-
-
Guys, I'm adding a 4th power supply, this one is the least, but I'm adding it because it's a good brand, which makes me think that... it will be a bit short in the long run.
Antec High Current Gamer HCG-620M plus bronze
This one, I don't know why, is the most expensive of all and the one with the least power.
I'm going to try to tinker a bit with the power supply by opening it up and removing the layers of dust that have gotten inside, to see if... the flute sounds
-
stick with the Cooler Master V650S Plus Gold
your Tacens broke, right? they are flimsy, that's why they are cheaper,
regards
-
stay with the Cooler Master V650S Plus Gold
did your Tacens break? they are weak, that's why they are cheaper,
regards
Anyway, thanks for the answer,
I'm really annoyed because I've run Prime95, Aida64, HWMonitor and it doesn't restart. But as soon as I play Battlefield in multiplayer, it crashes without any error message or anything (it just restarts without more) I've removed RAM, from tri-channel to dual and it's the same. I've disabled HT, Turbo... and it's the same
I don't want to keep guessing and spend the money, it's over 100 euros. What did you have in mind for RAM. It hurts to spend money twice on the same thing.
But mostly, it's the fear that if I change it, it will keep doing it and with 100 euros less.
Everything indicates that maybe the voltage isn't stable under full load (even though it doesn't reach 100% while gaming) but when I run torture tests, it doesn't budge. That's what's bugging me... why does it happen when I'm gaming but not when I'm stressing it?
I've cleaned the dust with a shovel, but it's happened again. Inside, the PSU looks intact. I just have to pray and assume it has voltage drops.
I'm now waiting for a 6-core Xeon, and I was thinking of waiting because the TDP is 90 and see if consuming less... I can fool it for a while.
I have too many assumptions as you can see
-
First I would say that if you have overclocked the micro, memory or graphics, remove it. Play Battlefied with everything set to default to see if it restarts. An overclock that is not 100% stable can cause problems.
If on the other hand you do not have overclocked, run various torture tests like 3dmark for the graphics and OCCT for the micro, and let us know. Also check the memory as I mentioned, at what speed you have them set. And it would not hurt to check the motherboard if everything fails, to see if it has any faulty capacitors.
Check the temperatures of the micro and graphics also while running the tests and while playing. I know that as you indicate you have already run some tests, but to be sure it is better to run several.
-
First I would say that if you have Overclock done to the micro, memory or graphics, remove it. Play Battlefied with everything set to default to see if it restarts. An Overclock that is not 100% stable can cause problems.
If on the other hand you do not have Overclock, run various torture tests like 3dmark for the graphics and OCCT for the micro, and let us know. Also check the memory as I mentioned, at what speed you have them set. And it would not hurt to check the motherboard if everything fails, to see if it has any capacitors in bad condition.
Check the temperatures of the micro and graphics also while running the tests and while playing. I know that as you indicate you have already run some tests, but to be sure it is better to run several.
I think I have checked it... since I changed the heatsink to the 212evo it has not gone over 65 degrees, until a month ago I had a simple overclock to 3.6ghz fsb 167 1.21v and the mems at 1333. For a while it went well, but suddenly it started to fail and I removed everything, until I left it at stock, and at stock it still does it.
I have run several stress tests, but precisely the OCCT, gave me problems, all the time I saw the little message of vcore 1.21 value exceeded. I tried putting 1.30 and the same, and at stock also. So it does not let me pass it. With the same message maximum vcore value reached.... I do not know if this is symptomatic. On the other hand with the OCCT the vram appeared 1.65v when my mems only support 1.5. I tried to fix it from the bios, but the program kept saying 1.6. I did not want to give it much credibility.
On the other hand I passed the memtest and nothing.... -
Have you thought that maybe it's a problem with Battlefield and not the hardware? It seems that many people have experienced this with BF.
-
Have you thought that maybe it's a problem with Battlefield and not the hardware? It seems that many people have experienced this with BF.
You just killed me.
I have seen in several forums that more people are experiencing this, but... won't the same thing happen to all of us? A power supply problem? That actually requires a lot of juice for 64-player maps?This morning I received the x5670, I assembled it, put its paste... cleaned a bit the socket that had dust. I've been 2 hours with the game on, and now it doesn't restart :S
Either one of two things, either the 40w less consumption (which I understand requires less voltage) has been noticed, or the processor would be defective?
I must also say that I was amazed to see the temps, or I went too far with the paste... or it's the first time I see a processor at idle at 16 degrees.... while playing battlefield 30 degrees the core that's hottest.... I mean, you have to see it to believe it o_O -
It would be strange if it was the micro if OCCT didn't give errors, some magical factor of computing has appeared…
As for the temperature, unless you are wearing an anorak around your house I don't think it is correct. 16ºC at rest in the best of cases (which it won't be) would need an ambient temperature of 16ºC in air cooling. -
It would be strange if it was the micro if OCCT didn't give errors, some magical factor of computing has appeared…
As for the temperature, unless you are wearing a parka around your house, I don't think it's correct. 16ºC at rest in the best case (which it won't be) would need an ambient temperature of 16ºC with air cooling.Well… the truth is that at home it's cold, I don't have heating. I've moved to a big, old house (30 years old) in a village. And I'm ashamed to say it but, when I come back from work at 6 in the morning, on the ground floor it's 11-12 degrees. During the day it may rise to 14-15 without turning on the fireplace. Even so, I also think it's exaggeratedly low. The i7 had between 25 and 30 at rest. Could I have put too much paste on and not have good cpu-heatsink contact? But with the i7 it was the opposite, too much paste... too much heat, it didn't dissipate.
OCCT did give me problems, I mentioned it before. It just doesn't let me pass it. With the error message of vcore too high or maximum value reached. That just starting the test program. And what I mentioned above about vram that appeared 1.65-1.66 when my memos only go to 1.5. Even having fixed it in bios.
I think I'm going to download it again and see if it says the same thing. I'll try to take a screenshot.Thanks
-
Nothing either lets me pass it, the test has lasted 14 seconds
Subefotos.com - Share your photos quickly
Here's a bench with the error and the vcore screenHere's another one as I was saying the Vram value is set to 1,6XxX how can that be??? :S
Subefotos.com - Share your photos quicklyI haven't found a way to upload the photo any other way
Edited, I'm going crazy. In power saving mode, it runs without any problems. The Vram drops a little as does the Vcore, look at the temperatures :SSSS
http://subefotos.com/ver/?ff1ae115fb81e8c0dc9f886a408f2f8bo.png
-
The X5670 in Turbo (which I imagine will trigger when you turn on OCCT as it increases CPU usage to 100%) has a vCore of 1.35V according to the Intel specs. Try disabling Turbo Boost in the BIOS or manually increasing the vCore as it seems to cause severe vDroop as soon as CPU load starts (dropping from 1.2 to 1.07V) when in reality it should be the other way around (increasing the vCore when Turbo kicks in).
The problem is that increasing the vCore will increase the micro's consumption and if you doubt your power supply you may have 2 problems at once and be racking your brains over one of the 2.
Before increasing the vCore check if you have Load-Line Calibration enabled in BIOS, if not, enable it and you will avoid some of that voltage drop under load. -
I had the load line calibration set to auto, I changed it to enabled, and removed the turbo, and it still does the same thing. It won't let me pass the test.
On the other hand, when I put it in energy saving mode... it passes without any problems.According to you, what part of the source could it be? The board??
The Dram is still at 1.7... why? Shouldn't it be at 1.5??
Edit: now it does this.Subefotos.com - Share your photos quickly
For all that, it doesn't restart anymore. Honestly, I don't understand anything...
-
Hello again, after messing around with the bios, voltages and some peripherals.
I have managed to stabilize it.
I don't quite understand why, but from the bios, I couldn't quite fix the voltage and I downloaded the additional board software.
Specifically, I downloaded the turbo V and set the voltage to 1.35 Dram bus voltage that was at 1.6 stock, I lowered it to 1.5 and in turn Qpi/Dram core volt that appeared low at 1.35 I raised it to 1.5v.
With Ht deactivated and intel speed step too.
Of course if I lower the qpi/dram it blows up immediately, what does this value refer to? -
Hello guys, I'm reediting the post.
After blackouts, resets and a never-ending headache, my... swelled up and I decided to take it to a tech. Something I never thought I would need.
What initially seemed like a power supply.... (after testing the ram and other components) when I took it to the tech, they did several tests with other sources, another graphics card and another ram.... and in the end it turned out to be 2 capacitors on the board. 15€ for labor and capacitors. And it's working great
Now I can overclock without breaking a sweat.I say this for people who have similar problems and think it's the power supply, first check the capacitors. It seems that after 6 years, some voltage spikes, heat and other... factors had two that, even though they weren't bulging or leaking liquid, their charge wasn't stable and caused voltage spikes.
-
Wow… what a surprise! Motherboard failures are the most hateful because it's usually the only thing you don't have on hand to try a replacement without getting too confused. I'm glad you finally have it solved, go for it!
.jpg)
