Category: Gaming

Key Takeaways

1. Connector Melting Issue: RTX 4090 users are experiencing melting 12VHPWR connectors, leading to burnt plastic odors and fusing to graphics cards and power supply units (PSUs).

2. Thermal Imbalance: The lack of current monitoring across the 12V lines results in uneven current distribution, causing thermal hotspots and connector degradation.

3. Resistance Variability: Significant inconsistencies in resistance values among the power pins can lead to localized heating, generating enough heat to compromise connector integrity.

4. Vendor Accountability: Inno3D has denied responsibility for connector issues under warranty, while other vendors like ModDIY have shown willingness to assist, highlighting inconsistent vendor responses.

5. Design Criticism: Der8auer criticizes Nvidia’s power connector design and calls for better protection and support for consumers, emphasizing the need for a redesign to prevent future failures.

The ongoing issue of melting 12VHPWR connectors has come back into the spotlight, this time impacting a user of the RTX 4090 who reached out to overclocker and YouTuber Der8auer. While enjoying some gaming, the user detected a familiar burnt plastic odor and found that the high-power connector had fused to both the graphics card and PSU ends. The graphics card in question was an Inno3D model that had since been modified with water cooling, operating with an undervolt and connected via a ModDIY cable that used 16 AWG ultra-soft wires. Even with careful assembly and a PSU (SilverStone Extreme 850R) considered suitable for a 4090, the plug became stuck, indicating a thermal issue within the connector. Inno3D’s customer service, however, declined to provide assistance, claiming that the damage was on the cable and PSU sides—an argument that Der8auer finds technically flawed.

Current Monitoring Issues

Der8auer elaborates that the main problem focuses on the absence of current monitoring across the individual 12V lines in the 12VHPWR interface. It seems that RTX 40 and 50 series graphics cards are unable to identify uneven current distribution among their six power pins. This imbalance, often caused by varying contact resistances in the wires or pins, can create thermal hotspots. In this particular instance, Der8auer managed to take apart and replace the damaged connector on the GPU side himself, using an H++ version of the 12VHPWR plug. His measurements showed significant inconsistencies in resistance values—one pin was as low as 12 milliohms, while another reached 43 milliohms, and one had no connection whatsoever. These differences directly lead to uneven current flow, which results in localized heating and ultimately causes connector melting and degradation. Even with a modest power draw, this can generate up to 4.6 watts of heat concentrated in a small area—sufficient to compromise the connector’s integrity over time.

Vendor Responses

Adding to the frustration is the inconsistent response from vendors. ModDIY displayed some readiness to assist, including an offer to pay for repair costs, but Inno3D consistently denied any responsibility, even though the card was still under warranty. Der8auer emphasizes that only the GPU-side design can effectively reduce these risks, as neither the cable nor the PSU can monitor load distribution. He notes that while high resistance can effectively “cut off” a wire, it causes more current to flow through the remaining wires, which leads to a cycle of overheating. Ultimately, Der8auer criticizes Nvidia’s poor power connector design and the lack of accountability from its partners. With high-end graphics cards priced over 2,000 euros, consumers deserve better protection—and quicker support—when failures happen. Until Nvidia revisits this design, users face ongoing risks, regardless of how meticulously they assemble or undervolt their systems.

Source:
Link