Is it somehow possible my 650W PSU is not enough for a relatively lean system?

Dlaw said:

That doesn't apply to apartments, though. If you are renting your living space, it is the landlord's responsibility to maintain the property. That includes upgrades to the apartment/living space to keep it up to code as codes change. I believe that once a code changes, you have up to a year to conform without penalty.

If you own your own property, then yes, you can get away with things like wiring not conforming to the current codes.

Click to expand...

it applies to apartments where i live.
because even a 2 year old apartment here would no longer be up to code, all doors and windows would need to be replaced, and the entire complex rewired, to meet up with the new code, plus a number of other various things, and nothing around here has any upgrades.
i have renovated 30 year old apartments where everything was all original still.
seems crazy that they can force a building owner to keep a building up to date with new code down in america. buildings should only be held to the code from when they were built.
heck, most old apartments would need new staircases built, as they have changed the rules for tread size and step height in the last code update..

The laws in poland, make me want to open up an apartment complex there. Think about it, if you were to follow american regulations, and actually keep things current, and upgrade things........your apartment ratings would be stellar, and you would stand to make a tidy profit, as people would gladly pay for not having their stuff fried lol.


Not too keen on the being so close to russia bit tho, as they are going back to expansionist tendencies.

dyckah said:

it applies to apartments where i live.
because even a 2 year old apartment here would no longer be up to code, all doors and windows would need to be replaced, and the entire complex rewired, to meet up with the new code, plus a number of other various things, and nothing around here has any upgrades.
i have renovated 30 year old apartments where everything was all original still.
seems crazy that they can force a building owner to keep a building up to date with new code down in america. buildings should only be held to the code from when they were built.
heck, most old apartments would need new staircases built, as they have changed the rules for tread size and step height in the last code update..

Click to expand...

If you've ever been to a major city, you'd understand why. If someone plugs in too many things on a breaker (or even scarier, a fuse), and the wiring isn't up to code, it very well could start a fire. If you start a fire in a very urban area, there's no telling how much damage it's going to do. Once the damage is done, it's the landlord's insurance's responsibility to pay for all of it (not even including any lives lost). Where do you think the insurance company is going to pull that money from? Everyone else paying insurance, by raising the rates.

And if there's a lot of damage, then the National Guard may get involved. So now, the Government is spending money on damages caused by someone's apartment not being up to code. Where do you think that money is coming from? Taxes.

So you see, it's much better when landlords are required to upkeep their property.

Anyway, back to topic, 650w PSUs are enuff for lean systems, as long as they are really 650w

*hopefully avoids JackBauer PM*

Dlaw said:

If you've ever been to a major city, you'd understand why. If someone plugs in too many things on a breaker (or even scarier, a fuse), and the wiring isn't up to code, it very well could start a fire. If you start a fire in a very urban area, there's no telling how much damage it's going to do. Once the damage is done, it's the landlord's insurance's responsibility to pay for all of it (not even including any lives lost). Where do you think the insurance company is going to pull that money from? Everyone else paying insurance, by raising the rates.

And if there's a lot of damage, then the National Guard may get involved. So now, the Government is spending money on damages caused by someone's apartment not being up to code. Where do you think that money is coming from? Taxes.

So you see, it's much better when landlords are required to upkeep their property.

Click to expand...

i may not have been to a city as major as new york, but I live right outside the city of Vancouver, and have numerous friends who live there, and if you plug too many things into a breaker, even really old code required a breaker that should blow/trip before the wiring catches on fire. especially with old breakers, they get old and tired/worn out, and will blow/trip before they even hit their 15A rating. some old buildings we go into to do renovations on, our 14A air compressor will blow about 1/2 the breakers in the house before we find one with a short enough run of wire to the panel, and a breaker that is still good for 15A.
old buildings were built to code too, just not the current code, and thats the point of breakers, to trip before the wiring caught on fire, even back in the day.

Re: code & grid, I've just discovered that I don't actually have earth/ground in my power socket. How it has failed to occur to me for so many years living here is beyond me, but yeah. And the building is probably a two-wire one, with only perhaps some vague possibility of zeroing the earth if I'm lucky. (I'd be afraid to use the pipes here for earthing, as you can't know what's gonna happen, apart from the possibility that at some point the pipes are plastic.)

All is not lost though, back in communist Poland earthing was required in bathrooms, so I might go away with 20 m cable and some silliness.

The owner is family, so permission to get an electrician to come here and do the work would not be an issue. The problem is that real earthing could require a lot more work than just replacing the socket from what I've heard today, like pulling all of the cable out, connecting from the corridor outside the apartment etc. Ugh.

Still, do I want to buy a voltage regulator/stabilizer? (My $20 'anti-surge' power strip is suddenly looking so vulnerable.)

NZKshatriya said:

Anyway, back to topic, 650w PSUs are enuff for lean systems, as long as they are really 650w

*hopefully avoids JackBauer PM*

Click to expand...

How lean? There are some really good prices for used GF580 cards and sometimes also 670, 680, 770 and 780, even –90, though that's not that cheap. They tend to be cheaper than a 750ti, sometimes massively.

***

Regarding the mobo and BIOS, though, it seems I may have some luck yet — apparently, *one* of the *multiple* different causes of the same old P5Q-E problem can be the system trying to access a floppy drive that isn't there but is set by default as the first boot device. And that supposedly is enough to make the mobo/BIOS lose its wits.

There is some circumstantial evidence to confirm this: Actually, a couple of days ago, when things seemed to be working better after zeroing the CMOS for 10 minutes and activating the OV jumper, I noticed the POST message changed to something I had back when the BIOS didn't kill the CMOS on every boot. It seems as if the CMOS kill I did had restored the last working config... which would be something from 2009, apparently.

dyckah said:

i may not have been to a city as major as new york, but I live right outside the city of Vancouver, and have numerous friends who live there, and if you plug too many things into a breaker, even really old code required a breaker that should blow/trip before the wiring catches on fire. especially with old breakers, they get old and tired/worn out, and will blow/trip before they even hit their 15A rating. some old buildings we go into to do renovations on, our 14A air compressor will blow about 1/2 the breakers in the house before we find one with a short enough run of wire to the panel, and a breaker that is still good for 15A.
old buildings were built to code too, just not the current code, and thats the point of breakers, to trip before the wiring caught on fire, even back in the day.

Click to expand...

I'm really ignorant about electricity, so I've no way of knowing, but I'm beginning to wonder if the grid isn't somehow incapable of providing enough juice for the PC...

Here's an idea for grounding.

Get a steel/Iron bar thru center of building, wire all outlet/circuit grounds to bar, wire bar to a stake in the ground....There, prob fixed? And hey, bar could be used as lightning rod as well, double duty.

NZKshatriya said:

Here's an idea for grounding.

Get a steel/Iron bar thru center of building, wire all outlet/circuit grounds to bar, wire bar to a stake in the ground....There, prob fixed? And hey, bar could be used as lightning rod as well, double duty.

Click to expand...

11th floor in a block of 80 apartments. Can't really drive bars through the centre of the building.

attach to side of building, run cables through inner walls
*never argue with someone with too much time on their hands, and a vivid imagination*



- - - Updated - - -

Isn't it pathetic, that my classmates in my online courses plagiarize so often, and I get my assignments in early enough, that I have to compare my assignments to theirs, to make sure they haven't copied my essays?

*sigh*

They also blatantly copy MS Technet.

The CMOS seems to stay here ever since the following two things became true: overvolt jumper on, floppy not first boot device. Even after pulling the plug on the PSU, which by Asus's design triggers a wipe if something's bad with your settings. So now if it goes bad I'll know it's just whatever setting I've changed, so I'll take it in small steps, one setting at a time, and keep written records, starting from more voltage for mem and northbridge. And then I'm gonna dance with the Wolf. The other Wolf I had, an e7200, actually managed some really ridiculous OC in this computer back before I became unable to save any CMOS settings.

Okay, now I'm wondering about the PSU vs gfx.

I've recently bought a used Radeon 280X from Gigabyte, one of the WindForce3 editions, the Battlefield one (GV-R928XOC-3GD-GA). I've also replaced my mobo with the exact same model. Now I have no problems saving OC settings or any settings at all and booting, though the real OC headroom of this mobo is a different story and may well not be that big. But I have a problem with the gfx card:

I've specifically made sure that the extra power leads (6+8) are plugged in correctly. There is also no way the rated TDP of my system adds up to the rated 550W on the 12V line on my PSU. But the card seems to be very flimsy and fragile about going up from its idle 300 MHz GPU and 150 MHz mem clock. With various releases of AMD drivers that I've gone through, I've had problems like the card crashing faster if I set the fans higher or downclocked the GPU.

In some cases the card goes on a race to 100C out of the blue, for no discernible reason, e.g. me sitting here on OCForums, reading walls of text, no HDD or CPU activity, no nothing. GPU fans go roaring, I launch HWMonitor and what the heck? 70C! During some of those situations the PCB feels cool to touch (it never ever really gets warm apart from the metal strip on the back).

In Furmark, the card card goes up by about 1C to 5C per second at a stable rate, depending on the settings, but it always go up, the temp won't stabilize. And I'm getting 9 fps in Furmark in 1920x1080 (AA off) in those drivers that don't make the card crash, whereas I saw it lunch at like 60 fps and crash with older drivers. I sometimes see the card forget itself with the newer drivers and produce 20-30 fps at start, before dropping to 9. For comparison, my old HD 4850 (+ Accelero) (also from Gigabyte) could do a stable 5 without a fuss about the temps. Windows Experience index is like only 0.2 points above the HD 4850, and I think actually 0.1 points below my now-dead GF460 (also from Gigabyte). This makes me think that the card is throttling heavily, although it does show the full clock (whatever it's current set at) in Furmark. But I'm pretty sure that the fps I'm seeing in 3D Mark (06) is below my old GF 460. Heck, it's not clearly above the HD4850 and might be below it.

I'm aware that new cards don't necessarily mop the floor with DirectX 9 games, but this thing is a seriously beefed-up well-specced high-end construction with plenty of everything (384 bits and all IIRC), not one of those lean-specced nVidia cards like the GF960 that might be lacking in hardware for some purposes. I should be seeing some improvement over the GF460 and the HD4850 in DirectX 9 — unless the clock gets throttled, I guess.

I could also understand some measure of throttling from a modern card working in an older PC with an older CPU (yes, I know it's slower than recommender), but not this kind of thing here. Not HW Monitor showing the card with 84C at 500 MHz GPU clock, which is below half of its factory setting, and 35% GPU usage.

Obviously crashing 2 seconds into a DX9 game's intro/menu animation (a 3D map zooming and rotating) is something I can't understand at all.

What does this sound to you guys like? I can't help thinking it looks like not enough power.

Case air flow may be poor, but case wings are both removed. System and CPU temps aren't bad at all.

If the card feels cool when it's reading 70C, it's probably bad. You should open an RMA for that part too, if it's an option.

It sucks, but some times you get bad parts, even if they are "factory re-certified" (you didn't mention if this was a private sale or not).

Dlaw said:

If the card feels cool when it's reading 70C, it's probably bad. You should open an RMA for that part too, if it's an option.

It sucks, but some times you get bad parts, even if they are "factory re-certified" (you didn't mention if this was a private sale or not).

Click to expand...

Sorry, private sale from through an auction engine from a guy who claims it behaved 'normally in normal games'. I've given him two options: accept the return with postage on me, or the card goes to a repair shop for testing and the loser pays (if he loses, I get a refund along with all shipment fees) and am waiting for him to respond.

On the other hand, I can see that the heatsink is mounted on thick thermalpads that look bad (they actually look almost like coagulated glue, except for some details of shape), so I suspect applying some paste instead could help, along with replacing the WindForce fan shroud with my own 12cm fans. But the question remains how the heck he supposedly didn't observe any such problems when using that card since it was news. Told him to get things sorted out with the previous owner if it was a friend or something. I don't really have many options here, but the auction site has his address and record of the transaction (along with the category and description, which didn't mention defects), so if I don't get a reply from him any time soon, I'll just start the procedure, I guess.

And yes, the card feels cool despite huge reads, in some cases. It generally feels cooler than whatever the reads are. This is why I'm wondering if the GPU isn't heating disproportionately faster than the rest of the card. According to some reads changes happen very fast too, for example 5C a second (whether up or down). I remember such things happening back in the old times, but it still feels extreme on a modern card.

I'm tempted to remount the heatsink and see what happens, but then my position would probably be ruined with the seller and the auction site if I wanted a return.

Well, there is definitely something wrong with it, regardless if it's a simple fix or not. 70C is hot (to the touch), and you aren't going to heat something up that much without heating the things around it. That leads me to believe that it's the thermal sensor, which is something you can't really fix. Regardless of the actual issue though, since you just got the card, you should be able to return/replace it.