• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Power Supply problem?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

ZXIII

Registered
Joined
Jun 14, 2003
I'm kind of new to ocing so I may not know what I'm talking about, but it seems like my PSU is to its limit and can't handle anymore o/c, I'm not sure though... it's a PP-303X, 300W (ATX). My system specs are

AthlonXP1700+ Tbred B DUT3C, 0311MPMW @ 2083 (166x12.5 - 1.675v)
A7V333 Rev 2.0
SLK-700 w/ YS-TMD fan + AS3 thermal compound
MSI GeForce4 Ti4400 128mb
2x 256mb Corsair XMS 2700C2 @ 2-3-3-5 - 2.8v
SB Live X-Gamer
DFE-530TX ethernet card
2x WD 60gb 7200rpm hard drives
Philips CD-RW 8x4x32x

I can't seem to even boot up at above 170 fsb (lol) no matter what I do (raised ram voltage, put timings to 2.5-3-3-6, raised cpu voltage) and I doubt it's my temps (cpu 34C idle, 39C load, MB temp 34-35) or my ram being pushed too far.
My voltages are 11.936 | 4.838 | 3.264 | 1.696

I run 3dmark2k1 for a few hours and they don't seem to fluctuate much, like + - 0.05 at the most but I don't even know if these above values are good or bad.. I don't have enough experience to tell if that indicates my psu being overloaded or what. Most people tell me my psu is way too low for what I'm running, but I'm having no crashes or reboots at my current settings, and the air coming out from the back of it isn't even hot (slightly warm). I had to raise my voltage from the stock 1.6 to 1.675 for stability over 1.9 ghz or so, otherwise I'd get blue screens while booting WinXP (I tried 1.650 first but that didn't work either).

So I'm fine now @166, but if I try to go a little higher (like 175 fsb), the system suddenly refuses to even boot. It boots and loads winxp fine with 170, but I haven't done extensive testing at that fsb. I know not being able to even do 175 seems pretty pathetic, so like does anyone have any clue what could be doing it, if not psu? I've seen most people go to 180+ (up to 192-195) with my board, and about that much with my type of ram too (some even with pc2100 corsair). I'm aiming for like 185x12/12.5 for long term usage and maybe slightly higher (if my board can) for benchmarking. I just don't see how I'm stable at one point and then I raise fsb slightly, it can't even POST all of a sudden (I assume it would start crashing/rebooting first, before getting to the point of not post'ing). Is this a common psu issue or something? Oh, and I haven't tried doing memtest or prime95 yet, but I hear some people that are perfectly stable, only get errors/crash when running prime95 sometimes so I don't know if I should trust it too much.

I'm probably going to get a fortron/sparkle 350W (would be enough?) cause I read those are like really good price/performance value, if this is for sure a PSU issue... but if anyone has other suggestions to try out, that wouldn't hurt either.
 
I used a PP-303X for several years, and it was overwhelmed with the load presented by my first P4 setup, a 1.6a at 2.4GHz. I replaced it with a 350 sparkle, and have just now hit its limits with my current P4 1.8a at 3.2GHz.

The 350 Fortron/Sparkle is a great deal more powerful than the PP-303X. I suspect it would be enough output for your application, considering you have any luck at all with the PP-303X and the fact that I know the 350 Fortron to be a lot stronger.

Whether the power supply is responsible for your lack of overclock, that is hard to say. There are so many possible causes for a problem like this that it is hard to say without testing the rig. You pop in another supply, and you have the real answer very quickly. It is dangerously unreliable to try to guess the results from afar.
 
ok I remembered I had an old compaq that still has its mb and psu so I took the psu and hooked it up to my hard drives, also unplugged my cd-rw and did more testing, then I plugged in the hd's and my cd-rw back, did the same testing and same results... and I figured out some stuff.

One of my problems was the 12.5 multiplier setting, it wouldn't let me boot up at all @ 175+ fsb .... but at 12 I could boot into windows at up to 185 fsb... I know the multiplier itself can't be the problem but for some reason I can run 185x12 (2.22) but not 175x12.5 (2187)... it isn't stable though. It freezes in 3dmark... even at 1.8v (max setting for my mobo), and at 39C idle, 43C load temps. This happens regardless of my hd's being connected to the other psu or to my original psu. It's stable at 180x12, but when I turn my system off and then back on, it wouldn't post....have to turn it off/on again and change 5 fsb lower in bios (when it starts in safe mode). When I reboot (at 180) it posts and loads windows... just wouldn't when I turn it off then on... don't know what could be causing this. At 175x12 it posts and runs stable in both instances (when I reboot and when I turn it off/on).

I also noticed that at higher voltages (above 1.70 or so), my monitor blacks out for about 1 second when the PPPoE service loads to connect to the net, on startup.

I changed to stock speed (133x11@ 1.6v) and my voltages are
11.872 | 4.892 | 3.28 | 1.632
with the vcore voltage fluctuating the most, to 1.616... the rest are almost perfectly stable. At 2+ ghz speeds both the vcore and 12v voltages are constantly changing about 0.02 - 0.05v lower.

Would a better power supply keep my system more stable at the same (maybe lower) voltages?....this might be a dumb question but I heard that higher quality psu's give better current and would allow for a higher o/c with the same voltage, than a lower end one...? What about more stable voltages, that should help right?
 
Yes, a good power supply generally helps stability and overclock results, but you are only going to see a difference if your present one is innadequate. To be honest your problems don't sound to be power related, but it's really impossible to tell for certain. Once you try a different supply you will know how much of a difference it makes for your exact circumstance. I'm sorry I can't give you a definitive answer, but it's just guessing to try to blame or absolve your current supply based on the symptoms you describe.
 
mm I think I'm gonna buy another one and try it, I have a feeling it will help ... even though it's hard to tell. I highly doubt any of my other parts aren't letting me go higher.... I did more testing and I can run stable at 190x11, but for now I'm at 185x11.5 (1.7v) as long term use cause I fear my hd might die at 190+ fsb.....I'd need a motherboard with agp/pci lock to safely go higher. I forgot to say last time, that at 1.8v and 2.2+ ghz is the only time I freeze in 3dmark....every other setting below those, if the voltage is too low, 3dmark crashes to desktop or gets a blue screen... I raise the voltage and it's stable... but if I raise it to 1.8v it starts freezing.

The only other reason I could think of is if my chip can't go any higher, but that's highly unlikely unless I have very bad luck and got a low quality chip or something. It should do 2.4+ without difficulty, and I know it's not guaranteed but I'm almost sure I can push it farther... it would also be cheapest for me to buy another power supply than anything else, so I should start with that. I think I have the highest chance of finding out what exactly my problem is that way and if a new psu doesn't help it's still a plus to have a better one for future upgrades and stuff... and I'd know that I'm in the 1% or so with a very low-oc'able tbred B chip (I'm hoping not).
 
Back