Heartbreaker
Diamond Member
- Apr 3, 2006
- 4,803
- 6,317
- 136
Some interesting tidbits from the connector specs in the PCIe_CEM_R5.1_V1.0 documentation:
To me this shows increased reliance on the testing and the quality standards of each manufacturer. The 9.2A per pin load is actually the minimum requirement before testing. The final max rating can be anything based on connector side-load testing
"The specific wire, connectors and manufacturing process used for a cable assembly must be designed to accept the current imbalance due to contact resistance variability and side loading"
How the heck are they supposed to do that? It may be that they are expecting minor imbalances, so It can handle 10 or 11 amps, but we are seeing much worse in some cases.
It seems like they need some kind of Apple-like cable/connector approval process, where many samples are torture tested.
I remember the recent issues Der8auer had with his properly plugged in cable with no side loads. These were not minor current differences, he had 2, 4, and 5 amps on some wires and 23 amps on one of them.
To get those kind of imbalances, he had to have multiple very poor conductor/connection in the set. Which kind of points to a poor manufacturing process somewhere.
I was a little disappointed that he didn't investigate into the source of that cables issues (He plugged in a new cable and everything was fine). My curiosity would be why was the original cable so bad? It's rare to catch something this bad before it melts obscuring the original problem, so it was a unique investigation opportunity.
Last edited: