr 2022)
- USB-C UCSI Errors on Boot — related known problem (UCSI errors may appear alongside USB-A issues)