[SOLVED] Corsair CX750M can't hold i7 4790K & 780 Ti playing The Witcher 3

[GreiverBlade]

my M12II 750 Evo is a bronze unit, fully modular and quite cheap (price not quality ) also my 1st Seasonic ... i couldn't be more pleased with it ...
previously i ran a FX6300+R9 290 on a Fractal design integra R2 650w (bronze closer to a silver just like the M12II Evo ) and both CPU and GPU were overclocked ... never ran into an issue, so i tend to think, even a good quality bronze is enough

short: 80+ bronze silver gold platinum titanium =/= quality of the PSU (the efficiency does a lot but the quality is not only the efficiency )

 

[R-T-B]

Speaking of the EVGA's... I got a 50% off EVGA PSU's coupon with my recent EVGA GTX 980 purchase... very tempted to grab one personally just to see how they fair, the reviews look stellar. Thinking about one of the Seasonic Platinum models (SP is the model number for them) but the Superflower ones are supposed to be crazy good as well. At 50% off it should be an insane deal. Don't be surprised if it shows up in the for sale forum after I'm done playing with it for a very low price, naturally

 

[Saru]

Guys... I was wondering.

What if its the PSU's fan?

You know, when I play Witcher 3 earlier in the day, the PC is cold, and it takes a while to shutdown. In the second shutdown, do not delay as much as the first time, and so on.

This may be the temperature of the PSU, right? What if it needs only exchange the fan of it?

 

[GreiverBlade]

Guys... I was wondering.

What if its the PSU's fan?

You know, when I play Witcher 3 earlier in the day, the PC is cold, and it takes a while to shutdown. In the second shutdown, do not delay as much as the first time, and so on.

This may be the temperature of the PSU, right? What if it needs only exchange the fan of it?

nonononope ... changing a fan on a PSU is a bad idea ... altho i did it, but it was on a LC-Power 350w for a blue led fan mod after the original fan got 3 blades on 7 flew out of the aforementioned PSU ... and i know that opening a PSU without being qualified is not a good idea ... (tho it works as intended ... i would not use that one on any of my rig, as minor as they could be )

unless you RMA it and Corsair see it's the fan and either change it or replace the whole PSU

 

[Saru]

nonononope ... changing a fan on a PSU is a bad idea ... altho i did it, but it was on a LC-Power 350w for a blue led fan mod after the original fan got 3 blades on 7 flew out of the aforementioned PSU ... and i know that opening a PSU without being qualified is not a good idea ... (tho it works as intended ... i would not use that one on any of my rig, as minor as they could be )

unless you RMA it and Corsair see it's the fan and either change it or replace the whole PSU

Yeah, you are right.

So, I changed the position of the PSU (I dont know if it matters) but the fan was down in the case and is now facing up

I tried running Prime95 and Furmark.

Last time, the PC survived 10 seconds. Now its 20 min and counting...

It can be the PSU, although, dont know, it might be playing a trick. hahaha

But, the fan was trying to getting air with the case's obstacles.

What do you think, guys?

 

[Rowsol]

20 minutes of that seems good to me. Did you play W3 after?

As far as PSU recommendation, I like any brand as long as it's gold or platinum. I think the Rosewill capstones are good for the money.

 

[Saru]

Forget about it, it was a trick...

But I've found something. It only shutdown with Furmark test.

The Prime95 test doesn't take it down...

 

[Rowsol]

GPU needs more juice than CPU.

 

[Saru]

GPU needs more juice than CPU.

So... PSU?

I dont think that the GPU itself can shutdown the computer, right? (temps were fine, with 71ºC at max)

 

[Saru]

But... I can't understand that.

The PSU could hold Furmark and Prime 95 test at the same time in 20 minutes...

And then, I stopped, restarted the PC, put only Furmark and then there was a shutdown in 5 seconds.

 

[EarthDog]

You shouldn't be running Furmark in the first place.. its a power virus.

But the fact that it did NOT die using both P95 and Furmark should tell you its likely not the PSU.

 

[Saru]

Well, doing some tests here...

I've changed the PCI-E cables (yeah, doing everything... hahaha), and running Furmark (6 minutes and counting...)

 

[Saru]

 

[Saru]

Hum... putting Prime95 test at same time, and 20 minutes with Furmark now... (I've just changed the cables)...

LOL, PCI-E cables with malfunction can shutdown a PC?

 

[newtekie1]

LOL, PCI-E cables with malfunction can shutdown a PC?

If it was slightly loose, and not making a good connection, yes. A cable that isn't making a good connection will heat up at the point of the bad connection, the connector will got hot, the hotter it gets the more current is required to push through the connector. Eventually, the current draw overloads the PSU and it trips the protection.

If that truly was the case, and I've seen it happen plenty of times so it is very possible, then the fact that the Corsair was actually built decently probably saved you.

 

[HWTactics]

Very interesting. Does your case have a vent cut out in the metal below the power supply? If there is just solid metal there, the fan was too close to a flat surface to breathe correctly. Sounds like you are on the right track.

 

[Saru]

If it was slightly loose, and not making a good connection, yes. A cable that isn't making a good connection will heat up at the point of the bad connection, the connector will got hot, the hotter it gets the more current is required to push through the connector. Eventually, the current draw overloads the PSU and it trips the protection.

If that truly was the case, and I've seen it happen plenty of times so it is very possible, then the fact that the Corsair was actually built decently probably saved you.

Well... after 33 minutes with Furmark, Prime95, Chrome, MSI Afterburner, GPU-Z opened, the PC shutdown...

I dont know if it's good or it was the cables... but I did my record running these tests with 33 minutes!

But... its not good enough, right? A PSU 750 could handle more? Or is it normal for my rig?

Just to remind:

CPU: i7 4790k 4.0
GPU: GTX 780 Ti
RAM: 16 GB DDR3
WC: Seidon 120V
HD: 1x HD 1TB WD Blue
DVD: 1x DVD RW
Fans: 4 fans
MOBO: GA-Z97X-SLI

 

[Saru]

Very interesting. Does your case have a vent cut out in the metal below the power supply? If there is just solid metal there, the fan was too close to a flat surface to breathe correctly. Sounds like you are on the right track.

Hey, HWTactics...

Well, my case has an air filter on the bottom.

I dont think its heat. Furthermore, there are pins that leave it with room to breathe (like the image I sent)

 

[Mussels]

that PSU should handle that load just fine, a loose cable could definitely have caused your intermittent issues.

 

[Saru]

that PSU should handle that load just fine, a loose cable could definitely have caused your intermittent issues.

Even with a shutdown after 33 minutes with all these programs opened?

I mean, it was normal? I never performed these tests before...

 

[Mussels]

Even with a shutdown after 33 minutes with all these programs opened?

I mean, it was normal? I never performed these tests before...

a loose cable doesnt care how long the PC was running, only when it was knocked loose. Don't forget you could have more than one issue here too - could be a loose or damaged power cable, AND something overheating (wait til its just powered off, then poke everything with your finger - RAM, VRM's, NB heatsinks, etc etc)

 

[HWTactics]

Saru you might want to try playing Witcher now with the changes you have made. 33 minutes running a stress test isn't a bad result in your case if you ask me; the 780 Ti's cooler is just barely capable of acceptably dealing with its 250W TDP, and a stress test like Furmark pushes it harder than games will, kicking the heat up even higher.

Edit: Noticed the heat is 66°C; running Witcher should reign it in closer to the 60°C mark, which is a very good temp and the card shouldn't have any problems @ that temp.

 

[Saru]

a loose cable doesnt care how long the PC was running, only when it was knocked loose. Don't forget you could have more than one issue here too - could be a loose or damaged power cable, AND something overheating (wait til its just powered off, then poke everything with your finger - RAM, VRM's, NB heatsinks, etc etc)

So... do you think the problem with the PSU is solved? We are discussing now something that can be overheating?

 

[Mussels]

So... do you think the problem with the PSU is solved? We are discussing now something that can be overheating?

One of the most common troubleshooting mistakes is to assume there was only one problem. Two weeks ago i was making threads about why an R9 280X was faster in my stock i5 than my OC'd i7

After a week of troubleshooting and testing i found two broken fan power adaptors, a faulty power cable to my CPU waterblock/pump causing it to throttle occasionally, and that the i5's board has a slightly OC'd PCI-E bus. Perfect example of why complicated issues are so damn complicated.

edit: and the 280x actually died as well, VRM fault.

 

[Saru]

Saru you might want to try playing Witcher now with the changes you have made. 33 minutes running a stress test isn't a bad result in your case if you ask me; the 780 Ti's cooler is just barely capable of acceptably dealing with its 250W TDP, and a stress test like Furmark pushes it harder than games will, kicking the heat up even higher.

Edit: Noticed the heat is 66°C; running Witcher should reign it in closer to the 60°C mark, which is a very good temp and the card shouldn't have any problems @ that temp.

Hahaha you read my mind.

I'm playing right now, some 20 minutes and no shutdowns... well, lets see and I will tell you.

Guys, I dont want to interrupt your sleep. Go back to bed (its 1:57 am here in Brazil hahahaha).

Thank you so much for helping me, its been great to see people helping each other. I'm very happy being in this forum.