Fun and small videos like this are what we want to do here -- not full deep-dive, but something simple! As the description says at the very top, remember not to open a PSU unless you know what you're doing.
Yo Patrick, I love your power supply videos! I'm always interested to watch but, there is one thing I'd love to see you explain; what are the components that hold a charge / we should not touch? Many youtubers (including GN) say not to fiddle with PSU's, can you show / explain why?
Really appreciate the variety! Those long 30 minute deep dives are really cool, but sometimes there's cool stuff going on that doesn't need to be that long.
That's exactly what warranty is for. And they offer a 10 year period for this model. It would be interesting to see how NZXT handles a post-disassembly warranty repair or replacement, of course in incognito mode.
"Oh Boy! 4 more screws!" This is exactly the kind of content I was hoping to see on this channel. Not in-depth, but knowledgeable, informative experiences. Mr. Stone is very good with his explanations and descriptions. He makes a great teacher. Seeing day-to-day operations and the work you put in outside the big main channel videos is super interesting to me. Or maybe I'm just envious of your work.
I've had a EVGA 650 BQ powering my system throughout the multiple PC's I've used. It amazes me to this day how dependable it has been over the last 8 or so years with very long on periods! I was just thinking to myself and about to open a new tab about how feasible it would be to replace a broken mosfet and it was answered in the next sentence. GN Extras was a awesome idea. Thanks Stone and team :D
Yeah I was using a evga g2 for 6+ years till I decided to buy a new one just out of paranoia with other new pc parts and such, but it held on really well and I ended up selling it to another guy
@@Sithhy Same I've been using the same HX750 80+ Gold for 8 years now. First I had it in an OEM (don't ask me how) running an APU + R9 280 and it's since been in my main system for 6 years seeing multiple CPU & GPU upgrades. Currently a 5600X + RTX 3070 -- both are undervolted however. I like cool and quiet. I'll probably try and upgrade the PSU at some point in the next few years but it has been a fantastic and stable unit.
finally, Patrick being the host. MOOOORE, just moooore! I like Steve but I also like PS. They've both got the qualities of hosts and have a unique personality.
Even GN doesn't have many sub channels like Linus, having at least 2 channels can diversify their content & video duration. Some people like short & some don't mind with long videos. As long as there is valuable information shown, it's fine. Don't chase quantity without taking care of quality, I saw that happened to Linus & all his other multiple channels.
I definitely like these types of videos. To me, there is a lot of value in the charm of the quick, lower production, off the cuff videos. There always seems to be something lost as channels grow, I'm glad GNX exists to hold on to that scrappy feel.
1:20 I'm not an electrical engineer...just an IT professional...but I don't see why letting a power supply run at low power would cause a failure. The only reasoning I can think of would be it's one of those "silent" PSU's that doesn't spin the fan up and it overheats and fails. Or... The mere fact it was on for months and months it was exposed to the crappy power grid inside your building and died because of dirty power/surges/etc (ie it wasn't attached to a sine wave UPS or worse it was attached to a square wave UPS).
When I saw the title I was expecting something more in-depth than this. At least you (kind of) showed the failed component which redeemed the video a bit. There are a lot of people who like a really in-depth investigation of this kind of thing. You only have to look a a few electronics youtube content creators to see this. If you were to go that far you'd pull in viewers from those areas as well. I would certainly be very interested in such content. You could even go so far as building custom electronics for a PC... You got a decent load testing rig and other proper equipment so why not explore other possibilities? You have the talent.
The reason the APFC MOSFETs blow up is the bad 1920-vintage circuit. The boost circuit uses "hot switching". This requires switches to be rated at 2 to 3 times the actual voltage to prevent switching transients from destroying them. A modern design would use an AC to DC PF circuit with NO BRIDGE. (The bridge is an artifact of poor education and refusal to look at designs less than 50 years old.) The switching occurs when the voltage or current are near 0. Also the LLC (not PWM) regulators are very inefficient at low loads. Modern designs have a uniform efficiency from low to high loads. Modern designs are also cheaper, smaller, more efficient, less ripple, less EMI, and better transient response than the 1920 circuits. One difference between Gamers Nexus and Matt Cremona is that Matt Cremona reads the comments and responds. Gamers Nexus does not.
I have not watched the video at all and I will instinctively start warning everyone against buying this PSU and start massive drama because there is no way the video will explain why I'm wrong and this is not a massive issue that has to be made big news of.
I will also not watch the video but instead read the comments section and start beef with random people about how their views are stupid and they shouldn't be allowed on forums.
I also have not watched the video, as I am in fact a bot that's here to spam advertisements for all kinds of scams by leeching onto the most upvoted comments. While I have your attention, may I interest you in some fine monkey png's?
I appreciate the nigh scheming weasel whimsical tune added to the first half of the video. Really brings up a nice dichotomy between something named with the suffix X-treme as a mountain dew flavor from the 90s like GN-EX up a few notches. Always a pleasure seeing the team have their own humor weaved a different way on this channel versus the GN one. That training Steve mentioned in their HW news for the weak video really shows. With practice they look like natural hosts. Great job team and THANKS STEVE!
Please do more power supply reviews! We need a true successor to JonnyGuru! His reviews were always my go-to, helped me pick out my EVGA P2 650 all those years ago. This thing is still running in my PC 7 years later, hasn't skipped a beat!
Great video. I know this PSU was sitting for a long time and was safe, but perhaps it's good to include basic warning about how caps can remain charged for a while even after being unplugged.
The 5VSB supply in ATX PSUs will drain the HV caps down to however low it will go before shutting off (50-80V in PSUs I have measured) in 15-20 seconds. There isn't much of a shock hazard unless you plug it in after the screws have already been removed.
@@gnextras Don't forget we have DIYer's making epoxy art in a closed unventilated rooms. Even if GN flashed a warning start middle and end of video people still going to hurt themselves, too many have HUAS (head up ass syndrome)
This was a nice surprise on my feed! Always happy to hear from Mr Stone. Maybe he could try to fix one of these on camera? Sorta like a “will it run?” that the car guys have?? I expect (hope?) that the average GN viewer is smart enough to know not to zap themselves. Maybe I’m too much of an optimist. 😅
Great vid Patrick! I have a question that may/may not have been answered in the past. I recently had a PSU that I replaced because it started a coil whine sound when placed under load by the GPU when I was gaming. The sound stopped when I replaced the PSU. Just curious as to what the issue could have been.
Been pretty lucky with PSUs and haven't had one die in use on me yet. Last one I had was a Corsair AX860i that ran 24/7 through multiple gaming systems even driving every generation of SLI from the 780 Ti to the 2080 Ti and it never had an issue, fantastic PSU. Only changed it as the PSU was getting on a bit after such a long usage and may have had issues driving a 3090 and now 4090. Hopefully the Corsair AX1600i lasts just as long without issue, so far 2yrs of 24/7 use and still rock solid.
1:32 - that moment destroyed my whole perception of how GN office looks like. For some reason my mind always assumed it isn't on ground floor, or at least part of it is above ground. It makes much more sense now.
Replacing a MOSFET is an iffy proposition. The original manufacturer tries to use FETs from the same batch. A replacement transistor will have different characteristics. This imbalances the current through the inductors and can cause saturation. Replace all the MOSFETs in the PFC or LLC regulator (hopefully with units from the same batch).
So what would the GN team do with a dead PSU like this? Are you going to try to warranty the unit, try to replace the mosfet, or just replace with the stack of PSUs that you have sitting around and e-recycle the dead unit?
when a fet fails, it usually backfeeds its control circuitry, like what happens sometimes when fets on GPU's fail, requiring the repair/replacement of the power management IC and/or totem pole driver circuitry, but as a precaution, its just easier to replace PSU's then to repair them
Nice informative bit. Not many will try to repair a PSU and I wouldn't recommend even opening one without a clue as that is one of the most dangerous parts in any appliance. Right after high voltage blocks in ancient TVs and similar stuff. But for those who know some electric basics this certainly is educational and entertaining.
Your NZXT E 850 is a Seasonic unit. It should have 10 years warranty . Why not send your dead unit to claim warranty ? I think a content on after sales service on PC components is a legitimate one, making sure manufacturers honor their warranty.
My friends EVGA supernova 750 popped a few nights back and killed his 2080 in the process, he thought it was all the psu's fault. I came by to take a look, it turned out that the power strip he was using didn't have a surge protector and had caused a chain reaction when it failed from a power surge his apartment had. Electricity is a wild thing.
Personally, I am comfortable tinkering with low amp electronics, but tinkering with anything responsible for power delivery SCARY. I would rather resolder half of SMDs on a complex board than change one MOSFET in a PSU or, admittedly, install a single mains outlet. Electricity kills; even if the operation by itself may be much simpler, you just can't let your awareness slip for a moment, and this is an enormous stress. One of these "do not try this at home" things: possible, but you absolutely need a proper workbench at a minimum, and the result is not worth much for an average user. Fixing a $500 TV by resoldering an EEPROM feels rewarding, and the worst that could realistically happen is that you end up burning $400-ish (well okay, you might drop a soldering iron and get burnt I guess), but meticulously trying to NOT DIE in the process of fixing a $100 part is just in an another ballpark of stress levels entirely.
We used to have regular power failures at work due to the company buying a lot of cheap PSUs that they figured were cheaper to replace often. I doubt we got more than three years out of most (we left them on 24/7). I've got Seasonic and BeQuiet at home.
As Marvin the Martian would say "an Earth shattering kaboom". I had an Antec power supply running a system for 14 years. It only died when I disconnected things to replace the motherboard and CPU (and RAM). I'm not sure how long it would have kept running the old system if I had just left it alone, but it definitely served its time.
Yo Patrick, I love your power supply videos! I'm always interested to watch but, there is one thing I'd love to see you explain; what are the components that hold a charge / we should not touch? Many youtubers (including GN) say not to fiddle with PSU's, can you show / explain why?
when you show the broken part of a device in the future please try to get it from several angles. from the one shot we got I could not actually tell what I was looking at and visually identify the evidence of the break. That said I would very much like more videos that show the disassembly and failure identification for instances of broken computer parts, including peripherals, as well at failures of tools used to work on fixing computers like multi meters and soldering stations.
From my experience (at least with other SMPS, not computer ones), if the main switching mosfet has blown it's almost always taken the controller IC with it, or perhaps the IC took out the mosfet but same difference. I've learnt my lesson and just order new parts for the whole section instead of acting surprised when the new mosfet violently explodes.
I worked on a very high current, low impedance industrial ps years ago that used a bank of tfets in parallel. (Used in electroplating) The operator standing very near it when it let go said he damn near pooped his pants! Chain reaction catastrophic failure. Carnage in a box. They now have fresh air supply to these devices.
The only thing that has been a little annoying using the mod mats is finding black screws when i drop them, probably not a issue with studio lighting, white modmat when? gn blue modmat? i have a 3 faulty Units that could be used to make this type of content, turn on and seem to work but cause a system crash if you want them
Why do you not shut down the streaming setup if you're not using it? Doesn't really fit with your emphasis on power efficiency when benchmarking components
Hi GN/Patrick, sorry if this has been asked already. Is the fact that it smoked when failing a particular indicator that it was a MOSFET that failed, or was that just an educated guess?
had a current control IC in a antech PSU spark and put out a nice light show in an old home server, but the UPS did not trip, had to yank power cable. because of that i now require every UPS in my house to have FULL access to the UPS power button.
tbh a sticker like that on a psu should still be legal .. this is a device that has the full blown mains power on it and is potentially dangerous and if you drop something in there or cause a short you risk yourself, others, burning down whatever building you are in, and dammaging it
A thought on the potential reason for the failure, with it running on a low power out for a relatively long time like that could it be overheating of the mosfets from the fan being off at that low a power, or potetially that the fan only goes on at a specific load level and that level is determined by the load on the 12v rails and you had a higher load on the 5v rails with the leds?
Months on end idle is normal operation and doesn't lead to PSU failure. You probably just caught one with a dud component in it, it's just unavoidable. That's what warranty is for.
I dub thee GNX. Another thing that causes failiure in PSU's is a brown out or other issues with electrical either from the utility company or issues with wiring in the building. I love the sound of that, GNX...It has an X. Make it happen.
"Maybe that's what got it; just running low power for a long, long time. We're talking months at a time" Slight concern as I look at my FSP PSU that's been running 24/7 since 2012, running a mostly idling PC.
I have a EVGA SuperNova 1000 G2 that I bought in 2013. It started out in an i7 4770K system, and it's in my present rig which is a R9 5900X / 2070 Super, blah, blah, blah, etc. I never shut my systems down (so 24/7) running World Community Grid (100% CPU usage) and gaming. So this PSU has been running continuously going on 10 years. Granted the PSU has been overkill in pretty much all the systems it's been in.
Was there a 5 volt load on the PSU? It's been a few years since I've seen it, but some PSUs need a 5 volt load for the 12 volt to work properly. I might be 10 years out of date, though 😂
Components usually die because other components died or are about to die. It becomes a chain of events that lead from one little component failing causing an array of others to fail because of it and eventually something blows up that will make the device stop altogether. So yes, replacing the obviously broken MOSFET will not fix it. It will most likely blow up the new one instantly because the reason it blew in the first place is still there.
For me there is one question that matters when power supplies fail. Did it inflict any damage to other components, or did the psu contain the failure, and just stop providing power?
is the NZXT C850 effected? was just about to buy this power supply "NZXT C850 (2022) 850W 80+ Gold Fully Modular Power Supply" just want to make sure its a good buy or not lol