Interactive Troubleshooter

Published: Sep 3, 2026

USB4 & Thunderbolt 5 Device Not Detected: Fix Guide

Time: 15 mins Level: Beginner

By StepDiag Engineering Team

Reviewed by Ibraheem Taofeeq Opeyemi

An interactive root-cause diagnostic guide to fix undetected USB4 and Thunderbolt 5 devices, docks, eGPUs, and storage drives. Follow step-by-step solutions to resolve PCIe tunneling failures, DMA security blocks, cable PAM3 signal degradation, and driver power state issues.

Full Diagnostic Tree & Step-by-Step Overview

What specific behavior or error occurs when you plug in your USB4 or Thunderbolt 5 device/dock?

Device gets power, but Windows displays 'USB device not recognized' or shows a yellow bang (Code 10 / Code 43) in Device Manager.
Device connects initially, but frequently disconnects, drops performance, or crashes during high-bandwidth tasks (transfers/displays).
Displays connected to a USB4/Thunderbolt 5 hub output video correctly, but PCIe peripherals (NVMe SSDs, eGPUs, 10GbE LAN) remain missing.
The device is completely unpowered (no LEDs), laptop fails to charge, or Windows states 'Display connection might be limited' / 'PC isn't charging'.

When inspecting Device Manager or system logs during the detection failure, what error condition is present?

Device Manager reports 'Code 10 (This device cannot start)' or 'Code 43' on the USB4 Root Router or Thunderbolt NHI Controller.
The device is visible in Device Manager, but Kernel DMA Protection or security policies block PCIe tunnel authorization.
The USB4 Host Router stays locked in a low-power D3cold idle state and fails to wake up upon device insertion.

Reset USB4 Host Router & Reinstall Thunderbolt NHI Controller Drivers

Solution:

Root Cause: Corrupted USB4 Native Host Interface (NHI) Driver State or Enumeration Timeout

USB4 and Thunderbolt 5 host controllers (such as Intel Barlow Ridge or integrated AMD USB4 routers) rely on the Native Host Interface (NHI) driver stack to manage high-speed domain tunneling. When a power surge, fast-startup resume glitch, or firmware mismatch occurs, the host router enters an unrecoverable Code 10 or Code 43 error state, refusing to negotiate new PCIe or DisplayPort tunnels.

Step-by-Step Fix:

1.Uninstall Host Router & NHI Controllers in Device Manager:

Press Win + X and select Device Manager.

Expand Universal Serial Bus controllers and System devices.

Right-click USB4 Host Router (or Thunderbolt(TM) Controller / NHI) and select Uninstall device.

Check the box for *Delete the driver software for this device* (if available) and click Uninstall.

2.Perform System Bus Rescan via PowerShell:

Open PowerShell as Administrator and run the following command to reset PnP enumeration:

powershell
     pnputil /scan-devices

3.Re-initialize USB4 Host Stack:

Restart the PC. Windows will reload the inbox usb4bus.sys and nhidriver.sys drivers.

4.Consult OS Technical Guides for Stack Architecture:

Review technical specifications detailed in the Microsoft Official USB4 Troubleshooting Guide regarding host router domain state diagnostics.

Configure Kernel DMA Protection & Authorize Thunderbolt 5 Peripheral Security

Solution:

Root Cause: Kernel DMA Protection Blocking Unauthorized PCIe Tunnel Access

Thunderbolt 5 and USB4 utilize Direct Memory Access (DMA) to grant high-speed PCIe peripherals direct access to system RAM. To prevent DMA side-channel attacks, modern operating systems enforce Kernel DMA Protection (e.g., Intel VT-d / AMD-Vi IOMMU policies). If a connected peripheral's firmware lacks strict DMA remapping support or if OS peripheral authorization is disabled, the system drops the PCIe tunnel while allowing basic USB 2.0/3.2 fallback.

Step-by-Step Fix:

1.Verify Kernel DMA Protection Status in Windows:

Press Win + R, type msinfo32, and press Enter.

Check System Summary > Kernel DMA Protection. Ensure it reports On.

2.Adjust Peripheral Authorization Settings in Windows 11:

Open Settings > Privacy & security > Windows Security > Device security.

Click Core isolation details and verify Memory integrity is enabled.

Open Settings > Bluetooth & devices > USB > USB4 hubs and devices.

Locate the connected device and toggle Allow connection to Always.

3.Verify Security Level in BIOS (SL0 vs SL1):

Enter UEFI BIOS (Del/F2) > Advanced > Thunderbolt/USB4 Configuration.

Ensure Security Level is set to User Authorization (SL1) or No Security (SL0) if using non-DMA-mapped legacy expansion chassis.

Disable PCIe ASPM Power State Suspension on USB4 Downstream Ports

Solution:

Root Cause: USB4 Host Router Trapped in D3cold Low-Power Idle State

To save laptop battery power, aggressive Active State Power Management (ASPM) puts USB4 host routers into deep D3cold sleep states when idle. When a Thunderbolt 5 or USB4 device is plugged in, the host controller must receive an ACPI-generated sideband wake interrupt (via CC pins) to wake up to D0 state. If power management settings prevent this wake interrupt, the port remains dormant.

Step-by-Step Fix:

1.Disable USB Selective Suspend in Windows Power Options:

Open Control Panel > Power Options > Change plan settings > Change advanced power settings.

Expand USB settings > USB selective suspend setting and change it to Disabled.

Expand PCI Express > Link State Power Management and set to Off or Moderate power savings.

2.Modify Driver Power Management Properties:

Open Device Manager > expand Universal Serial Bus controllers.

Right-click USB4 Root Router (or USB Root Hub (USB 3.0)) > Properties > Power Management tab.

Uncheck Allow the computer to turn off this device to save power.

3.Disable Fast Startup in Windows:

Fast Startup preserves corrupt sleep states across power cycles. Open Control Panel > Power Options > Choose what the power buttons do > uncheck Turn on fast startup.

Under what operational workload or physical setup does the USB4 / Thunderbolt 5 device disconnect?

Disconnections occur when using a passive USB-C cable longer than 0.8m or an older non-PAM3 USB 3.2 cable.
Drops occur under high power loads (e.g., charging a high-wattage laptop while driving 8K displays or eGPUs).
Connection drops when daisy-chaining multiple high-bandwidth Thunderbolt/USB4 devices through a single hub.

Upgrade to PAM3 80Gbps/120Gbps Electronically Marked (E-Marker) USB4 Gen 4 Cables

Solution:

Root Cause: PAM3 Signal Attenuation & Physical Layer Cable Mis-specification

Thunderbolt 5 introduces PAM3 (Pulse Amplitude Modulation 3-level) signaling to achieve 80Gbps bi-directional or 120Gbps asymmetric bandwidth. Unlike legacy NRZ (Non-Return-to-Zero) signaling used in Thunderbolt 3/4 and USB 3.2, PAM3 requires extremely strict high-frequency impedance matching. Using passive USB-C cables longer than 0.8m, non-E-marked cables, or legacy 10Gbps/20Gbps cables results in massive bit-error rates (BER), causing physical layer link collapse under sustained traffic.

Step-by-Step Fix:

1.Identify Cable Marking and Physical Rating:

Inspect the cable connector housing for official 80Gbps or 120Gbps logos alongside the Thunderbolt 5 or USB4 80Gbps emblem.

2.Verify Electronically Marked (E-Marker) Chip Output:

Use a hardware cable tester or OS software (e.g., USB4 Hubs App) to check if the cable reports an active E-Marker IC capable of 20V/5A (100W/240W) and Gen 4 (PAM3) operations.

3.Enforce Cable Distance Constraints:

For passive Thunderbolt 5 connections, ensure cable length does not exceed 0.8 meters.

For longer runs (up to 2.0 meters), switch exclusively to Active Optical or Active Re-timer Thunderbolt 5 Cables.

4.Review Hardware Standards:

Learn more about physical layer requirements via the Intel Official Thunderbolt Technology Overview.

Resolve USB PD 3.1 Extended Power Range (EPR) Over-Current Trips

Solution:

Root Cause: USB Power Delivery 3.1 Extended Power Range (EPR) Negotiation Collapse

Thunderbolt 5 docks support high-wattage host charging (up to 140W–240W) using USB PD 3.1 EPR (Extended Power Range) operating at 28V, 36V, or 48V profiles. When the dock attempts to deliver peak wattage while simultaneously powering internal PCIe buses, improper power supply decoupling or voltage ripple causes a brownout on the Configuration Channel (CC) lines, resetting the USB PD controller and dropping the connection.

Step-by-Step Fix:

1.Verify Dock Power Adapter Wattage Headroom:

Ensure the external DC power brick connected to your Thunderbolt 5 dock meets or exceeds the dock's peak rated input (e.g., 180W to 330W bricks).

2.Isolate Power Charging from Data Connection (Temporary Isolation Test):

Connect the laptop's OEM AC power adapter directly to a separate dedicated charging port on the laptop.

Plug the Thunderbolt 5 dock into the secondary USB4 port to test if eliminating dock power pass-through stops data disconnections.

3.Update Power Delivery Controller Firmware:

Download the latest dock firmware utility from the manufacturer (e.g., CalDigit, Lenovo, Dell, Anker) and flash updated USB PD controller microcode.

Reconfigure Inter-Domain Routing & Daisy-Chain PCIe Bandwidth Topologies

Solution:

Root Cause: Inter-Domain Routing Overhead & PCIe Adapter Mapping Over-Subscription

USB4 Version 2.0 and Thunderbolt 5 support multi-host Inter-Domain routing and multi-tier daisy-chaining. However, each hop through a downstream USB4 hub introduces protocol translation latency and consumes PCIe protocol adapters within the host router. Daisy-chaining high-bandwidth devices (such as a Thunderbolt 5 NVMe array through an eGPU chassis) exceeds available PCIe lane mappings, forcing the host router to drop downstream domains.

Step-by-Step Fix:

1.Flatten Connection Topology:

Connect high-bandwidth PCIe devices (eGPUs, dual-NVMe enclosures) directly to dedicated root ports on the laptop/PC motherboard rather than through a secondary hub.

2.Re-order Daisy-Chain Nodes:

Place highest-bandwidth PCIe peripherals first in the chain (closest to the host), followed by DisplayPort monitors, and low-speed USB peripherals last.

3.Verify Host Router Tunnel Allocation:

Open Device Manager > View > Devices by connection.

Expand the USB4 Host Router tree and confirm that connected hubs do not exceed 3 downstream hop levels.

What system hardware configuration or software status exists while PCIe tunneling is missing?

Hardware Virtualization (Intel VT-d or AMD-Vi IOMMU) is disabled in the motherboard UEFI/BIOS.
Multiple 4K/8K displays consume all available tunnel bandwidth, leaving no allocated bandwidth for PCIe data.
The peripheral uses an outdated Thunderbolt 3/4 controller firmware incompatible with USB4 v2.0 Host Routers.

Enable Hardware IOMMU / VT-d Virtualization in UEFI BIOS

Solution:

Root Cause: IOMMU / VT-d Hardware Virtualization Disabled at Motherboard Level

USB4 and Thunderbolt 5 enforce strict PCIe packet isolation through Input-Output Memory Management Units (IOMMU). If Intel VT-d or AMD-Vi is disabled in motherboard BIOS, the operating system kernel cannot create IOMMU translation domains. As a safety precaution, Windows suppresses PCIe tunneling entirely (disabling NVMe SSDs, eGPUs, and LAN controllers) while allowing DisplayPort Alt-Mode video tunnels to pass through.

Step-by-Step Fix:

1.Reboot into Motherboard UEFI BIOS:

Restart the PC and press Del or F2 repeatedly during POST.

2.Enable Virtualization & IOMMU Settings:

Intel Systems: Navigate to Advanced > CPU Configuration or System Agent (SA) Configuration > set Intel VT for Directed I/O (VT-d) to Enabled.

AMD Systems: Navigate to Advanced > AMD CBS / PBS > set AMD-Vi and IOMMU to Enabled.

3.Enable Thunderbolt / USB4 PCIe Tunneling:

In BIOS, locate PCIe Tunneling or Thunderbolt Technology and set it to Enabled.

4.Save and Boot into Windows:

Press F10 to save changes and restart. Verify that PCIe expansion devices now enumerate in Device Manager.

Rebalance DisplayPort 2.1 and PCIe Bandwidth Tunneling Allocation

Solution:

Root Cause: Dynamic Display Tunnel Bandwidth Preemption

Thunderbolt 5 host routers feature dynamic asymmetric bandwidth adjustment (boosting up to 120Gbps in one direction for displays). However, DisplayPort 2.1 uncompressed high-resolution streams (e.g., dual 8K @ 60Hz or triple 4K @ 144Hz) take priority over data tunnels. If display bandwidth allocation saturates available high-speed lanes, the USB4 connection manager starves or closes the PCIe data tunnel to prevent display flickering.

Step-by-Step Fix:

1.Enable Display Stream Compression (DSC):

Open your GPU Control Panel (NVIDIA Control Panel or AMD Software).

Ensure Display Stream Compression (DSC) is enabled on all connected high-resolution monitors to reduce raw video bandwidth demand by up to 66%.

2.Adjust Refresh Rates in OS Settings:

Open Settings > System > Display > Advanced display.

Temporarily reduce monitor refresh rates (e.g., from 144Hz down to 60Hz) to free up bandwidth lanes for PCIe data.

3.Use Windows USB4 Hubs App to Monitor Bandwidth:

Launch the USB4 Hubs and Devices app in Windows 11 to view real-time percentage allocations across DisplayPort, PCIe, and USB3 tunnels.

Update Thunderbolt Host & Peripheral NVM Controller Firmware

Solution:

Root Cause: Legacy Non-Volatile Memory (NVM) Firmware Handshake Incompatibility

Connecting legacy Thunderbolt 3 or early Thunderbolt 4 devices (e.g., Alpine Ridge or Titan Ridge controllers) to modern USB4 Version 2.0 / Thunderbolt 5 host routers requires updated controller microcode. Outdated NVM firmware on the peripheral fails to complete the low-level handshake for PCIe protocol tunneling over USB4 host routers.

Step-by-Step Fix:

1.Check NVM Firmware Version:

Open Thunderbolt Control Center or launch PowerShell as Administrator and run:

powershell
     Get-PnpDevice -Class 'System' | Where-Object {$_.FriendlyName -match 'Thunderbolt|USB4'}

2.Download Firmware Flasher from Peripheral Manufacturer:

Visit the support site for your dock or device and download the latest NVM Firmware Updater (e.g., NVM Version 36+ for TB4/TB5 compatibility).

3.Execute Firmware Flash Routine:

Disconnect all other USB devices, attach the dock directly to the PC, and execute the firmware flash utility. Power cycle the dock completely after flashing finishes.

What physical hardware condition or power status exists on the unpowered USB4 / Thunderbolt 5 setup?

The laptop or host PC Embedded Controller (EC) state is frozen, blocking USB Type-C CC pin detection.
The dock power supply works, but Power Delivery fails to negotiate wattage for laptop charging or peripheral power.
Using a desktop PCIe Add-In Card (AIC) where internal headers (TB_HEADER, GPIO, SATA power) are disconnected.

Perform Hard Embedded Controller (EC) Power Reset to Clear USB-C CC Pin Latch

Solution:

Root Cause: Embedded Controller (EC) Latch-Up on Configuration Channel (CC) Logic

Type-C physical ports utilize Configuration Channel (CC1/CC2) pins to detect device insertion, orientation, and initial power roles. A static discharge, ground loop, or abnormal voltage spike can cause the motherboard's Embedded Controller (EC) to enter a latched fault state. In this state, the EC disables VBUS power output and ignores state changes on the CC pins, making the USB4 port appear entirely dead.

Step-by-Step Fix:

1.Perform Hard Battery/Power Drain Reset (Laptops):

Unplug all cables, power supply bricks, and USB devices from the laptop.

Press and hold the laptop power button continuously for 30 to 45 seconds.

(For laptops with dedicated pinhole reset buttons, insert a paperclip into the EC reset pinhole on the bottom case for 10 seconds).

2.Perform CMOS and EC Reset (Desktops):

Turn off the PSU power switch and unplug the AC power cord.

Remove the motherboard CMOS battery (CR2032) for 5 minutes or bridge the CLR_CMOS jumper pins.

3.Reconnect Power and Test Port Initialization:

Reattach the AC power cable, power on the host system, and reconnect the USB4/Thunderbolt 5 device.

Calibrate USB Power Delivery 3.1 EPR (Extended Power Range) Negotiation

Solution:

Root Cause: Failed Power Delivery Protocol State Machine (PD Contract Mismatch)

USB4 and Thunderbolt 5 support bidirectionally negotiated Power Delivery (PD 3.1). Before powering up high-speed physical layers, the host and device perform a hardware handshake over CC lines to agree on a power contract (e.g., 5V/3A up to 20V/5A or 28V/5A EPR). If the connected peripheral attempts to draw current before the PD contract is finalized, the host controller triggers over-current protection (OCP) and cuts VBUS power completely.

Step-by-Step Fix:

1.Reset Power Delivery Controller State Machine:

Unplug both the Thunderbolt cable and the dock's external power brick from the wall socket.

Wait 60 seconds to allow internal power supply capacitors to discharge fully.

2.Enforce Correct Cable Plug Sequence:

Plug the dock's power adapter into the wall outlet *first*.

Ensure status LEDs on the dock light up in standby mode.

Connect the Thunderbolt 5 / USB4 cable to the host laptop *second* to initiate a fresh PD handshake.

3.Test Passive USB PD Fallback:

If using a multi-port dock, plug into the designated Host Port (marked with a laptop icon) rather than downstream peripheral ports.

Verify Desktop Motherboard TB_HEADER, GPIO, and PCIe Auxiliary Power Cables

Solution:

Root Cause: Missing Internal Header Connections or Auxiliary Power on PCIe Add-In Cards

Unlike laptops with integrated USB4 routers, desktop PC setups utilizing PCIe Thunderbolt 5 / USB4 Expansion Add-In Cards (AICs) (e.g., ASUS ThunderboltEX 5 or Gigabyte GC-MAPLE RIDGE) require physical sideband signaling cables connected directly to the motherboard. If the internal TB_HEADER (GPIO) cable or 6-pin/8-pin PCIe auxiliary power cable is missing, the card cannot receive power-state signals or negotiate high-wattage port output.

Step-by-Step Fix:

1.Verify Internal TB_HEADER / GPIO Jumper Cable:

Power down the PC and inspect the expansion card.

Ensure the supplied sideband cable is connected from the card's header to the motherboard's matching TB_HEADER / USB4_HEADER header pins.

2.Connect 6-pin/8-pin PCIe Auxiliary Power:

Thunderbolt 5 desktop cards require supplemental 12V power to support 100W+ Type-C Power Delivery. Connect a dedicated 6-pin or 8-pin PCIe power cable directly from the PSU to the card.

3.Verify Motherboard PCIe Slot Generation & Routing:

Ensure the card is seated in a motherboard PCIe slot running at PCIe 4.0 x4 or PCIe 5.0 x4 speed directly routed to the CPU or Chipset as specified by the motherboard manufacturer manual.

Related Diagnostic Guides

Ubuntu Black Screen After Nvidia Driver Install - Secure Boot Diagnostic

SSH Connection Refused on Port 22 But Service Running - Firewall Tree

Linux Boot Stuck at GRUB Rescue After Windows Update - Dual Boot Fix

SmartScreen Can't Be Reached Right Now Error Fix - Windows Diagnostic Guide

Account Locked After Too Many Attempts in Microsoft 365 - Admin Unlock Tree

Subscribe to StepDiag

Embed Guide on Your Website