H1 susceptible to catch fire - contact NZXT

[LukeSavenije]

6 hours ago, marcolopes said:

Are there any H1 cases on the market without the SAFETY issue? https://blog.nzxt.com/details-on-h1-case-safety-issue/

If so, HOW CAN I CHECK if the cases are NOT affected by this problem?

Thanks

P.S.: I contacted nzxt but no response whatsoever!! (the community is non existent! https://support.nzxt.com/hc/en-us/community/topics

they can supply you with a replacement screw kit (plastic ones), but then they should come back to you first... might want to try your retailer as well

[Meronar]

Just got my repair kit

[shaz2sxy]

Got my email yesterday waiting for it to be shipped.

[Moonzy]

Gamer nexus did a video on this

I have the same feeling about this as Steve, I still won't feel comfortable using it even with nylon screw.

[Quackers101]

dammit was going to post GN, steve is on fire?!

Grounding and lucky or unlucky screw draw before you meet the fire demons.

Sparks left and right.

[Quackers101]

also in short what steve said in the video. Screw is not the problem the extension is, and it should be changed. the kit doesn't take away the risk, it makes it less of a problem perhaps, but not remove the dangerous part.

DO NOT USE THE RISER CABLE FROM THIS CASE, UNTIL FIXED WITH A NEW VERSION.

Fire hazards away!

[514_AppleGuy]

How in the Holy Name of F@ck does a case catch fire?

[Moonzy]

13 hours ago, 514_AppleGuy said:

How in the Holy Name of F@ck does a case catch fire?

Badly designed pcie riser, the PCB have exposed 12v line in the screw hole, where screws are normally grounded, creating a short, allegedly, all speculations and not confirmed, but looks right

 

Doesn't trip OCP SCP because the case provides enough resistance

 

13 hours ago, Quackers101 said:

also in short what steve said in the video. Screw is not the problem the extension is, and it should be changed. the kit doesn't take away the risk, it makes it less of a problem perhaps, but not remove the dangerous part.

DO NOT USE THE RISER CABLE FROM THIS CASE, UNTIL FIXED WITH A NEW VERSION.

Fire hazards away!

I would say replacing it with nylon screws is a cheap fix

if there's nothing to ground the exposed wire, then there's almost no risk

 

But yes I would sleep better at night if they changed the riser and not the screw, as GN mentioned, people might use metal screws or something in the future when they re-use the part and cause a fire.

 

 

edit: i have an issue with their demo of how it happens

the screw thread doesnt really move up and down relative to the piece, it just rotates as it screws in, so if it doesnt touch the trace, it shouldnt touch it

unless the pcb becomes loose and shifted up and down slightly

 

regardless, moving the screw does seem to start the issue so there's definitely something wrong there

[SFFDesigns]

Honestly I feel like there is no reason to buy an H1 in almost any situation right now. 

 

Meshlicious – ssupd

This case (which is also vertical) is better in almost every way, and will be released very soon.

 

The hardware compatibility is much better, the price is better (although you do have to buy a cooler and PSU separately) and the thermals are better 

 

If you do have an H1 though, I'd probably go for the PCIe 4.0 riser made by Louqe. It's pricy, but looks quite good and is PCIe 4.0

Edited January 24, 2021 by Hymenopus_Coronatus

[Bombastinator]

[Spotty]

8 hours ago, Moonzy said:

Doesn't trip OCP because the case provides enough resistance

I think you meant Short Circuit Protection (SCP), not Over Current Protection (OCP). The PSU doesn't sense the short because there's enough resistance provided by the case/motherboard tray.

The NZXT S650 is exclusive to the H1 case so there's not a lot of information out there on it, but it's manufactured by Seasonic for NZXT and is very likely just a rebrand of the Seasonic Focus SGX 650W. It even has the same fan grill. According to this Tom's Hardware review OCP wouldn't trip until around 65A, which as you saw from the consequences in the video is probably not helpful in this scenario Would have been interesting to have measured the current before it caught fire.

[Moonzy]

2 minutes ago, Spotty said:

I think you meant Short Circuit Protection (SCP), not Over Current Protection (OCP). The PSU doesn't sense the short because there's enough resistance provided by the case/motherboard tray

man im just so not familiar with PSU stuffs

 

can you link some page that explains both in detail? because i cant imagine how they would be different

[Bombastinator]

2 minutes ago, Spotty said:

I think you meant Short Circuit Protection (SCP), not Over Current Protection (OCP). The PSU doesn't sense the short because there's enough resistance provided by the case/motherboard tray.

The NZXT S650 is exclusive to the H1 case so there's not a lot of information out there on it, but it's manufactured by Seasonic for NZXT and is very likely just a rebrand of the Seasonic Focus SGX 650W. It even has the same fan grill. According to this Tom's Hardware review OCP wouldn't trip until around 65A, which as you saw from the consequences in the video is probably not helpful in this scenario Would have been interesting to have measured the current before it caught fire.

So not Oatmeal Pie.  I am saddened.

[Spotty]

Just now, Moonzy said:

man im just so not familiar with PSU stuffs

 

can you link some page that explains both in detail? because i cant imagine how they would be different

https://www.tomshardware.com/reviews/power-supplies-101,4193-21.html

 

Quote

SCP (Short Circuit Protection)

Short circuit protection (SCP) constantly monitors the output rails, and if it finds an impedance of less than 0.1Ω, it immediately shuts down the power supply. In other words, if somehow the output rails are short circuited, then this protection kicks in and shuts down the PSU to prevent damage or fire.

 

There's explanations for the rest of the protections found in PSUs included in that article on Tom's Hardware if you want to read up on the rest.

[LukeSavenije]

3 hours ago, Spotty said:

The NZXT S650 is exclusive to the H1 case so there's not a lot of information out there on it, but it's manufactured by Seasonic for NZXT and is very likely just a rebrand of the Seasonic Focus SGX 650W.

I'll even confirm this for you

[Kisai]

23 hours ago, Moonzy said:

Badly designed pcie riser, the PCB have exposed 12v line in the screw hole, where screws are normally grounded, creating a short, allegedly, all speculations and not confirmed, but looks right

 

That was nearly the exact speculation I had on it.

Quote

My guess is (see where those four bands are in the middle of the picture) the screws there made contact with the PCB.

However my thinking was that the screws were the wrong size. And if the 12V trace on the PCB is that close to the screw hole, that actually does sound like the screw holes were not drilled in the place expected (eg off by 1mm), and does kinda collaborate the reports, because only someone who didn't screw straight down (eg slightly angled) would have worn into the 12v.

 

23 hours ago, Moonzy said:

edit: i have an issue with their demo of how it happens

the screw thread doesnt really move up and down relative to the piece, it just rotates as it screws in, so if it doesnt touch the trace, it shouldnt touch it

unless the pcb becomes loose and shifted up and down slightly

 

regardless, moving the screw does seem to start the issue so there's definitely something wrong there

I think you're mis-interpreting the image, it's a 2D image, thus if you rotate the screw, the thread action is in fact moving up/down. So if it was screwed in multiple times, maybe not straight, it might have ground the edge of the pcb.

 

Regardless, I'd not use that PCIe riser, if any reason because it's not a question of "if" it will catch fire, but "when". Plastic screws doesn't solve the problem 3 years down the road when you upgrade the GPU or break the head off the screw from overtightening it.

 

[Moonzy]

33 minutes ago, Kisai said:

if you rotate the screw, the thread action is in fact moving up/down

If you rotate, the thread does indeed move up and down

Only if your screw isn't moving up and down along with it

 

If you're just rotating it, then yes, but if it's being screwed in, then the thread doesn't move up and down relative to the screw hole.

[Chiyawa]

Wow. Can't believe there is a fatal flaw in NZXT case. Well, this is quite unexpected.

 

Anyway, kudos for NZXT trying to rectify the problem, but I do think they can do it better. At least they tried.

[leadeater]

1 hour ago, Moonzy said:

If you rotate, the thread does indeed move up and down

Only if your screw isn't moving up and down along with it

 

If you're just rotating it, then yes, but if it's being screwed in, then the thread doesn't move up and down relative to the screw hole.

Have a watch again at this time stamp, it'll make it more clear what is going on.

 

 

 

It's not really the thread moving up and down that is the issue, although that is technically accurate, it's that the contact point between the thread and the PCB is rotating as you turn the screw. If the contact point between the screw thread and the PCB is 180 degrees away from the exposed trace then no problem however if you rotate the screw 180 degrees then you will be contacting the PCB trace. This won't change if the screw itself is moving up or down or not. So basically for many when the screw has been tight the screw thread has been some degrees rotation away from the trace and thus all is good, for a very small few that was not the situation.

 

Whether or not the screw hole itself is misaligned and is causing the issue I rather doubt it, the way those are manufactured I'd expect 99.9% of them to have the exposed trace in the screw hole. Simply bad design or improper parts selection. For how far away that screw is from the PCIe slot there is simply no reason a trace should need to be that close to the hole, what I personally suspect is this is a reused part or board mounted differently originally and that screw hole was never there when first designed and the PCB layout was created, most design tools would warn or prevent this issue, or at the very least be super obvious that it would happen.

[shaz2sxy]

I have had my NZXT H1 case since April 2020

 

I have just installed my "repair" kit. I noticed when taking out the metal screw from the rise you can hear the PCB being screwed in, so for me it look like it either had a screw that was too large, and/or the design of the PCB on the rise was not very good and the traces were too close to the holes.

 

The Riser has def got a design fault with it BUT the repair kit does fix the issues. Not the ideal way (i.e. replacing the PCIE rise) but as long as there is no metal screws touching the traces it is fine.

 

some notes on the video by Gamers Nexus.  for them to get the riser to catch on fire they had to mess about with it a lot. No one would do that to a case. 

 

The amount of cases that have been effected are around 10 

 

I do think it was a bit sensationalist. but it is good that they found out the root of the problem and will like to see how NZXT reply.

[linuxChips2600]

On 1/23/2021 at 9:27 PM, 514_AppleGuy said:

How in the Holy Name of F@ck does a case catch fire?

(Likely) Poor QC (quality control) and/or design...

Edit - oops didn't see that there was replies saying same thing already lol

P.S. also gives a whole new meaning to the term "being screwed" lol.

[Spotty]

1 hour ago, linuxChips2600 said:

P.S. also gives a whole new meaning to the term "being screwed" lol.

I guess you could say this case has some of the worst thermal performance Gamers Nexus has ever tested.

[Moonzy]

GN have a small update on the situation:

Spoiler

In one hand, they have a dangerous product

but they did solve the issue for it's intended purposes

GN's concern is with future uses which I'm not too sure if it's 100% justified, even though people do re-use things.

 

Now, if there's also an exposed Ground plane, then the story changes as having both of those plane exposed in the same screw hole is just another fire waiting to happen.

We might learn more in the 2nd video.

[leadeater]

1 hour ago, Moonzy said:

Now, if there's also an exposed Ground plane, then the story changes as having both of those plane exposed in the same screw hole is just another fire waiting to happen.

That doesn't matter, the riser is grounded, the motherboard is grounded, the chassis is grounded, everything is grounded as it's supposed to be. Everything is grounded to a common plane/point, ground. The thing that is wrong is a live 12V plane that can be bridged to a ground point by a conductive screw, you should never have an exposed 12V like this irrespective on how things are grounded as it's entirely possible for a relative to that 12V ground point to be introduced (literally anything falling on to it or touching it).

 

Your finger is a relative ground point, we're just lucky 12V doesn't really do anything to us and we generally are wearing clothing between us and the ground or path to ground. Here's something stupid to try, go get a power cable, chop the end off, expose the ground wire, plug it in to a wall outlet that is switched off, hold on to that ground wire, now touch this screw in the NZXT case holding in the riser. You know what you will feel? A nice tingle as current travels over your skin to ground. Now just be thankful it's 12V DC and not 230V AC.

 

TL;DR Grounding is irrelevant to the safety issue here. Live wires/power should never be exposed, it's a safety issue.

[Moonzy]

5 minutes ago, leadeater said:

by a conductive screw

I'm assuming it's being changed to a nylon screw, so after the "fix"

 

but yes, having an "exposed" 12v is not the safest thing to do

though most PCIE slots have their 12V solder point exposed on the back -thonk-