• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

12VHPWR &12v 2x6 Fire Risk Discussion

d3athf1sh

[H]ard|Gawd
Joined
Dec 16, 2015
Messages
1,454
so i recently had to replace an aging power supply in my parents computer which uses a graphics card with 2 x 6pin pci-e. so no problem. i end up going with an MSI MPG A850GS which is ATX 3.1 and both 80 plus Gold and Cybenetics Gold. reviews and specs here: Hardware Busters & Cybenetics

but then as i'm looking at it while installing i notice what you can clearly see in the picture. now i'm not sure if the new ATX 3.1 spec has anything to do with the connectors and they are interchangeble between brands but if this IS the standard, i can clearly see why so many people have had problems. you can clearly see that the posts for the 12v 2x6 (revised HiPwr) are clearly HALF THE SIZE of the standard ones we've been using for years! so a little background, i'm an electrician of 30yrs, why in the Funk would they use smaller posts and push more power through them? it's no wonder there are problems, those posts look to be around 22 awg(?) wire and they're steel, not copper, but the bigger problem is that little post has probably HALF the surface area of the bigger ones, now being steel doesn't add a whole lot of resistance, but the small diameter and contact area SURE DOES! especially when pushing twice the current! you get what's called thermal runaway. because the more resistance there is, the more heat is produced, which more heat in turn, adds more resistance which adds more heat until FINALLY you get melt down. man whoever designed this spec should be dragged behind the shed and shot. and that's the new standard? what happened to UL certification and whatever the Canadian equivalent is? did ANYONE test it or is everyone buying into the newest ATX standard a working beta tester?

so this is my call to anyone with a brain. buy cards with standard pci-e connectors. save yourself a future head ache and do the world a favor by making 12vHiPwr go the way of the dodo.

mpg850gs.jpg
 
Last edited:
yeah i was thinking about it and another thing that's probably adding to the number of things causing meltdown, is the same reason we stopped wiring houses with aluminum wire, when it heats up and cools down (when running high loads and getting hot) it expands and contracts and eventually the connectors won't be making good contact, which like i said before, will add more resistance which adds more heat.... etc etc. plus that happens gradually over time so that's probably the reason you see people say "it was running fine for a year, then all of a sudden..." it makes perfect sense now.

still blows my mind that this "standard" ever made it out the underwriters lab. they're job is to torture test stuff.

who knows, maybe jensen lined some people's pockets to get the standard rushed through so they could save money on their "tiny pcb" design that's not big enough for standard 8 pin connectors. which still doesn't make much sense, to me, being the card still takes 3.5 spaces. how much are they really saving by shrinking the pcb? it can't be that much?

edit: interesting, just looked at the box and there's no UL stamp anywhere on the box. i thought that was a requirement for all electrical appliances sold in the US? there is a C and a CE right before like 5 or 6 "recycle" logos, so not sure if those are the equivalent canadian testing labs? doesn't say anything below it.
 
Last edited:
I'm pretty sure most of the savings is coming from the PCB design- the single bus instead of separating each pin into its own trace.
 
m pretty sure most of the savings is coming from the PCB design- the single bus instead of separating each pin into its own trace.
Seems so counter productive though. There are already designs popping up that are built around monitoring each pins power delivery. So insanely dumb...Full circle solving a problem that shouldn't of existed in the first place.
 
Seems so counter productive though. There are already designs popping up that are built around monitoring each pins power delivery. So insanely dumb...Full circle solving a problem that shouldn't of existed in the first place.
Yes, but with a difference. nVidia isn't paying for it, us consumers are. And the kicker is that a 3rd party solution costs more than if nVidia had simply done it right the first time and passed the expense on.
 
so i recently had to replace an aging power supply in my parents computer which uses a graphics card with 2 x 6pin pci-e. so no problem. i end up going with an MSI MPG A850GS which is ATX 3.1 and both 80 plus Gold and Cybenetics Gold. reviews and specs here: Hardware Busters & Cybenetics

but then as i'm looking at it while installing i notice what you can clearly see in the picture. now i'm not sure if the new ATX 3.1 spec has anything to do with the connectors and they are interchangeble between brands but if this IS the standard, i can clearly see why so many people have had problems. you can clearly see that the posts for the 12v 2x6 (revised HiPwr) are clearly HALF THE SIZE of the standard ones we've been using for years! so a little background, i'm an electrician of 30yrs, why in the Funk would they use smaller posts and push more power through them? it's no wonder there are problems, those posts look to be around 22 awg(?) wire and they're steel, not copper, but the bigger problem is that little post has probably HALF the surface area of the bigger ones, now being steel doesn't add a whole lot of resistance, but the small diameter and contact area SURE DOES! especially when pushing twice the current! you get what's called thermal runaway. because the more resistance there is, the more heat is produced, which more heat in turn, adds more resistance which adds more heat until FINALLY you get melt down. man whoever designed this spec should be dragged behind the shed and shot. and that's the new standard? what happened to UL certification and whatever the Canadian equivalent is? did ANYONE test it or is everyone buying into the newest ATX standard a working beta tester?

so this is my call to anyone with a brain. buy cards with standard pci-e connectors. save yourself a future head ache and do the world a favor by making 12vHiPwr go the way of the dodo.

View attachment 809806
Same been a sparky my whole life and I was pissed when I saw how chicken shit the power delivery for 600w was on a 3k card
 
was just watching the latest derbauer video and the ad for Seasonic Prime 2200w power supply was playing. took this screen grab, but you can clearly see that the pins for the 12v2x6 HiPwr are clearly half the size as the normal pins. so i guess that's the standard on all psu's (not just MSI) and we will continue to have the meltdown problem for as long as we continue to use it. especially on hi power draw cards like the 600w ones from nvidia.

seasonic_2200.jpg


someone should really should start a class action lawsuit and get these things pulled from the market. but i guess like every other electrical code ruling, it's gonna take someone's house burning down or loss of life before anyone gets off their asses and does anything.
 
Last edited:
someone should really should start a class action lawsuit and get these things pulled from the market. but i guess like every other electrical code ruling, it's gonna take someone's house burning down or loss of life before anyone gets off their asses and does anything.
I mean, go talk to a lawyer if you feel strongly... but don't be surprised if they aren't that interested. Screamy posts on the Internet are not proof. You need to be able to show that these are melting at a rate that is causing an increased number of RMAs compared to previous connectors. Can you demonstrate that? Do you have any evidence to show it is the case?

Likewise with the safety angle, do you have any evidence they are a hazard? The fact that they melt doesn't mean they are a fire hazard. Part of testing from places like UL and TUV frequently are things like "What happens if this melts?" or "If the components catch fire, will they put themselves out before the unit burns?" What is the safety testing that the companies have done?

Vibes don't win lawsuits, data does. Do you have the data that shows that indeed these new connectors are a big problem, or do you have some Reddit posts? That's not a snarky rhetorical question, if you have data on this I'd be really interested. However if you think it is an issue that reaches lawsuit level, you'll need it.
 
My PSU started to smell so I ordered the Wireview, I'm not taking any chances any more.
 
I mean, go talk to a lawyer if you feel strongly... but don't be surprised if they aren't that interested. Screamy posts on the Internet are not proof. You need to be able to show that these are melting at a rate that is causing an increased number of RMAs compared to previous connectors. Can you demonstrate that? Do you have any evidence to show it is the case?

Likewise with the safety angle, do you have any evidence they are a hazard? The fact that they melt doesn't mean they are a fire hazard. Part of testing from places like UL and TUV frequently are things like "What happens if this melts?" or "If the components catch fire, will they put themselves out before the unit burns?" What is the safety testing that the companies have done?

Vibes don't win lawsuits, data does. Do you have the data that shows that indeed these new connectors are a big problem, or do you have some Reddit posts? That's not a snarky rhetorical question, if you have data on this I'd be really interested. However if you think it is an issue that reaches lawsuit level, you'll need it.
the one i bought a few weeks back didn't have UL certification or the Canadian equivalent

The fact that they melt doesn't mean they are a fire hazard.
LOL :smh:
 
My PSU started to smell so I ordered the Wireview, I'm not taking any chances any more.
yeah that should help, sucks though, you shouldn't need that. we've had psu's for 40 years that didn't have these problems but then again, they weren't trying to push 600w through a post half the size of the old standard. and think about it, the old post size has been used on extreme overclocking competitions since their inception and they never had this problem. i just can't believe there are still consumers out here defending it. got to justify their purchase, i guess?
 
I noticed that the Wireview modules are at Microcenter. I think I might get one for my 5090. Easier than attempting to roll my own solution. Seems like the way to go. Price is overall fair.
 
the one i bought a few weeks back didn't have UL certification or the Canadian equivalent


LOL :smh:
No matter how you feel, he's not wrong. At this point, the only thing that will force nVidia to act rapidly is if it actually sets fire to a house. Maybes and possibilities mean nothing in the legal world without proof. If you want to spearhead a legal challenge that has any chance of winning, you will have to demonstrate that the melted connector will set fire to things outside of the case.
 
No matter how you feel, he's not wrong. At this point, the only thing that will force nVidia to act rapidly is if it actually sets fire to a house. Maybes and possibilities mean nothing in the legal world without proof. If you want to spearhead a legal challenge that has any chance of winning, you will have to demonstrate that the melted connector will set fire to things outside of the case.
Not necessarily. Technically, the fumes from the melting could be hazardous as well, and I think anything burning unintentionally is a fire hazard. I doubt it's as strict as that.

It's just an issue of someone with enough bank to take it to court.
 
yeah that should help, sucks though, you shouldn't need that. we've had psu's for 40 years that didn't have these problems but then again, they weren't trying to push 600w through a post half the size of the old standard. and think about it, the old post size has been used on extreme overclocking competitions since their inception and they never had this problem. i just can't believe there are still consumers out here defending it. got to justify their purchase, i guess?
I couldn't figure out what was causing the smell, only that it was 99% likely from the PSU. I pulled and inspected every cable and tested the entire 12V rail for EPS/PCIe, including ground and found zero issues. The PC is also running, but moving the EPS connectors results in instability.

Credit to Seasonic who are sending me a replacement.

Just got my WireView Pro II wired today, and this thing oozes quality. I can't seem to find any manual for the mounting bracket though, I guess I'll just poke around and see what works.

20260723_140006.jpg
 
Just got my WireView Pro II wired today, and this thing oozes quality. I can't seem to find any manual for the mounting bracket though, I guess I'll just poke around and see what works.

View attachment 816518
nice (y)

i've got the reversed non-wired version in the build i'm currently working on (see below); i had thought about ordering the wired version as well, but i'm going to wait until after the build is up and running to decide if i really want the wired version.

heavy-metal-build-071626-update.png

fire-alarm-instruction-red-2.png
 
Just installed it. I'm about to run a short test to see how my current plug is operating. I'm a bit scared tho given the previous smelly.

TBH I didn't really look at what this thing offers before buying. Holy shiz, this thing includes a jumper for your power and the wireview can hard-shutdown your PC. I also installed the latest HWMonitor and sure enough, there it is.
Amazing.
 
So I pretty much immediately got a current imbalance warning. 3 of my pins are pulling half an amp more than the rest. It's still well below the 10A limit though, I'll have a mess around with it when the new PSU turns up.
 
nice (y)

i've got the reversed non-wired version in the build i'm currently working on (see below); i had thought about ordering the wired version as well, but i'm going to wait until after the build is up and running to decide if i really want the wired version.

View attachment 816519

View attachment 816520
wow! that's quite the machine you got there. actually no, let me say it right, that's f*ckin awesome dude. (y)
 
Sigh. While I know it was just anxiety talking, I decided I too wanted per-pin current monitoring to make me feel better, even though I've had no issues. I'm not interested in the Wireview for a number of reasons but MSI has a power supply which will do the trick. So I got one and....

1784849703206.png


Absolutely no issues. Everything is nicely in balance and in range. Looking at a log during a Steel Nomad run (the only thing that draws more than about 470w with my undervolt) there was about 2% spread between pins which is not only within spec, it is around the noise floor of the shunt resistors so realistically the lowest measurable.
 
Back
Top