Zugg Software :: View Entry

Posted: Tue Apr 22, 2008 11:17 pm

That's definitely interesting. I know from my Physics background that cosmic-ray radiation is higher at altitude, but I thought it was only like a factor of 2 between Denver and sea level. But we *do* have a Radon mitigation system in the house because low levels of Radon in the soil of Colorado are higher than some other areas. Maybe it's a combination of factors.

And yes, we *do* have light dimming issues, so maybe there are some low voltage issues. Maybe I'll try to find a cheap voltage monitor to see what it's doing.

An update on the problem: It wasn't the video card after all. The new EVGA 8800GT did the same thing. The game I am playing is "The Witcher" and apparently it causes a lot of crash problems with various NVidia drivers. So now I get to mess with it until I find a stable driver set.

What's really weird is that once it crashes (sometimes a BSOD with the infamous N4_DISP.DLL error...definitely a driver problem, but sometimes with just a black screen), then after that the computer doesn't boot again for a while. If I wait like 30 minutes, then it will boot. Or, if I disconnect the main boot SATA disk, boot without it, allow it to complain that it can't find Windows, and then reconnect the drive, then it reboots normally.

So, I don't understand how a video driver crash issue could cause the computer not to boot for a while.

The new card *does* have a temperature monitor that I am using, and I have used RivaTuner to turn up the fan speed on the card. It comes set at 29% by default, which is too low, and as soon as I started Witcher, I noticed the GPU temp rising (past 72 deg C!!). When I set the fan to 60%, the temp dropped down to 64 deg-C, which is what it is normally just sitting at the Windows desktop.

I also found the 1.3 patch for Witcher and will see if that helps. I've got it running with the cursor moved to the edge of the screen so that the image constantly rotates just to stress the video card and I'll see if it crashes after a while or not.

The 1.2 version didn't crash on my MacBook Pro at all during the Austin trip, so I might also try to put the same NVidia driver version onto my gaming system (167.44 I think). NVidia releases a *ton* of bad driver updates, so finding a stable driver is always challenging.

Oh well, at least I have a fast graphics card that finally has HDCP support for an eventual Blu-ray drive. I was going to have to upgrade eventually anyway. But this is just taking up too much of my time!

Oh, also, another huge annoyance: The EVGA 8800 card didn't have any size specs on the box. When I went to install it in my gaming system, it was too long (by 2 inches!!). It was running into my central drive bay. Fortunately, I only had a floppy drive in the left bay (this is a desktop systems, so 3" drive bays are left and center, with 5" drive bay on the right). I removed the left drive bay with the floppy that I didn't need anymore and moved the center bay with my hard disks to the left. Now I don't have any drive bay at all in the center, so now the graphics card fits.

But you'd think they would mention this someplace! I read a lot of opinions on NewEgg.com and didn't see very many talking about how big this card was. In fact, I saw a lot *more* comments about the size of other cards, like the MSI cards. So I guess I should have measured before I went to the store.

But what does a "normal" user do about this? They go to Circuit City/Best Buy and buy a new video card. They look at the system reqs on the box and everything is fine. Then they get it home and it doesn't fit!!??? That's just crazy.

So yes, I have wasted a *lot* of time on this today :(

Posted: Wed Apr 23, 2008 12:23 am

Ugh, Nvidia drivers annoy the hell out of me.

Perhaps we should've mentioned that the cards are HUGE. There was a lot of talk about this at the time they came out (practically any review you find, the first thing they'll talk about is how massive it its). Fact of the matter is, though, normal users don't buy hardware - they get someone else to do it for them. The inside of computers scares normal users. If they were to do that, I guess they'd just take it back the next day.

On the topic of the Witcher...

Wizard Joined: 03 Mar 2001 Posts: 1127 Location: London

http://www.theinquirer.net/gb/inquirer/news/2008/04/01/nvidia-blame-vista-crashes

Posted: Wed Apr 23, 2008 4:09 am

I certainly wouldn't be surprised. There driver quality has definitely been poor lately. Trouble with ATI cards is that I hate the fact that they use .NET for their driver interface. Makes their control panel overbloated and slow. Driver control panels don't need to be fancy and "skinned"...they just need to do their job properly. Anyway, all of my other computers are using NVidia these days, so at least I can keep their drivers in sync.

The Witcher 1.3 update, along with the latest beta drivers (well, beta for 8xxx but certified for 9xxx cards...go figure) seem to have done the job. I'm actually enjoying The Witcher, although the Zero Punctuation guy is always good for a laugh. At least you Brits don't need to worry about the stupid censorship in the U.S. version. Seems that digital breasts is too much for people to handle here in the "land of the free". Oh well, that's a whole 'nother side track that we probably shouldn't go down.

Posted: Wed Apr 23, 2008 2:38 pm

Posted: Wed Apr 23, 2008 4:25 pm

Last night I finally figured out how the Rive Tuner application works. It is a *great* tool with one of the worst user interfaces I have ever seen! But I found a tutorial for creating a fan profile, and now I have my new card set up for quiet operation most of the time, with increasing fan speed as the GPU core gets hotter. It also puts a nice temp monitor in my system tray so I can easily keep track of it. Very nice.

I think my problems were partly driver issues, but I still think part of it was heat related. I noticed that the fan on the new card defaults to only 29%, which doesn't help at all. And it never seems to increase. So when doing some stress testing I got the core temp up above 72 deg C, which is pretty hot. Now that I have the automatic trigger profiles set up in RivaTuner, the temp never goes above 65. I know I mentioned this a bit above, but that was using the manual fan controls in RivaTuner. Now it's all automated.

I know that ATI has the ATI-Tool for doing similar things, but it still bothers me that the ATI control panels use .NET and that you need to use 3rd party control panels to avoid the bloated memory usage. I'd rather that ATI would just write a simple Win32 interface and concentrate on making it work well instead of trying to make it pretty.

But oh well, it's too late to do anything about it now. ATI and NVidia are always going back and forth. Both of them have driver issues (I notice that the date on the ATI driver is pretty recent too). I think it's kind of like the line at the grocery store...whichever line I get in will always be the slowest. Back when I switched from ATI to NVidia, I was really happy with the change. I'm sure if I switched to ATI I'd find problems there as well. Like I said, at least now all of my computers are using the same drivers so I don't need to worry about it as much (until the *next* game that does something the drivers don't like)

Posted: Sat Apr 26, 2008 1:16 am

Guess what?? Chiara's computer started making noise today. Sounds like the fan on the power supply was dying. Yeah, it never ends.

We bought her machine several years ago at Circuit City. It was a cheap eMachines system. This is the first time I have opened her system, and I have to admit that I'm impressed. It has a nice motherboard with 4 RAM slots (only 2 were in use for only 512MB total). It has 2 SATA ports that are not being used, and has a PCI-Express-16 graphics card slot that isn't being used. Lots of options for upgrading this.

So, it turns out that my old development machine still has a use after all. The development system that I replaced in Feb had a nice power supply in it. Easy to swap it into Chiara's system. Also, the development system had 2GB of RAM, which didn't fit into my new computer because of the different memory format being used these days. But they fit great into Chiara's computer! So I replaced her 512MB with 2GB!

Since we use her computer for our local LAN mail server, this has really given her a nice performance boost. And now it runs nice and quiet with the new power supply. This was an upgrade that actually went quick and easy like it's supposed to.

But it was a nice surprise to see how well an eMachines system was actually built. Now I know that I can put some SATA disks into her system in the future. It's an AMD 3200+ processor, so with a decent graphics card, it could actually be a useable gaming system too.

So, back luck that yet another computer part in this house was dying. But good luck that I already had some good replacement parts.

SubAdmin Joined: 18 Nov 2001 Posts: 5182

I am beginning to get the image of a dead machine closet. If you had enough time you could piece together 3 good computers from all the still working parts of the ones that have died recently.

Posted: Sat Apr 26, 2008 4:31 am

ROFL...Yeah, that's about right. Over the years I have accumulated a lot. Every now and then Chiara makes me go through the storage room and get rid of stuff. But I still have 4-5 computer cases with motherboards of varying age. I even still have the dead Windows NT server that we used to host at Superb.net that got infected by a virus and died. And I still have the original HP server machine that a very nice and generous zMUD customer donated to me about 10 years ago (also now dead). So there is definitely some history there.

Unfortunately, they keep changing everything in PCs. Memory formats change...I've got a lot of old RAM that can't be used anymore. Of course, RAM used to be smaller so it probably doesn't amount to much anymore. CPU sockets change, motherboard form-factors change. I'm still using most of the hard disks that I have that are larger than 100GB or so. Mostly SATA, but still a few IDE. Old graphics cards are mostly useless (AGP and even some PCI). Old networking cards are too slow. etc.

Yeah, I could probably put together about 3 full systems that would be so slow that nobody would want to use them ;) Or else they would also be flaky from unknown component problems (bad motherboards, etc).

I think that the computers that actually have "modern" motherboards that are still working are all still active in the house. It's really just my development system that died and had recent enough parts to salvage for Chiara's computer. My Toshiba laptop is fairly recent, but it's pretty hard to salvage much from a laptop. I just can't bear to throw it away either, even though it's pretty worthless with such a messed up screen/video. Not even sure I'd get much for it on eBay.

Posted: Mon Apr 28, 2008 12:47 am

OK, I really don't know what to do about my gaming system. That problem I had last week where it would just die to a black screen and then refuse to reboot for a while is back. So it wasn't the video card and wasn't the drivers. I normally leave this computer turned on all the time. When I came down to use it yesterday, it had rebooted at some point and was actually stuck at the boot screen. It rebooted fine and I was able to use it yesterday.

Today I came down to find it sitting at a black screen again. I rebooted and checked the various GPU, CPU temps. All looked fine. I checked the Event log for any errors...nothing at all. While I was sitting and thinking about whether to try a different driver, it immediately died back to the black screen again and refused to boot. So it was running only about 5 minutes before it died, and was sitting at the Windows desktop doing nothing at all when it happened.

I moved the computer to my work bench and it still wouldn't boot. Even disconnecting the hard disk like I did last week didn't help. After about ten minutes of trying to get it to boot, it finally booted. And now it's running just fine.

I just ran a 30 minute stress test of the CPU, memory, and GPU. I got the GPU temp up to 74degC with the fan at 100%. I got the CPU temp up to 68degC. Everything working fine.

I have no idea what to try next. I'm not getting an system errors in any log file. I'm not getting any crash dumps. And I still cannot understand what would cause it to refuse to boot for several minutes after this happens.

All I can think of now is that it's the motherboard, but I really hate to spend even more money on this without any real evidence. I still need to find some way to monitor the power in case maybe it's dropping too low to run the computer somehow, but that just doesn't seem to fit the pattern of this. It really seemed like a heat issue since it would start booting again after the case was open for a while, but I can't seem to force it to crash by raising the temp, so I don't think that's it either. I'm completely perplexed, and once again, I'm really pissed at the computers in this house.

How many times have I said how sick of all of this I am? I'm literally right on the edge right now. I can't take much more of this!

Posted: Mon Apr 28, 2008 3:17 am

OK, this time when it crashed I got a BSOD with the error "Page fault in nonpaged area".

So now I'm thinking maybe it's a memory problem? What do you think...memory or motherboard? Guess I'll buy some new memory and see if that helps.

By the time I'm done, I'll have replaced most of the components in *this* computer too. Sigh.

Posted: Mon Apr 28, 2008 9:51 am

Memory's cheaper.

I'm having similar troubles - my computer is still freezing occasionally. I've replaced practically every part in it now (using spares for most of it, thankfully) - just CPU and memory left now. I don't have spares of either, so I'll have to buy them. Memory's much cheaper than a new CPU, so I'll be trying that first.

Posted: Mon Apr 28, 2008 6:10 pm

Absolute nightmare! It sounds exactly like a problem I had, the computer would either just freeze or reboot, no entries in any logs or anything. I started replacing parts and still had issues. Eventually I had replaced every part except for the case, not including the PSU, so just a hunk of metal and the problem still occurred, at that point I gave up because I couldn't be bothered to spend any more time or money on it. I just chucked the lot into the loft and it's up there to this day!

Sorcerer Joined: 13 Jan 2006 Posts: 715

Zugg, trust me, it isn't just you. My friend Carol suffers from serious computer gremlins this year as well. And I am of the opinion that it IS because of declining quality. Computers built in 1975 were still being used in 1999 quite commonly among the hardcore hardware hacker community like myself who just love to tinker, and also with many major corporations and the government (remember all the Y2K craze because those machines were only programmed for 2 digit dating...) Those machines LASTED. In fact, they STILL last. I still have a working 386 and 486DX machines. Slow, old, but making GREAT cheap linux firewall/router machines for me since I had them anyways. But anyways, LATELY, newer PC equipment seems to only last 3-5 years. What happened to the commonality of systems lasting 10-15 years easily? Well, part of it I put down to the much much higher amount of heat generated by modern components, but I also am of the opinion that the quality is inferior. What was it the Russian guy said on Armageddon? "Russian components, American components... All made in Taiwan!". Then there was also from either that one or one of the other space movies about all that fuel strapped to a giant machine built by the lowest bidder. Crazy. And getting worse. So, keep chugging away. It will either get better or not, but we don't really have direct control over it these days. But it does reflect our screwed up economy.

On that note, a little off topic, but:
[rant]
When things were American made, sure they were more expensive sometimes than foreign made things, but they had higher quality.
Also, with things being American made, Americans had jobs, and could AFFORD things. Afford to live at least a comfortable life.
Now so many jobs are going overseas, and no one can afford anything anymore. And that IS getting worse. And we've done it to ourselves.
Buying cheaper foreign goods just costs American jobs, lowers quality, and makes the economy take a crap in the front lawn.

Oh, but wait, not many of us HAVE front lawns anymore because less young people each year can afford to own a home.
The American dream is gone, sold to the lowest bidder because everyone tried to save a buck here and there out of greed to have MORE.
Luxury instead of comfort. We've totally screwed up there big time.
[/rant]

Wizard Joined: 10 Oct 2000 Posts: 1551 Location: Australia

Not necessarily memory. I had a brand new motherboard which worked fine with the first 512 megs you gave it, but if you put in another stick of RAM it chucked a fit. I tested each stick of RAM in 2 different PC's using the free testing ISO image at www.memtest86.com and they were all fine, just the motherboard decided it couldn't deal with them. Page fault in a nonpaged area can be caused by flakey drivers too, I've experienced that at various times such as when the motherboard .inf drivers weren't compatible with the sound card drivers or the raid controller card didn't like using xp drivers in 2003, apparently even flakey antivirus software can cause it!

http://aumha.org/a/stop.htm has a list of the various blue screen errors and things to check/troubleshoot.

Posted: Mon Apr 28, 2008 11:08 pm

On the computer problem: Remember that this gaming system has been running very reliably for over a year. So this is a new problem. When the problem first happened last week I hadn't changed anything recently. No new drivers or anything. It was only after the problem originally happened that I started replacing video cards, drivers, etc. So the original problem was definitely hardware and not software.

The key to this problem is that when the computer crashes, it will not reboot again for several minutes. No POST screen, no nothing. I don't have a speaker so I can't listen for POST beeps. All of the fans spin and the hard-drive resets. But then nothing. Power-off/Power-on doesn't do anything. If I open the case, I can get it to reboot more quickly. To me that tells me it's a heat-related issue, and definitely hardware. I can't imagine how it would be possible for a driver problem to prevent the system from rebooting for several minutes.

Ralgith: I completely agree with your rant. Chiara and I talk about that all the time. The current economic crisis is just the beginning. Just a matter of time before the whole system implodes and collapses. The result of too many "short-term" solutions and not enough "long-term" thinking.

Posted: Thu May 01, 2008 11:21 pm

Cross your fingers...I just installed a new set of 2GB RAM into my gaming system. I had 4x512 and just replaced it with 2x1GB since memory is currently pretty cheap. I didn't go with the cheapest stuff though...I stayed with CORSAIR (Devastator series). They have tall heatsyncs on them, so maybe that will also help keep them from dying. Good thing I didn't have any space problems with it.

I also got a replacement motherboard in case I need it (and it's a direct replacement for the motherboard that I just put into my Development machine). Hopefully I'll never need it, but it was cheap enough to get one to have on hand, since the main problem in replacing motherboards is that they always completely change all of the hardware in new boards, requiring an OS reinstall. With this motherboard I can swap either computer without reinstalling Windows.

The UPS power conditioner is still "in transit". It was supposed to arrive today too, but the tracking system hasn't been updated and it didn't arrive with the other boxes. But at least I can install that at any time without opening the computer. My gaming system was the only computer in the house not connected to a UPS, so this is something I've been meaning to do for a while.

Hopefully the new memory will fix whatever heat issue I was having.

Wizard Joined: 10 Oct 2000 Posts: 1551 Location: Australia

This has nothing to do with bad computer luck, other than you totally need this as your new monitor for the games computer...

Posted: Sat May 03, 2008 11:41 pm

More travails from the land of nVidia:

I have a two-monitor setup - I was using one to play WoW and one to watch Dr Who (today's episode was ace, btw) and to browse thottbot. All of a sudden, WoW's FPS drops to about 0.5 and the video starts stuttering madly - when the game's fps goes up, everything's gone all weird - elements of the UI were all screwed up with random dots of colour everywhere and water wasn't rendering at all. Mplayerc crashed.

Given the fact that both programs using the overlay were screwed up, it's either a driver problem or a hardware problem. And I've never had any other problems with this card, and the temps were fine. So must be drivers. Grr.

Posted: Wed May 21, 2008 3:35 am

Came back from our trip to find the gaming machine crashed with a BSOD again. Something about a bad_poll_caller. Some Google research seems to point to something in the hardware. Some people found bad memory, but I've already completely replaced the memory. Some people talk about an overheated CPU, but my CPU temp is only 29-30 deg according to the BIOS monitor.

It's still very flaky. It reboots (unlike before when it would just refuse to reboot), so it seems like the problem has improved a bit with the memory replacement. But after booting, it crashed within 5 minutes in Windows even before I could run Firefox with a page fault error. Still sounds like memory, but I can't imagine that I'd get a bad Corsair replacement. Guess it's possible, but I think I'm going to try replacing the processor next.

I've still got the motherboard replacement as a last resort. But I think I'll try the processor first since that's the easiest to replace and won't effect the OS or other software. Seems like I'm going to have replaced everything in the computer before I'm done with this mess. Not a very nice thing to come home to.

Posted: Wed May 21, 2008 3:49 am

Hmm, before getting a new processor, maybe I'll play with the memory timings and voltage a bit. I have read that this particular motherboard doesn't auto-detect memory timings properly. The fact that the symptoms *have* changed since I replaced the memory (getting different BSOD and not having problems booting anymore) makes me think the memory has changed something, and maybe the motherboard timing or voltage is just off a bit, making the system unstable.

System just crashed again tonight, so it's definitely not doing very well.

Edited: I looked up the specs on the old memory and the new memory. The old memory was 1.9V and the new memory is 2.1V. So I increased the memory voltage by +0.2. The EasyTune still claims a voltage of 1.8V, which is screwy, but I don't trust the info from EasyTune. The BIOS doesn't actually show the absolute memory voltage anywhere I can find. But I'll see if it runs more reliably with this voltage increase or not.

Posted: Thu Jun 19, 2008 1:44 am

Annoying problems continued...

Since posting about these things actually seems to help, it's time to hear about my latest computer problems.

Btw, on the problem in the previous post, changing the memory voltage *did* fix the problems. I also replaced the processor just in case, but it hasn't crashed at all in over a month now.

But today's problems are completely new. I wanted to try Eve Online again, so I downloaded their client software. It's a 1.2GB download, which takes about 35 minutes on my high-speed cable modem. After downloading, I clicked and got the error from their NSIS installer:

Posted: Thu Jun 19, 2008 6:51 am

Well, copying the entire EVE folder from the working computer to the main gaming computer kind of worked. I was able to launch the game, but I ended up crashing when I entered a station.

I burned the DVD on the laptop and I am able to install using that file without any trouble. So somehow the large files are getting corrupted when transferred on the network, either downloaded with Firefox or IE, or even just when copied across the local network from another computer. That has me a bit worried. I'm going to install from the DVD image and then see if I still crash in the station or not.

Edited: No longer crashes after installing from DVD. So there was definitely some corruption from the copy, probably in one of the big texture files that was copied.

Anyone have any idea why large files transferred across the network might get corrupted like this? I haven't seen anything like this before. I'm using the integrated Gigabit network ports on the motherboard (Nvidia chipset and drivers). It has dual network ports, but I've got one disabled because I used to have network errors when both ports were connected (so much for improving network speed with multiple ports!).