500W not quite cutting it?

mixxedstuff

Junior Member
Jun 21, 2008
12
0
0
Apologies if this question has shown up recently- a quick glance with search didn't quite give me what I was looking for. Google also has lots of outdated stuff available.

Regardless this'll be a bit long.

I'm currently running on the following system:

Antec Earthwatts 500W (Came with Antec Sonata III)
ASUS M3A-H/HDMI Motherboard
AMD 9750 (125W)
Visiontek ATI 4850
4GB (2x2GB) @2.0v, 5-5-5-12
Razer Barracuda sound card
Encore wireless-N PCI card
160GB SATA HDD (planning on adding a 500GB in the future, but power may be another concern)
3 120mm fans running (includes CPU fan)
DVD Burner
Vista Home Premium x64

I just went into bios and it gave me the following readings:
Vcore- 1.2680 V (that one should be fine)
3.3v- 3.264 V
5v- 4.972 V
12v- 11.968 V

Now here are the issues I've been noticing:

Disregard the italicized- just got word from Visiontek that this section is normal behaviour.

First, when the system powers up with a cold boot, it often (though not consistently) does some sort of power cycles, where the fans power up, then down, up then down, while a red light on the back of the 4850 does the same in sync with the fans. This could sometimes (rarely) occur for up to somewhere around five to six seconds, which is where it really caught my attention. Usually it only cycles around three times. Since the system continued to boot/run, I figured it was okay, but it did concern me some. I shot a support email to Visiontek, in case it was an issue with the card, but didn't get any reply. Might be unrelated. Continuing on...


In addition, I found that hooking up a 4th 120mm fan, the system has issues booting, and asks for bios reset ("Overclocking failed!" thanks to the 2.0v and timings on the memory), or gets into windows, and locks up in short order. Disconnecting the fan seemed to remedy the problem so that I could run windows again.

Just an hour or so ago, I was playing Crysis, and during a rather active battle on the cruiser deck, I noticed that the system was slowing down extremely. I went to save (barely managed), and realized the entire system was slowing down. I (very slowly, progressively slower) navigated to the start menu and attempted to open up ASUS Probe to check temps, but it never managed to load- the system froze up again, akin to how it did with the fan attached. Touching the heatsinks, however, implied they were fine (GPU is manual set to 65, not auto), since they were quite cool- I've had them hotter stress testing. This also happened while the case was open (forgot to close it after disconnecting that fan)

EDIT: Gave Crysis a shot again. Managed to escape the complete lockup a few times by hitting esc at the first sign of slowdown, but eventually got a "Crysis has stopped working" error from windows.

I understand that there is a wide range of things that can be causing the issue, but the fan causing failure is what raised the red flag. I would have guessed that overloading the PSU would cause the entire system not to even boot (as one of my previous systems did a few years back when I attached both a cold cathode ray tube and some LED fans)- is this another possible symptom of too weak of a PSU? Searching "computer slows down" and "computer freezing" in combination with any keywords generally returns extremely generic things with Google =P

I'd like to have a little feedback before I go blowing 100 bucks or so (more than about 125 is out of the question for me at the moment). I am considering one of those supplementary power supplies to offload from the Antec- do you feel this is a reasonable choice? Or would I be better off going for some 650-750 watter within my price range (up to $125, hopefully)? Do you all suspect it's something else causing the problems? I'm sort of shooting in the dark here- it's been a long time since I've dealt with this stuff.

Thanks!
 

HOOfan 1

Platinum Member
Sep 2, 2007
2,337
15
81
Software readings of voltages cannot be trusted, just so you know. Besides all of those readings were well within specification anyway.
 

bob8701

Junior Member
Jan 21, 2008
15
0
0
the voltage reading is in spec, but pc use most dc power when play game, so it is difficult to say psu is fault, did you even run Prime 95 to check your pc stability?
cpu overheat can casue similar problem as yours.
 

mixxedstuff

Junior Member
Jun 21, 2008
12
0
0
I had run prime 95 for quite some time when I first put the PC together. But just to be sure, I'll run it again for a few hours.
 

BassBomb

Diamond Member
Nov 25, 2005
8,390
1
81
Originally posted by: mixxedstuff
I had run prime 95 for quite some time when I first put the PC together. But just to be sure, I'll run it again for a few hours.

check your gpu temps and cpu temps also to be sure
 

mixxedstuff

Junior Member
Jun 21, 2008
12
0
0
Just got word from Visiontek that the video card power cycling is normal behaviour, just a check to see if enough power is being supplied.

The comp's currently running Prime 95- coming on 5 hours now, 100% on all cores, 45C for CPU temp. One instance is running blend mode, the rest are small FFTs (since blend uses so much memory). Tests seem to be passing. This is cooler than I got with the stock cooler (which was usually around 55 load). As a side note, I've been running the CPU at stock speeds.

GPU tends to be at 56C or so idle, with fan at 65%, and stabilizes around 71C when at 100%, when using the ATI Catalyst automatic overclocker thing (although games don't seem to like that overclock- so I brought it back down to stock, which does me fine anyway).

Anyone have any ideas about that fourth fan causing problems with POSTing and, if it gets to windows, a freeze?