**A Power Supply Review: Testing the Gigabyte Model**

We recently had the opportunity to review a power supply model from Gigabyte, and our experience was quite enlightening. We were excited to see how this power supply would perform, but unfortunately, it did not meet our expectations.

One of the issues we encountered with this power supply is that it failed repeatedly when pushed to its limits. In fact, several users have reported similar experiences with this model, including dead or failing units. This has led us to wonder if there might be a more significant problem with this power supply than just a simple failure at the OP trigger point.

**Testing and Analysis**

Our testing revealed that the power supply failed on the way up to OP trigger point, which is unusual for this type of product. In fact, we found that only one other power supply model we tested failed in this manner, where it hit 60% load after an OP trigger and then died. This suggests that there might be a design flaw or quality control issue with the Gigabyte model.

We also noticed that the front end of the capacitors was missing, which is not typical for a power supply. We are not sure what this means in terms of functionality or safety, but it's definitely something we would like to explore further.

**The Story Behind the Review**

It's worth noting that we have had interactions with Gigabyte in the past, and they have repeatedly tested their power supplies to OP and claimed success. However, when we pointed out our findings and asked for a response, we were met with a lack of urgency or attention. This raises questions about whether Gigabyte is taking adequate measures to ensure the quality of their products.

We also want to emphasize that testing power supplies requires caution and expertise. If you're not familiar with the inner workings of a power supply, it's not safe to open one up for analysis. Our experience with this power supply serves as a reminder to approach these types of projects with care and caution.

**Conclusion**

In conclusion, our review of the Gigabyte model was disappointing, but it also highlighted some important issues that need to be addressed. We hope that our findings will help inform better design and testing practices in the industry, which ultimately benefits consumers like ourselves. If you're in the market for a new power supply, we would recommend doing your own research and carefully considering different options before making a purchase.

**What's Next**

We'll continue to monitor this situation and provide updates as more information becomes available. We also plan to review another power supply model that we tested earlier, which had some issues with its marketing materials. That video will be released soon, so stay tuned for more news from Gamers Nexus.

"WEBVTTKind: captionsLanguage: entoday we're going to do an uncut speedrun we're going to speedrun the explosion of a gigabyte power supply we're not going to have any Cuts in this video including the ad break and the whole point of this is going to be to address something that gigabyte said in its press release gigabyte when responding to an issue that we covered recently about the power supplies detonating said uh something very simple it said uh that the testing was done over an extended period of time and it kept saying over and over again that we were basically putting it under extreme load scenario now technically gigabytes had certain media outfits it was afraid to name us that's okay I would be too if I were gigabyte but uh it really meant to us cuz we most recently tested it and before us ays from Hardware buster did the testing back in October November found the same issue and I believe Igor's lab has found the same problem as well problem is when you hit op or in the process of approaching op which is over overow protection the power supply instead of shutting down like it should which is what a protection mechanism does will explode and like we said in the previous video it is no longer a protection if in order to prevent an explosion you preemptively explode it that's still a catastrophic failure it's that's not how it's supposed to work so uh we're going to do testing on this live basically uncut and see uh how quickly we can make this one explode to address gigabytes repeated statements of extended lengths of time and other mischaracterised so I'm going to take this in hand this off to Patrick Stone and he will hook it up to the DC load tester yes and we'll do we'll provide some commentary too now as we get into this this video is brought to you by Corsair and the 5000d and that is a case that we tested very positively it has an airflow version and a solid version we've liked the air flow version you can grab that at the link in the description below if you're trying to build a higher end midtower gaming PC okay so Stone do you want to explain the basics of what you're doing right now indeed yeah so the what I'm doing right now is just uh connecting all of the modular cables into the power supply and they've already been connected to our sm8 800 load tester uh the load tester's already been configured for the correct loads we're doing 20% Then 60% then 100% And then we'll go over those uh 100% values at 110 120 130 140 hopefully these things will all trip um at uh somewhere in that 130 to 140 range if not that's even more uh bad news for gigabyte yeah but um so we got it all plugged in here and I'm just going to and switch things on um so we just go to our uh uh varak yeah varak our input regulator and then from there it goes to the sm800 and then sm8 800 has a SM to20 electronic load on there um also make sure to switch on the power supply and that's good and then we turn this guy on when we turn the sm00 on um we should be able to look at the input voltage and it's at 116 so we'll go up here to the very to the varc and we'll adjust it down to 115 that's uh what it's spec for 115 so the whole point of the varak is to help us adjust for uh voltage coming from the wall that's right yeah changing loads yeah cuz uh you can you can guess that it's 115 or 120 but better to know yeah and uh so then if we look at the current current settings right now I'm at my first little piece I'm at 20% um and it's not turned on yet you can tell that by switching down here looking at the voltage no voltage except for 5 VSP and so as soon as we turn this guy on uh here with the power supply on button uh you see the fan spin up and you see voltage values and uh we'll we'll let it sit here for I don't know like maybe a minute sure um and and just hit this little stopwatch action um yeah and a note here too so our testing methodology we explained in the first piece talks about it a little bit more in the response to gigabyte where they tried to gigabyte was very strange in that response if you haven't seen it they basically said this is the fault of users and reviewers who are doing it wrong uh and ignored one very important Point clearly there's a lot of failures here without people ever going over power because if you look at the new egg reviews there's a lot of DOA yep and there's a lot of quick failures so they didn't really address that they retargeted the story specifically on overpower because they knew they could make it sound like it's ridiculous and they're hoping that people reading the press release will think I'm never going to go past 750 watts so I'm not worried about it yep but you should still be worried about it to be very clear the other thing we're addressing point to is uh again just the the fact that they're trying to kind of paint this as ridiculous thing um but in actuality it's I mean there's there's a lot of caveats here but you can always Spike above the expected load of your system in very brief intervals like when you first launch a heavy load application with a GPU we've plotted them before you'll you'll catch a spike that goes Way Beyond even the TDP of the GPU but it's only for a few milliseconds maybe and stuff like that can be a a bad scenario for the power Spike it's not designed for it yeah so what while you were talking there I went ahead and jumped from 20% to 60% and you could see that because of the increased load the fan kicked on to cool the cool this guy down um didn't think about it but maybe this is a good idea to stand the power supply up like this so when the Sparks come out if they do they shoot toward us was just about to suggest you don't do that it's a great idea right but it would make better video great video yes it's definitely true I will make sure insurance doesn't see this video or any of the other ones in the looking at the voltage coming out of this PSU you can see that at 60% uh things look similar to what we had at um at 20% and if we look at the current levels you can see that the currents jumped up on the 12vt rails significantly not so much on the five and the 12 vsb sorry the five and the right um 3.3 and the 5 vssb right um as we uh stop that one and jumped 100% load uh you can see I've got those current vales this is just about as high as we're going to go on 5 3.3 and 5 VB we we'll bump the these two numbers up just a little bit when we go for the real overp Power Protection um but these numbers the 12vt rail is what we're really going to be pushing cuz that that's what happens with your PSU like if you're going to be using it to reemphasize the point here the numbers that people are looking at on the screen uh so 11 11 11999 it's current that's right yeah right just to make sure that's clear and for people who aren't aware Watts current times volts gives you watts and so we're loading each rail a little bit differently exactly right yeah so if we look at the voltages again we can see the voltages here and uh We've also got the sm800 is nice feature where you can just pop that on and you can tell how much power is being pulled so right now we have 750 watts coming out of this guy right um and so now we'll just jump into uh probably something like like 10-c phases for the for the overpower so we're going to jump to 110% um you can see the power number jumps up uh and then if you're interested you can see the voltages um note that we were higher than this originally and it's not uncommon to see voltage drop down you know as you increase the load um and then you can see the values here again these guys stayed the same and these guys jumped up as we go to 120% um you can see again same these guys are jumping up and if we look at the wattage now we're up to 900 wat pull thing and I I want to explain a few things too while we get into this so uh first on the testing side we are changing our testing a little bit in this video not in a significant way the only thing we're changing is the um sort of the amount of increments we're doing and percentages as we approach our test for where does OP trip the reason we're doing that is simply because uh it's all the same we're just saving everybody time so and they're not necessary steps it doesn't if if anything right now what we're doing benefits gigabyte because this approach is putting it under less thermal load we're not testing it for as long so it's actually we're gigabyte we're giving the full benefit of the doubt by decreasing the test time uh so it's in a better position so did that just shut down it did at 140% a few seconds in 140% it shut down um and we can tell that not only uh by looking at the voltage numbers here you can see the voltages are all dropping like crazy uh this thing just shut off fan turned off right exactly and so now what I'll do is I'll turn the power supply off I'll hit the uh power supply off button on the tester and I'll even uh kill the ACN on the tester okay um and then so the next thing to do is okay so it it tripped now I wanted to see what was the exact trip point was it it was somewhere between 130 and 140 so now I've got to go back to 130 and then I've got to step up little bits at a time to figure out what's going on right is this where you normally leave it off for a few minutes uh so in in this case uh I would probably leave it off if if it got you know warm right now this is still fine in terms of temperature yeah I mean again we didn't really run it that long right and we're we're trying to in increase the amount of or decrease the amount of time we're taking here so uh I'm going to go ahead and flip things back on turn the sm00 back on uh fli the power supply back on and uh I've got got to bring the load back to the 20% level so now I'm back over here at 20% and I can verify that by looking at the current again um and then I will turn the power supply back on and we can see the fan flips on momentarily um and I think it's probably going to shut off again because it's only 20% load um once we start getting towards a potential catastrophic failure let's change the direction of that fan okay because even though it would better video commenters are going going to question my judgment even though you're the one who did that so um yeah so uh we should explain overpower protection a little bit I don't know at what point in this process you might have a a good opportunity to spend a minute on talking about why it exists and and what it's supposed to do yeah so uh in theory that moment would come after this trip if it doesn't explode yeah and if it does explode we'll explain it after there you go right yeah so I'm at 100% load I'm not going to stay there long I'm going to go straight up to 110 and then 120 after a few seconds and then I'm going to jump to 130 Andrew you might want to make sure the power supply is not shot at this point all right and then now that I'm at 130 now I'm going to go ahead and I'm going to jump over to a different rail I'm GNA go over to my 5volt rail and I'm just gonna just gonna push that up uh in terms of current um there we go I want about 10 amps is what I want on that and then I'm going to do the same thing to the 3.3 volt rail I just push that up 10% this is exactly what we did on the previous test so no changes here um and then the next thing to do is to go back to the 12vt rails and just start increasing them one at a time right and I'm not going to spend a much time on these I'm just going to go uh I'm going to hit maybe uh 1500s uh of of an amp on each rail or on each uh split of the 12vt rail and I'm not going to stay on there long I'll let it flash maybe three or four times for a few seconds and then I'm going to go to the next one do the same thing let's make sure power supply you got it in the shot okay cool and then this one same kind of thing all right and then the next one and what I would do generally is I would be using my spreadsheet right now and I would be documenting on the spreadsheet each each spot that passed right so I would I would say you know okay I passed here passed here passed here passed here pass pass pass and then I'm going to go right back around again and I'm going to go to the next level um and so yeah and to be clear as stone is saying here but uh testing for the lab for reviews that are produced written edited we have a lot of note taking we have a developed process we follow we have Sops we follow we have specific times we follow so everything's already controlled but again here the whole point is to address gigabytes claims of uh extended testing and show that they are factually incorrect especially considering the actual load time of this is relatively low still looking good at the moment we are surviving and should say this too Stone the uh op it is not good if it's set too high that's right I think it's it's really common misconception where and I used to think this too by the way where yeah so people think that power sply can push a lot of power without failure without shutting down so say it can go uh to 150% that's actually not good because the the chance of a catastrophic failure goes up potentially the the degree to which it fails gets worse and you really want it to shut down earlier in the range closer to its spec so where are we now for uh your test setup so this is uh this is just like the basically my third step along on the 12vt rail uhhuh um if I had my spreadsheet pulled up I could tell you exactly what um what the 12v rail Max was it looks like it just tripped again which is good do you need me to look something up um yeah if you'll go to the fourth tab on the top and this we can't show sorry at at the very top there uh the fourth tab over it should be large font maybe the next one over there I got one of them on huge font there oh go yeah so the 140% was 16 amps per Rail and you can see now that I didn't get there like so when I hit the 140% button it it completely shut down with OP which is good and then to get to exactly where it was I just do Little Steps like this and now I can add these numbers up to get an exact op trip point for this particular unit right um and so generally I would say like right now this this unit passed op correctly uhhuh because it it we were able to take it to op trip and then we were able to dial in to exactly where op be able to do it again though trip yeah and so now that it has successfully done the tripping I'm just going to bring it back online after like another minute or two right uh at like 60% cuz 60% is very common for a PC and then hopefully it's just going to keep running right along with no problems right we did have one explode at 60% though process yeah so successfully done the tripping by the way is uh the name of my Pink Floyd cover band oh yeah nice okay uh so I'm going to go back to 60% load right there and uh the power supply got switched off and the power supply on signal from here got switched off and uh it's not even really warm to the touch so so I don't have any real worry about did things get ridiculous inside this power supply so I'm going to go ahead and flip this guy back on and then hit this power supply on button it's oh that was the most impressive one we're going to get a blast shield for uh for future ones think I think I'm going to to get a blast Shield to go in that state is that expected Behavior Stone that's what that's what they're supposed to do right so after this one worked well after this one worked well oh my God which like the the that op testing that we just did yeah with like on the other power supplies that are not gigabyte GP DP dx50 GMS there's some uh there's some black soot flying around in there just flew in front of my face I going to check the security camera footage for that one we got a quick camera in the corner um I'm just going to I'm just going to why yeah just just in case I don't know are you scared um maybe they sent this unit to kill us I don't know that's the most impressive one so this I don't know if we bought this one on new egg or got it from a viewer it's is it this is one of the ones we purchased on New Egg and it's within the serial number range yeah uh which by the way our other Patrick went ahead and did the research for us and found out that which ones are returnable or not right and so separate video for this but uh gigabyte has been refusing warranties and rmas from what we've seen so far from you all those of you whove emailed us about your warranty at thas with gigabyte we'll cover that separately uh put some pressure on them to replace it but anyway this C can I go back though like all right so so again it Tri at 140 which it's supposed to like that that if every one of these the ones that have been successful and the ones that have failed tripped at 140 no problem right and then dialing it back in to find out the exact LP trip point at 130 this one looked like it was maybe like 134 132 somewhere in there right right and that that's a that that's an acceptable trip point it's not great but it's acceptable um and then from from there like I I honestly thought that I was going to turn it on at 60% and it going be fine yeah no big deal um can I that I was not ready for that I well I'm G to say one thing I'm glad I had you turn that up it looked like it shot electricity not fire like it looks like the blue flames oh yeah that that was I don't know if our camera is high speed enough to catch that but between this and security hopefully we got something um uh okay so the failure rate that we had in lab here was like 50% or so right uh and I I honestly do hope that when we plug in the next one here we got another one that we and we'll save that for another time we don't need to do that I think we got this is 22377 22400 so these are like really close in serial number and uh like maybe this one won't fail uhhuh right but after seeing that maybe it will fit so we need to wrap up say a couple quick things here um let's make sure everyone's clear on what just happened and why it is valid testing a couple points to bring up gigabyte kept using the phrase DC electronic load over and over and over in their press release okay uh I'll say this stone is is maybe a little more a little less skeptical than I am of these companies I've been in this industry a long time when I read that what I see is let's try to make this look ridiculous by showing users uh it's not real world kind of like how Intel does when it loses in benchmarks and then when it doesn't it it likes them again uh so they kept saying DC electronic load we should very quickly address that and at this point by the way if we have Cuts in the video there might be some here to speed it up because you're going to hit a 30 minute limit at some point right 21 right now all right and uh but it's already exploded so I think we can cut at this point um so DC electronic load can we quickly recap why that's uh not like some evil Boogeyman or something that they're trying to make it seem so this is electronic yes and it it's it's pulling DC out of the power supply that's right it's very it's very complex very complicated would you agree with the statement uh and this is complex algebra uh if power is X okay so we're defining power of the variable as X does x equal x is power power I think I think that that would be an accurate statement I'm pretty sure that's accurate so in other words your computer wants power uh your video card wants power DC power right AC comes out of the wall and converted to DC right and then the cables terminating at the load tester it's almost like your computer's a DC electronic load almost almost it's almost like gigabyte uses DC electronic load testers as does its Factory in the production of every single one of its power supplies that's weird I wonder why they don't like it now that we're using it yeah did I I was trying is my acting okay it it seem like I was actually confused I think it was pretty good it's pretty good so um all right so we've adjusted that power is power computer wants power this thing wants power it's all the same and by the way a realistic quote unquote workload as in in uh a computer is potentially more stressful than this is because the load is much more variable you have a lot more uh there's potentially more transients there's more Peaks and and the amplitude of sort of the demand changes so it's it's actually maybe even worse yeah and I'll throw one last thing in here um you know ays has clearly been doing this longer than we have right AIS is an expert like 100% yeah the the company cybernetics if you guys haven't heard about them maybe go look them up um they've got some some good stuff out there uh and his his point was that uh the problem is probably with the timings mhm the of the controller that controls the mosfets right and and so it's not like uh it's not like like this thing just has has a small issue that there's a design issue and I think he said that in his video and so you know go go go check out his statement don't just take my word for it and uh he published the whole statement he wrote it sent it to us it ended up published on tech power up so they've got it you can read the whole thing um another contact in the industry that we can't name but uh also believe that there might be a management issue on the mosfets specifically and uh part one showed what specific components failed if you want to see the mosfets that failed right and additionally um gigabyte said it's adjusting op down right to like 110 to 12% Eris said hey that's not possible that's too tight of a range and I read that statement in our response part two video uh additionally it it seems well it seems like maybe that's not the only problem with this power supply because again the user reviews indicate that there's a lot of people with dead or failing ones uh not all of these people are going to op and then back oh I should say that too gigabyte kept saying repeatedly they kept saying we're repeatedly testing them for long periods of time oh yeah yeah no like this is what you just saw is what is what we did yeah that's it's not like we're just like op test do it again it's not like do it again do it again no that's not what we're doing we should reemphasize one point here too so out of all the failures the rest of the power supplies we've tested failed on the way up to op trigger point testing uh whereas this one only one other one we tested failed in this manner where it hit 60% load after an OP trigger and then died is there anything else we want to say um I I I have to believe that gigabyte can do a better job yeah um you it we we're we are always trying to work with the companies to say hey we found something in your product yeah and you probably should check this out and then when they work with us on it products get better right and then you know maybe it's not a big story for us but that's okay what we're trying do is review products for all the end users anyway we just did this with another power supply it's not you'll see it online soon but we we were able to change their marketing before we produced the video they probably saw this series and went oh no uh AIS emailed gigabyte in October before he published a review they had a chance to fix it we emailed gigabyte I don't know in March or something originally we were talking to them and then again in June they had like six months with us almost a year between the reviewer community and the response to Aris was basically um nope we don't see it don't see the problem conveniently I don't know if you caught onto this but AIS did uh they gigabyte said they tested their power supplies to op and they succeeded they never said that they turned them back on oh yeah yeah yeah yeah so it's like well I if you only do have to test yeah great well and I think you mentioned too that maybe that could be one of the reasons why some of the units came out as DOA right if they tested it at the line and then it shipped a customer plugs it in turns it on and then it does what we just saw yeah I can't speak to gigabyte and their Factory standards but the I have been in power supply factories and we have talked to a lot of them done tours with them all this stuff everyone I'm aware of tests when 100% of the units off the line to make sure they meet some spec so if they're testing 100% of the units and including including OPB test maybe that's the source of doas I don't know but I think that covers it uh can I get you to touch this capacitor that sounds questionable but okay it'll just make your hair stand up for a few seconds yeah I think oh by the way don't open power supplies if you don't know what you're doing don't don't do that don't do this got to do other stuff to even work on this anyway uh cuz they're potentially unsafe so um is it is there an obvious failure oh did we save that for a different video this is the main switchers definitely we'll do a Boll shot of this we're going to cut and Boll at this point it's already failed so we can do cuts um the front end of the capacitors is missing not capacitors the let's take Boll of this separately we'll drop it in we'll talk about this more in another video but that'll wrap it for this one thanks for watching the speedrun we enjoyed it uh he might need to change his pants from the reaction I saw and I honestly was expecting it to work like I really was I was like I don't know if this one's going to fail that's what I thought that was exactly what I thought so uh that'll wrap it if you want to support us and our work we do here and the purchase of new underwear for Patrick Stone you can go to store. Gamers nexus.net and grab one of our Mouse mats mod mats or other items there shirts uh or you go to patreon.com Gamers Nexus we've got behind the scenes videos talking about this testing he's hosted several of them and you can check those out it's a great way to support us and our work and also get something in return thanks for watching\n"