Whenever you have diodes or transistors in parallel, it's really best to replace both the failed one and the good one together with two diodes you bought at the same time(increasing their chance of coming from the same batch). This is because the diodes have a negative temperature coefficient(temp goes up, resistance goes down, current goes up, heat goes up,...) and the only thing keeping a little balance is the shared heatsink. Really make sure to order at least two of those diodes and two of those transistors, maybe even four, just in case it happens again :P
HX650 psu had burnt mosfet and diode 1n5406, replaced them and psu came back to life. All voltages ok but under 40-50% load same components died again, one transistor and diode. Any ideas?
a friend gave me a RM1000 that would not turn on, it would click, but no power, no lights. I sent it in to Corsair, and they sent me a brand new one in the box WITH all the cables and accessories... I just had to talk to their tech support. I made out like a fat rat.
Hello, nice video ! i have approx the same, but where do i find the 5vsb and 12v rail within those output connector ? i do not have any cable and i do not see any yellow wire. thanks
So six months ago found a similar fault with my HX750 PSU. Looking at your video, the fault is almost identical, except both larger IPW60R190 FETs had died, and the fuse had exploded. I bought new ones, replaced them, but when I powered it back up, they were killed again. Also, the (single) C3D06060A diode blew apart (front came off with one leg, the other leg stayed on the heatsink). I did a bit more fault finding and found that the smaller diode had also failed. Replacing all three components (this time only one FET so I don't lose two devices if it blows again), I powered it up and enabled the PSU with no load. The fan spun for a second or so, and then turned off. I'm guessing something further along the chain is stuffed since I can see the boost converter output reading the right output. Not sure if I should spend much more time I should spend, and am I really going to trust this supply with anything other than a cheap PC now, but I'll try a few more things.
@@rPaguHaP From memory, I had to place a small load on the the output, but I think there may have been one other smaller diode that had failed. I got it back up and running, and it's powering my secondary PC since I am too worried to use it with my newer parts. But it works again!
got an hx1050 clicks on the off i get 5v on power ok line tge 5v and 3v boards are damaged i removed them off the pcb where should i look to figure out whats causeing the power on then power off
I don't recall if there was anything specific going on in this video, but I usually power everything off of my bench PSU. It's a TTI EX752M that'll do CC/CV up to 150 V/2 A.
The root cause of the failure was the diode between the rectified grid and the PFC boost output. It fails if there's a giant spike on the grid, or if it's of poor quality. Lightning strike is improbable; would have been more damage. Random failure seems like the most reasonable explanation given the minimal amount of cascade failure.
valdarmort , The reason for the 600 V rated slow bypass diode is to pre-charge the boost capacitors on power up. Without it, the build up of current in the choke/fast diode/capacitor loop would charge the capacitors to DOUBLE the peak voltage. Probably all practical implementations of the boost topology use this arrangement. But the bypass diode will pass an immense inrush current which depends on the line voltage and impedance -- including any limiter circuit possibly added to the input of this power supply. A malfunctioning limiter could be the root cause. A thermal NTC limiter does not work if it is not allowed time to cool down before the next power up. Similar problems arise on relay type limiters.
I heard that testing a Power Supply with a meter doesn't always show faults because theres no load on it but I figure if you're getting the 5v and the 12 v readings it at least a good sign
People say don't touch a PSU for days after it's plugged out as there can be power in the capacitors. How long did you leave it before touching it? Could you vaccum dust out of it straight after plugging it out?
I'm trying to rescue a HX1050 with a completely blown Thermistor at TH1 (along with shorted mosfets and diodes). It's a NTC, I assume. Did you ever find out what to replace it with?
Great fix! You should post the video on /g/. I'm always skeptical about Corsair Power Supply Quality. I've seen them fail ever so many times whilst a cheaper Cooler Master or Xigmatek seems to just keep chugging on just fine. I think they are just very sensitive to the quality of the mains input voltage. So kids here is the life lesson "Always use a mains voltage power conditioner."
If you want to post the video on /g/, feel free - I don't usually post my videos on there any more unless it's in relation to something specific. I would not recommend any "mains voltage power conditioner". They're usually just bullshit, and will not do much to help you if something actually goes severely wrong on the grid that the protection circuitry in your power supply doesn't already do.
FFcossag Don't voltage regulators such as these help reduce both transience and voltage spikes from the mains? www.fsp-group.com.tw/index.php?do=proinfo&id=61 I don't know it seems to be the main reason for PSU failures considering we have alot of power outages and constantly flickering lights in my country.
Dirty Reviews Are you sure that's a voltage regulator? It's advertised as a UPS, and it definitely doesn't look big enough to be a double conversion UPS. I can't find any info on it, but I'd be about 90 % certain that's just a line-interactive UPS that's basically going to be a straight pass-through unless you get a long-lasting brownout or power failure.
Actually It definitely is not a UPS system. I own the 750 watt rated version and it merely indicates that it regulates the mains voltage and prevents spikes. I think its a back to back converter so something on the lines of AC>DC>AC. Its definitely heavy and probably has some large coils/caps inside so not one of those bogus Chinese "conditioners". I actually would love it if you did a video on these because they are right up your alley. Sort of like a UPS but without the battery and for people such as myself who have mains stability issues in their region.
FFcossag I just did some quick duckduckgo searching and all I could find was that it contains a "Boost and Buck Stabilizer" according to a marketing document.
also i have a cooler master(non pfc) 500w (many say it is actually a 450w), it has double bridges & double switch mosfets in parallel. it works sometimes when switched on , sometimes not only 5vsb is available, the fans turn on for .5 sec & then off then i have to hard restart it. i checked the voltages in bios sometimes the 5 v & 12 v fall below the lower 5% tolerance, i.e 4.7V & 11.3 v, what could be the problem the mosfets or the schottky diodes or caps, i don't want to discard it, it is beautiful ps.
I got some NOX psu from a friend that said it fried his motherboard. I checked all voltages and it seemed fine, tryied it on a motherboard and damaged it somehow, now it doesnt boot because an error message appears while on POST.. As well as some crazy temperatures and voltage readings on BIOS.... Cpu over voltage! Cpu fan error! Chassis intrusion! Fatal error! System halted! What could it be?
which one is best brand for psu among antec,corsair , thermaltech...???i want to buy antec vp450w psu.i have no ups . so please tell will this psu save my pc from over voltage or current?????
even i have no gpu in my pc .becoz i am not interested in gaming...so i think 450w good enough ....but having no ups iam concerned for voltage up down ....
Hi! How much do the parts cost that you need? This thing costed (if it's the silver edition) around 180€ here in germany (gold 150€) - so it's a pretty pricey PSU. Nice Video!
Nice video! I have one Enermax PRO 550W (with no warranty) but it is as new.. i used it for a short time. Now it seems have a problem at 5V gray wire (Power_OK), the other voltages are ok, only the gray wire have 0V (and i think this is the cause of computer dont start).. but i dont know how to fix it..
Nice repair! Wish i had the knowlage! Just brought home a Silver Power 650w PSU that just stopped working. Hoped it would be a simple "visible" fix, but as a noob i cant see anything wrong.. :P All the caps looks fine. And it looks really advanced inside, not as easy as the older PSU's.. Several small controller boards and stuff. So its probably trash, atleast for me. :-)
What a shame. PFC failures like the one in this video seem to be quite common for active PFC power supplies in general, might be worth just measuring the primary side FETs and diodes for shorts.
One bad mistake he made it turning on the pc then connecting the processor power plug while it powered on ya can blow something easily i seen it been done before nasty
Oh yeah, don't get me started on the cheap ones. I've seen cheap ones labelled for 600W or more, yet the guts inside can barely do 200W!!! Those always blow up nice n good, usually frying the motherboard in the process!
Your accent was difficult to follow unfortunately for me. The only word I really heard was at the end, I heard "Excellent"! So what ever you were saying in the video I guess it is fixed. Yes? No? Excellent?
Whenever you have diodes or transistors in parallel, it's really best to replace both the failed one and the good one together with two diodes you bought at the same time(increasing their chance of coming from the same batch). This is because the diodes have a negative temperature coefficient(temp goes up, resistance goes down, current goes up, heat goes up,...) and the only thing keeping a little balance is the shared heatsink.
Really make sure to order at least two of those diodes and two of those transistors, maybe even four, just in case it happens again :P
Solid advice, I'm of course replacing the failed pairs in pairs for those exact reasons.
People who buy a 1000W PSU, rarely use more than 400-500W.
i DOUBT that any of PSU can give 1000w wires are sooo thin they can not give so much amperes.
Miners?
Nice step by step repair, thumbs up!
HX650 psu had burnt mosfet and diode 1n5406, replaced them and psu came back to life. All voltages ok but under 40-50% load same components died again, one transistor and diode. Any ideas?
Who throws a product with a 7 year warranty in the trash? people these days... (at least it was good for you)
But nice repair!
a friend gave me a RM1000 that would not turn on, it would click, but no power, no lights. I sent it in to Corsair, and they sent me a brand new one in the box WITH all the cables and accessories... I just had to talk to their tech support. I made out like a fat rat.
I do have a 3 years old Corsair HX850 that burn up my entire PC(it was beautiful) and ive lost the payment voucher. So no warranty...
Hello, nice video ! i have approx the same, but where do i find the 5vsb and 12v rail within those output connector ? i do not have any cable and i do not see any yellow wire. thanks
What bridge rectifier it is using?
Have you got schematic diagram this ?
so if the pfc fails can you use the psu just by connecting the rectifier o.p to the caps.
It requires a lot of guts (experience) to bare hand touch everything on a exposed power supply. Congrats, man.
I'd rather say it only requires guts if you don't have experience. Once you know what you're doing, it becomes a whole lot less scary.
@@FFcossag and only use 1 hand...
my seasonic 850watt platinum starts showing drop in voltage
12v became 11.4v
5v shows 4.4v and 3.3v became 2.9v
what part do you think cause it?
So six months ago found a similar fault with my HX750 PSU.
Looking at your video, the fault is almost identical, except both larger IPW60R190 FETs had died, and the fuse had exploded.
I bought new ones, replaced them, but when I powered it back up, they were killed again. Also, the (single) C3D06060A diode blew apart (front came off with one leg, the other leg stayed on the heatsink). I did a bit more fault finding and found that the smaller diode had also failed.
Replacing all three components (this time only one FET so I don't lose two devices if it blows again), I powered it up and enabled the PSU with no load. The fan spun for a second or so, and then turned off. I'm guessing something further along the chain is stuffed since I can see the boost converter output reading the right output.
Not sure if I should spend much more time I should spend, and am I really going to trust this supply with anything other than a cheap PC now, but I'll try a few more things.
I had the same problem. Did you fix it?
@@rPaguHaP From memory, I had to place a small load on the the output, but I think there may have been one other smaller diode that had failed. I got it back up and running, and it's powering my secondary PC since I am too worried to use it with my newer parts. But it works again!
@@rownadoherty Thank you! Smaller diode like SMD diode? I Need to know here to look at :)
@@rPaguHaP I'm fairly sure all the parts I worked on were through-hole. I normally start by investigating the large semiconductors and resistors/caps.
Seems that most of the modern PFC based PSUs fail in the PFC area. Bad caps on the output isn't much of an issue at all now.
nice one,could you tell me where you found the data sheet for this psu.i am looking for the corsair rm1000i.cheers
I just used Google.
Lucky that the driver ic wasn't shorted! buuut great repair!
what were the costs of the broken transistor?
New transistors and diodes came up to about 30 € from Mouser.
got an hx1050 clicks on the off i get 5v on power ok line tge 5v and 3v boards are damaged i removed them off the pcb where should i look to figure out whats causeing the power on then power off
Very nice diagnostic work. The unit may indeed still be under warranty, but no need. It sounds like $30 did the trick.
what are you using for the limiting of input current? can you show us in a future video?
I don't recall if there was anything specific going on in this video, but I usually power everything off of my bench PSU. It's a TTI EX752M that'll do CC/CV up to 150 V/2 A.
@@FFcossag you are powering the psu under repair straight after the bridge rectifier with DC voltage from you bench power supply?
what would be your best guess as to what caused the power supply to fail. lighting strike or just bad luck with parts or overload
The root cause of the failure was the diode between the rectified grid and the PFC boost output. It fails if there's a giant spike on the grid, or if it's of poor quality. Lightning strike is improbable; would have been more damage. Random failure seems like the most reasonable explanation given the minimal amount of cascade failure.
valdarmort ,
The reason for the 600 V rated slow bypass diode is to pre-charge the boost capacitors on power up. Without it, the build up of current in the choke/fast diode/capacitor loop would charge the capacitors to DOUBLE the peak voltage. Probably all practical implementations of the boost topology use this arrangement. But the bypass diode will pass an immense inrush current which depends on the line voltage and impedance -- including any limiter circuit possibly added to the input of this power supply. A malfunctioning limiter could be the root cause. A thermal NTC limiter does not work if it is not allowed time to cool down before the next power up. Similar problems arise on relay type limiters.
I heard that testing a Power Supply with a meter doesn't always show faults because theres no load on it but I figure if you're getting the 5v and the 12 v readings it at least a good sign
For diagnosing bad caps, that's true. This isn't a case of bad caps, however. We just want to know if the unit runs.
@@FFcossag will the voltage drop when loaded in case of bad caps?
What's the part number of the blown components ?
CSD06060 and 35N60C3.
FFcossag Ok, by 650 volts I have expected IGBTs, and thank you :)
FFcossag Does this PSU use a half or a full bridge to drive the output stage, and at which frequency ?
Alex
High Voltage
I don't know, none of that was relevant to repairing the unit.
FFcossag Ok :)
People say don't touch a PSU for days after it's plugged out as there can be power in the capacitors. How long did you leave it before touching it? Could you vaccum dust out of it straight after plugging it out?
Just don't poke around too much.
Would you know the value of TH1 (thermistor) on the corsair HX1080 as it is burnt
I'm trying to rescue a HX1050 with a completely blown Thermistor at TH1 (along with shorted mosfets and diodes). It's a NTC, I assume. Did you ever find out what to replace it with?
@@TheGiddeGaddNo idea of the value of TH1 would you know the near value or equilivant value??
@@ehmd2912 I don't know, I was hoping you would know :) Mine was blown to pieces, no way to read the part number
what is wrong if a pc power supply keep blowing mosfets? (1600W)
All kinds of things.
I hope you got the bag of modular cables with that power supply, because that is another added expense just to use it.
Great fix! You should post the video on /g/. I'm always skeptical about Corsair Power Supply Quality. I've seen them fail ever so many times whilst a cheaper Cooler Master or Xigmatek seems to just keep chugging on just fine. I think they are just very sensitive to the quality of the mains input voltage. So kids here is the life lesson "Always use a mains voltage power conditioner."
If you want to post the video on /g/, feel free - I don't usually post my videos on there any more unless it's in relation to something specific.
I would not recommend any "mains voltage power conditioner". They're usually just bullshit, and will not do much to help you if something actually goes severely wrong on the grid that the protection circuitry in your power supply doesn't already do.
FFcossag
Don't voltage regulators such as these help reduce both transience and voltage spikes from the mains? www.fsp-group.com.tw/index.php?do=proinfo&id=61
I don't know it seems to be the main reason for PSU failures considering we have alot of power outages and constantly flickering lights in my country.
Dirty Reviews Are you sure that's a voltage regulator? It's advertised as a UPS, and it definitely doesn't look big enough to be a double conversion UPS. I can't find any info on it, but I'd be about 90 % certain that's just a line-interactive UPS that's basically going to be a straight pass-through unless you get a long-lasting brownout or power failure.
Actually It definitely is not a UPS system. I own the 750 watt rated version and it merely indicates that it regulates the mains voltage and prevents spikes. I think its a back to back converter so something on the lines of AC>DC>AC. Its definitely heavy and probably has some large coils/caps inside so not one of those bogus Chinese "conditioners". I actually would love it if you did a video on these because they are right up your alley. Sort of like a UPS but without the battery and for people such as myself who have mains stability issues in their region.
FFcossag
I just did some quick duckduckgo searching and all I could find was that it contains a "Boost and Buck Stabilizer" according to a marketing document.
also i have a cooler master(non pfc) 500w (many say it is actually a 450w), it has double bridges & double switch mosfets in parallel.
it works sometimes when switched on , sometimes not only 5vsb is available, the fans turn on for .5 sec & then off then i have to hard restart it. i checked the voltages in bios sometimes the 5 v & 12 v fall below the lower 5% tolerance, i.e 4.7V & 11.3 v,
what could be the problem the mosfets or the schottky diodes or caps, i don't want to discard it, it is beautiful ps.
Nice repair! Which power supply with current limiting have You used to test this PSU?
I have a TTI EX752M - they're great for this kind of stuff.
I got some NOX psu from a friend that said it fried his motherboard. I checked all voltages and it seemed fine, tryied it on a motherboard and damaged it somehow, now it doesnt boot because an error message appears while on POST.. As well as some crazy temperatures and voltage readings on BIOS....
Cpu over voltage!
Cpu fan error!
Chassis intrusion!
Fatal error!
System halted!
What could it be?
Could be bad caps; trashy output.
which one is best brand for psu among antec,corsair , thermaltech...???i want to buy antec vp450w psu.i have no ups . so please tell will this psu save my pc from over voltage or current?????
Nothing saves your PC. All the brand names are good as long as you buy an expensive unit. Cheap ones are trash.
even i have no gpu in my pc .becoz i am not interested in gaming...so i think 450w good enough ....but having no ups iam concerned for voltage up down ....
Hi!
How much do the parts cost that you need?
This thing costed (if it's the silver edition) around 180€ here in germany (gold 150€) - so it's a pretty pricey PSU.
Nice Video!
Parts were about 30 €. I think it's older than the 80+ models, though; there's a newer HX1050 on the market now.
Nice video! I have one Enermax PRO 550W (with no warranty) but it is as new.. i used it for a short time. Now it seems have a problem at 5V gray wire (Power_OK), the other voltages are ok, only the gray wire have 0V (and i think this is the cause of computer dont start).. but i dont know how to fix it..
I am feeling very nervous when you started touching the internal parts of an active PSU.
You just need to know where to touch them. ;)
Nice repair!
Wish i had the knowlage! Just brought home a Silver Power 650w PSU that just stopped working.
Hoped it would be a simple "visible" fix, but as a noob i cant see anything wrong.. :P
All the caps looks fine. And it looks really advanced inside, not as easy as the older PSU's.. Several small controller boards and stuff. So its probably trash, atleast for me. :-)
What a shame. PFC failures like the one in this video seem to be quite common for active PFC power supplies in general, might be worth just measuring the primary side FETs and diodes for shorts.
Is it possible to measure the diodes when they are still soldered to the board? Or do i have to take them out?
Often it's fine, if they do measure short then you should take them out to verify.
One bad mistake he made it turning on the pc then connecting the processor power plug while it powered on ya can blow something easily i seen it been done before nasty
Very interesting. Thanks for the video.
What language do you speak here?
lmao
Bro I Have a Thermaltake Toughpower XT Platinum 1275 Watt That Power Supply Load Taking Problem?
I Need Your Help
There isn't much I can do for you.
Good vid as usual 👍🏻
if you measure a short over a diode in circuit, it is 99% not normal, as having a diode directly across an inductor makes no sense.
I've never had a PSU fail on me I guess Im lucky but it's a common complaint that especially cheap ones tend to fail often
Oh yeah, don't get me started on the cheap ones.
I've seen cheap ones labelled for 600W or more, yet the guts inside can barely do 200W!!!
Those always blow up nice n good, usually frying the motherboard in the process!
Nice, thumbs up :)
not adiod a voltage regulator
Your accent was difficult to follow unfortunately for me. The only word I really heard was at the end, I heard "Excellent"! So what ever you were saying in the video I guess it is fixed. Yes? No? Excellent?
It works well!