der8auer finds major problem with RTX 5090 power connector
- Thread starter ZZLEE
- Start date
gingerbee
Well-known member
Yup, I posted it in fav/YouTube video, its a problem.
Bond007
Well-known member
Actually hardcore overclock has a bit of a follow on/add on video that shows how the FE cards have all wires going to just two connections on the card (one positive and one ground), so the card has no way of controlling or knowing this issue…the power is just going to take whatever the easiest path is through the cables (no balancing/detection for how much load each cable has).
Last edited:
gingerbee
Well-known member
12VHPWR, AHO "Nvidia got tired of trying to current balance the connector so now they just melt" wtf. Even the old 6-pin has a better design, this is crazy.
Last edited:
FreeKnight
Well-known member
Just as the Cheetoh dismantles the Consumer Financial protection bureau, so for Americans if your 5090 starts a fire, your only hope is a class action or private law suit (though admittedly the CFPB was often hobbled by Congress' refusal to allow anything resembling the dreaded 'regulation' in more recent years).
Real shame with the bonkers margins and profits Nvidia is making simple things like 'make products that don't combust' feels like a step too far for them now.
Real shame with the bonkers margins and profits Nvidia is making simple things like 'make products that don't combust' feels like a step too far for them now.
I was thinking about this some more beyond what was posted in favorites thread.
In a lab situation the 6 wires in parallel should be perfectly balanced every time regardless of number of shunt resistors.
Multiple wires in parallel are simply just seen as 1 larger sized conductor. In debaurers video he is even getting reading on the other wires meaning there isn't a break in the wire like being explained in the second video.
For the load to shift so dramatically in this case on paper the other wires would need to have a higher resistance to force this behavior but it isn't like there is in line resistors in the wires. Even small manufacturing differences in this case shouldn't cause that much of an imbalance.
It feels like right now a lot of big pieces have been exposed but feels like there is a key detail still missing here. Especially since it seems like it was the same conductor for both people.
In a lab situation the 6 wires in parallel should be perfectly balanced every time regardless of number of shunt resistors.
Multiple wires in parallel are simply just seen as 1 larger sized conductor. In debaurers video he is even getting reading on the other wires meaning there isn't a break in the wire like being explained in the second video.
For the load to shift so dramatically in this case on paper the other wires would need to have a higher resistance to force this behavior but it isn't like there is in line resistors in the wires. Even small manufacturing differences in this case shouldn't cause that much of an imbalance.
It feels like right now a lot of big pieces have been exposed but feels like there is a key detail still missing here. Especially since it seems like it was the same conductor for both people.
moocow
Well-known member
Falcon Northwest did their own testing and couldn't replicate der8auer's result.
videocardz.com
RTX Cablegate update: GeForce RTX 5090 claims another 3rd party cable, MODDIY now recommends to upgrade to newer 12V-2x6 cables - VideoCardz.com
Not just RTX 5090 but also 4090 started to melt again We gathered some news related to overheating and melting accidents. Should we rename to Cable News Network? Perhaps that’s not a bad idea, given that we have been talking about RTX 50 cables for a few days now. And indeed there was a lot […]
videocardz.com
Bond007
Well-known member
Very true. Maybe resistance at the connection between the pins and the metal bar they are all attached to? No idea, but there must be something.
gingerbee
Well-known member
Ya I was wondering the same thing, but I don't know enough about shunt resistors, but it seems clear either I don't understand why it does this or it is not explained
I wonder if it's a certain board spec, but I think they would all be the same or maybe a certain PSU/cable type.
I wonder if it's a certain board spec, but I think they would all be the same or maybe a certain PSU/cable type.
freeagent
Well-known member
Just curious.. out all of these failures..
How many were using Seasonic PSUs?
Just saying, the cable and plug look pretty cheap to my 4 eyes on this big screen.
How many were using Seasonic PSUs?
Just saying, the cable and plug look pretty cheap to my 4 eyes on this big screen.
