Hello,
I would like to ask for assistance regarding the warranty and replacement process for a Raspberry Pi AI HAT+ 2.
The product had been in my possession for only about one month when it failed, and it had only been used once for an actual deployment.
Failure timeline:
- I received the Raspberry Pi AI HAT+ 2 approximately one month before the failure.
- It was used for only one actual deployment session.
- Before leaving work, I completed the model and application deployment on a Raspberry Pi 5.
- At that time, the AI HAT+ 2 was detected normally, and the deployment was completed successfully.
- When I returned to work the following morning, the Hailo device could no longer be detected.
- There was no additional deployment, hardware modification, repair, or further operation between the successful deployment and the discovery of the failure.
The product therefore failed after only minimal use and within a very short period after receipt.
Hardware information:
- Host: Raspberry Pi 5
- Product: Raspberry Pi AI HAT+ 2
- Accelerator: Hailo-10H
- PCI ID: 1e60:45c4
- PCI address: 0001:01:00.0
The PCIe device remains visible through lspci, but:
- /dev/hailo0 is not created
- hailortcli scan reports no available device
- the hailo1x_pci driver fails during Stage 2 firmware initialization
- the same failure remains after reboots and full cold power cycles
The relevant kernel log includes:
customer_certificate.bin written successfully
scu_fw.bin written successfully
Board SKU-ID is: 6
u-boot-6.dtb.signed written successfully
Timeout waiting for firmware file
Failed writing SOC firmware on stage 2
SCU log could not be read from device
Firmware load failed
Failed activating board -110
probe with driver hailo1x failed with error -110
Software environment:
- Raspberry Pi OS 64-bit
- Kernel: 6.18.34+rpt-rpi-2712
- Official Hailo H10 packages
- h10-hailort 5.1.1
- h10-hailort-pcie-driver 5.1.1
Troubleshooting already completed:
- Clean Raspberry Pi OS installation
- Official repositories and packages
- Multiple reboots
- Full shutdown and cold power cycles
- USB-C power disconnected for more than 60 seconds
- PCIe FFC cable reseated
- GPIO header reseated
- Booted without USB storage and unnecessary peripherals
- Checked for undervoltage and throttling
- Confirmed that the PCIe device remains visible
- Tested more than one HailoRT and driver version
The device was returned for inspection, and I was informed that the onboard DDR was damaged.
Despite this hardware diagnosis, the reseller refused my request for replacement.
The important points are:
- The product was received only about one month before the failure.
- It was used for only one actual deployment.
- It was functioning normally when I completed the deployment before leaving work.
- It could no longer be detected when I returned the following morning.
- The subsequent inspection identified damaged onboard DDR.
Given the extremely short period of ownership and minimal use, this appears to be an early-life hardware failure rather than damage caused by normal wear, long-term operation, or repeated use.
Since the AI HAT+ 2 is a Raspberry Pi-branded finished product, could Raspberry Pi staff or forum moderators please clarify:
1. Is confirmed onboard DDR damage considered a hardware defect that requires replacement of the complete board?
2. Is there any software recovery for this Stage 2 firmware error, or is hardware replacement required?
3. What is the correct warranty or RMA escalation route when a reseller refuses replacement despite confirming DDR damage?
4. Can Raspberry Pi provide a technical confirmation, support case number, or written statement that I can submit to the reseller?
5. Is there an authorized distributor, regional support contact, or warranty escalation channel that can review this case?
6. Should I contact Raspberry Pi directly when the reseller refuses to provide replacement or an appropriate RMA route?
For privacy reasons, I have not included the purchase record, product serial number, reseller details, inspection documents, or correspondence in this public post.
I can provide the complete boot report, dmesg output, invoice, order information, product photographs, serial-number information, inspection record, and the reseller's written response privately to authorized Raspberry Pi or Hailo staff upon request.
I would appreciate guidance from Raspberry Pi staff or moderators regarding how this warranty case should be handled.
Thank you.
I would like to ask for assistance regarding the warranty and replacement process for a Raspberry Pi AI HAT+ 2.
The product had been in my possession for only about one month when it failed, and it had only been used once for an actual deployment.
Failure timeline:
- I received the Raspberry Pi AI HAT+ 2 approximately one month before the failure.
- It was used for only one actual deployment session.
- Before leaving work, I completed the model and application deployment on a Raspberry Pi 5.
- At that time, the AI HAT+ 2 was detected normally, and the deployment was completed successfully.
- When I returned to work the following morning, the Hailo device could no longer be detected.
- There was no additional deployment, hardware modification, repair, or further operation between the successful deployment and the discovery of the failure.
The product therefore failed after only minimal use and within a very short period after receipt.
Hardware information:
- Host: Raspberry Pi 5
- Product: Raspberry Pi AI HAT+ 2
- Accelerator: Hailo-10H
- PCI ID: 1e60:45c4
- PCI address: 0001:01:00.0
The PCIe device remains visible through lspci, but:
- /dev/hailo0 is not created
- hailortcli scan reports no available device
- the hailo1x_pci driver fails during Stage 2 firmware initialization
- the same failure remains after reboots and full cold power cycles
The relevant kernel log includes:
customer_certificate.bin written successfully
scu_fw.bin written successfully
Board SKU-ID is: 6
u-boot-6.dtb.signed written successfully
Timeout waiting for firmware file
Failed writing SOC firmware on stage 2
SCU log could not be read from device
Firmware load failed
Failed activating board -110
probe with driver hailo1x failed with error -110
Software environment:
- Raspberry Pi OS 64-bit
- Kernel: 6.18.34+rpt-rpi-2712
- Official Hailo H10 packages
- h10-hailort 5.1.1
- h10-hailort-pcie-driver 5.1.1
Troubleshooting already completed:
- Clean Raspberry Pi OS installation
- Official repositories and packages
- Multiple reboots
- Full shutdown and cold power cycles
- USB-C power disconnected for more than 60 seconds
- PCIe FFC cable reseated
- GPIO header reseated
- Booted without USB storage and unnecessary peripherals
- Checked for undervoltage and throttling
- Confirmed that the PCIe device remains visible
- Tested more than one HailoRT and driver version
The device was returned for inspection, and I was informed that the onboard DDR was damaged.
Despite this hardware diagnosis, the reseller refused my request for replacement.
The important points are:
- The product was received only about one month before the failure.
- It was used for only one actual deployment.
- It was functioning normally when I completed the deployment before leaving work.
- It could no longer be detected when I returned the following morning.
- The subsequent inspection identified damaged onboard DDR.
Given the extremely short period of ownership and minimal use, this appears to be an early-life hardware failure rather than damage caused by normal wear, long-term operation, or repeated use.
Since the AI HAT+ 2 is a Raspberry Pi-branded finished product, could Raspberry Pi staff or forum moderators please clarify:
1. Is confirmed onboard DDR damage considered a hardware defect that requires replacement of the complete board?
2. Is there any software recovery for this Stage 2 firmware error, or is hardware replacement required?
3. What is the correct warranty or RMA escalation route when a reseller refuses replacement despite confirming DDR damage?
4. Can Raspberry Pi provide a technical confirmation, support case number, or written statement that I can submit to the reseller?
5. Is there an authorized distributor, regional support contact, or warranty escalation channel that can review this case?
6. Should I contact Raspberry Pi directly when the reseller refuses to provide replacement or an appropriate RMA route?
For privacy reasons, I have not included the purchase record, product serial number, reseller details, inspection documents, or correspondence in this public post.
I can provide the complete boot report, dmesg output, invoice, order information, product photographs, serial-number information, inspection record, and the reseller's written response privately to authorized Raspberry Pi or Hailo staff upon request.
I would appreciate guidance from Raspberry Pi staff or moderators regarding how this warranty case should be handled.
Thank you.
Statistics: Posted by songhaoyan3 — Wed Jul 29, 2026 8:29 am — Replies 0 — Views 10
![[QLD, NSW, VIC] Samsung 65](http://files.ozbargain.com.au/n/31/978831l.jpg?h=8d64f0c6)
