Our name is Hardware Busters because one of our many jobs is to bust hardware-related myths. Today’s article and video, are a fine example of that!
It hit me once I noticed an article at Tom’s Hardware with the title: User Reports 12VHPWR Connector Melting From the PSU Side (Updated)
This looks crazy for two reasons: from when a report from a single user hits the news on such a large site, and secondly, we usually hear about problems on the other end of the 12VHPWR cable, which goes to the graphics card, and not the one on the PSU side.
I prepared a quick set-up with four Array electronic loads, capable of pulling up to 900W from the PSU, and since I happen to have the same PSU as the user that reported the problem on Reddit, a be quiet! Dark Power 13 1000W, I thought to load its 12VHPWR fully while monitoring the temperatures on both connectors. I didn’t intentionally use a special fixture because I wanted to go over the limits. Moreover, I wanted to simulate a scenario where the other end of the connector doesn’t make an ideal contact, to see what will happen.
I started with 60A, around 720W, which was fine for the 12VHPWR connectors, but it applied massive stress on the mod I used to transfer the power to the loads, with the soldering starting to melt! Moreover, its temperature rise was notable because of the not ideal contact on the load side of the 12VHPWR connector. In contrast, the PSU side of the 12VHPWR connector had an operating temperature close to 30C!
I lowered the load to continue the experiment until the connection I made broke off because of the high temperature. Still, no problems at all on the 12VHPWR cable, with its PSU side being a little above 30C.
The conclusions of this short experiment are the following:
- A non-ideal contact of the 12VHPWR connector on the load side can notably increase its operating temperature, so it can lead to melting if you apply excess loads, especially at high ambient temperatures.
- A non-ideal contact of the 12VHPWR connector on the load side cannot affect the connector on the PSU side if this is properly connected!
According to my sources, the PCI-SIG plans to take measurements to ensure the proper connections of 12VHPWR connectors. Unfortunately, I cannot reveal more since I am not a leaker, but something is already in the works.
How large batch of cables are you going to test with your setup to ensure you get proper glimpse of standard deviations between actual production units?
There might be a lot of variance in production in the beginning when producing something brand new. Lots of new stuff to fail and learn from.
I am suspecting that failure rates for this connector type are way too high for something meant to be installed by consumers themselves. Especially when manufacturer error or user error might wipe out thousands worth of equipment.
I am not going to test anything more, because we will have new connectors at some point.
• What was the ambient temperature? 28°C?
• The graph shows that the temperature on the PSU side drops when a load is applied. Are you sure that your equipment measures properly?
• Are you planning to release an article that explores the actual load limitations of the 12VHPWR? ASUS stated that 1000 W are a non-issue months ago and backed this with an OC BIOS.