Amiga.org
Amiga computer related discussion => General chat about Amiga topics => Topic started by: Hodgkinson on May 15, 2008, 09:51:43 PM
-
Hi all,
I've just bought a new graphics card (Ok, its an AGP board for my PC, but Amiga's can also use graphics cards?) and every now and then I get what appears to be a line of noise across the screen, the same kinda thing as if you have a analogue TV and a noisy thermostat kicks in somewhere in the house - I've only ever noticed it at the same point about 1/4 the way down the screen. The old graphics card never did this, and the PSU (550W) should be more than enough to run the system.
The hardware GPU core temperature monitor never exceeds ~40C, and that’s with an additional "bay cooler" fan positioned beneath the heatsink on the card (No fan when supplied, but I thought it might be worthwhile to do anyway).
I’m using the latest drivers for the card, and the BIOS settings are set as close as I can get them to match the cards specification (It’s a fairly old system, so I can only get close to the popper settings)
Could this be a sign of an error or hardware fault within the Gfx cards GPU?
Advice appreciated,
Hodgkinson.
@Moderators - Please feel free to move this thread if necessary - I felt that if it is a hardware issue that it may be of benefit to any computer system; PC or Amiga (Besides, the warranty on the card won’t last that long).
-
Assuming everything else is OK, faulty RAM on the card can cause problems like you describe. These things can be a little difficult to pin down, but under-clocking the GPU and/or RAM is one way to check.
-
Underclock? I can't say that I've seen settings for that anywhere yet. Ideas? The board is this one (And from this company):
http://www.novatech.co.uk/novatech/specpage.html?NOV-GF6225
It should be capable of AGP 8x, but the BIOS AGP speed only goes up to 4x and the driver is reporting the card running at only 2x.
The problem certainly is hard to pin down, even to just notice. I've only seen it when pottering around in XP's desktop, but Im guessing* that it might still be occurring in games etc. Mind you, it's hard to notice when you're trying to land a 747 at Heathrow :-D
I presume that this is probably a hardware fault, though? If it is, im getting a replacement (Shame, it's a nice little card).
-
Sounds like the memory might be overheating as I've had similar display corruption issues when overclocking the memory on nVidia gfx cards.
It looks like it's passively cooled so one thing you could do is screw a little fan onto the heatsink. Or just make sure that you've got plenty of air flowing through the case.
Either way I'd RMA it and see if another one is any better. If not then try one of my suggestions.
Andy
-
There's plenty of airflow in the case...
-PSU + Rear case fan + Bay fan blow out at the back
-Front case fan + 2 small fans in HDD caddy blow in at the front.
I don't remember seeing any IC's on the opposite side of the board, so I assume that the RAM is somewhere under the heatsink, even if its bonded to the heatsink or not there's the additional bay fan (Can't think of a better way to describe it) right next to the Gfx card, sucking air over and past the heatsink area. The GPU core is struggling to reach 40C, so im guessing maybe a manufacturing fault somewhere?
EDIT: Ah, Rkauer had the same idea re the "slot" fan with this (http://www.amiga.org/gallery/index.php?n=541=11) A3000T.
-
ATITool (http://www.techpowerup.com/atitool/) should work fine. (I'm not too familiar with NVidia cards, if someone wants to suggest a better tool.)
-
If you connect it to your monitor via the VGA connector,
it might be worth checking if it does the same when you
connect it via DVI.
Edit: Or try another (better shielded) vga cable.
Regards,
Lloyd
-
Ah, it’s DVI->Adaptor->VGA Monitor at the moment. I'll give the other connector a test run. Oh, a trial showed that the outputs are separate to one another, e.g., extended desktop can be implemented with this card... (/Me wonders if this is standard these days...)
Thanks,
Hodgkinson.
-
Right then.
Just the other day I saw the GFX problem again, although this time accompanied by audible clicks from the speakers.
Hence, I’m wondering if the Q-Tec 550W PSU is suffering (CD-Reader, DVD Burner, 2x 80Gb HDD's, 7 fans total (!), PCI modem, PCI sound, PCI USB2.0, 2x 3.5" floppy, and the below Gfx card).
I've got a £40 Amazon voucher to use up. I'd like an extra 100W, and there is some "Atrix" 650W/720W PSU's that look interesting.
Now, we’ve always bought cheap PSU’s (~£20) over the years, and as a result we don’t really know what brand would be considered for decent, reliable PSUs.
Suggestions, folks? :-D
-
Sounds unlikely.
I've got a DVDRW, 6xHD, Geforce gfx card, 1394b card, sound card, 2x Gig E cards, TV card, SATA RAID card, old power hungry Pentium IV Northwood and I only have a £20 350W PSU.
It is a common mistake to evaluate different PSUs by comparing their maximum total power output. This is not a very accurate way of comparison. It’s a bit like comparing different cars by how high their speedometer has been marked. For a modern system the 12V will be the most important load. PSUs are labelled with max power output for the different lines. Comparison of ATX 2.01/2.2 PSUs can be a bit more complicated. The problem is to find out the maximum 12V output when having more than one 12V line. Unfortunately it is not always correct to just add up the maximum output of the different 12V lines.
Invest in a £10 antisurge extension cord? Add a ferrite ring to the Kettle lead going into the PSU?
Names to look for? Enermax, ThermalTake, Qtec, Nexus, Seasonic, Corsair.
-
The PSU has more than enough juice (but in theory it still can be the PSU if it's failing). I ran ASUS P5B-E, E6320 @2.8GHz, 2GB DDR2-800, DVD-RW, HDD and 8800GT with noname 350W and it was just fine.
Remember that even the most efficient PSUs have only around 80% efficiency, meaning that with 550W PSU you will be losing minimum 110W, even when totally idle.
With that 650W/720W it would be 130W/144W lossage... Ouch.
-
True, efficiency is an issue, but I assumed that the efficiency figure was based on the power being drawn from the PSU as opposed to that which the PSU is rated at?
Qtec as a good brand? I wouldn't of guessed judging by their prices alone...(The one in there at the minute is a Qtec)
One other one that im looking at is a Ezcool 650W...
-
Would it be worth while downrating and going for, say, a Thermaltake 420W?
EDIT: How about this one? (http://www.amazon.co.uk/Thermaltake-W0093-PSU-500W-TR2-500W/dp/B000UH0544/ref=sr_1_2?ie=UTF8&s=electronics&qid=1214142316&sr=1-2)
On a similar note, would anyone trust this PSU? (New seller in particular)
http://www.amazon.co.uk/Pimp-My-Player-1124-Micro/dp/B001B1X3Q0/ref=sr_1_41?ie=UTF8&s=electronics&qid=1214133696&sr=1-41
(Eg, say to run a 2GB SCSI HDD, SCSI DAT Tape and SCSI HP CD Burner?)
-
The 500W (Actually 450W rebadged) ThermalTake PSU arrived the other day and I've swapped out the old one. Surprisingly, the old PSU looked in good shape inside. I was under the impression it was a Q-Tec, but it was actually a ColoursIT PSU.
Let's hope this new PSU fixes the problem.
PS. I’m kinka pleased how well the wiring has tidied up inside. Might upload one or two photos (On my own site) ;-)
EDIT (Large-ish images):
http://www.booni.info/downloads/PC_Wiring.jpg
http://www.booni.info/downloads/PC_Wiring_Close.jpg
-
Just beacuse the Wattage is adequate does not mean that the power supply is keeping steady volts or amps on the rails.
From past experience of a lot os psu's, you defo get what you pay for.
Cheap £20 - £40 650w psus generally are not up to the task.
The PSU is the only part of a pc i would defo suggest that you spend decent money on, you will not regret it
-
Piru wrote:
Remember that even the most efficient PSUs have only around 80% efficiency, meaning that with 550W PSU you will be losing minimum 110W, even when totally idle.
With that 650W/720W it would be 130W/144W lossage... Ouch.
Eh? These are switching power supplies. The wattage ratings are maximum-output-before-fire, but the actual power varies based on load. A given design has a sweet spot for load (and indeed a minimum load) along with its own operating losses, but taking 80% of the maximum rated load won't give you a realistic idle figure except by dumb luck.
SilentPCReview (http://www.silentpcreview.com/article28-page4.html) has some real-world numbers. The "output" is the actual current demanded by the idle components, which, with spinning drives, fans, and dozens of hot'n'hungry transistors, adds up. The "input" is what the PSU sucks from the wall to feed that demand.
Those are some trendy, expensive, probably 80%+-certified PSUs, but you can see that the PSU is only eating about 20W (~70% efficiency, per the curve) to support the system at idle. That's a couple (incandescent!) light bulbs off from your estimate.
...
Anyhow, if someone wants to blame the PSU, my FSP (Fortron Source) "Green" units are still going strong, probably still one of the cheaper options out there. YMMV.
-
I'm not too sure on this but....
If your mobo only does AGP x4 then you can only install an AGP x4 or x2 card (which I think run on 5V).
An AGP x8 card runs on 3.3V (or something) and (afaik) shouldn't be installed in an AGP x4 or x2 slot.
Your card is possibly complaining at being overvolted and you're probably very lucky it's still working at all.
-
In relation to the power questions -- the symptoms are a line on the monitor and a crackle in the speakers, two items that are also separately powered. Speakers (and potentially display cables) can pick up noise from anywhere, but if this is really transient, is it possible that the monitor might just have a favorite spot to flicker when a nearby A/C compressor or something clicks on?
There can also be a lot of subtle timing issues with video that aren't so much defects as SNAFUs depending on different hardware's tolerances. Is the monitor in question a CRT or LCD? Does it get worse if you run the digital DVI output to a DVI-supporting monitor, or does it disappear?
-
DrDekker wrote:
I'm not too sure on this but....
The Internet can help. (http://www.ertyu.org/steven_nikkel/agpcompatibility.html) Really, it can. (http://episteme.arstechnica.com/eve/ubb.x?a=tpc&s=50009562&f=67909965&m=750004173631) (If the card weren't compatible with the slot, there would probably be much bigger problems than the occasional line.)
-
Thanks for the replies.
If it were solely interference related, I'd of expected to see it before I started beefing-up the system, whereas im my case it only started appearing afterwards, which is why im suspecting the PSU so much.
The screen is a Sony Trinitron MultiScan E500 CRT, and with using either the DVI output (Via adaptor to VGA) or the straight VGA output yields the same problem (Granted, I haven't actually tried DVI yet, but I'd have to yoink a LCD off another system for a good couple of days to test it).
Re voltages, the card has both the 1.5v and 3.3v keys, so I assume it should work quite happily (I read that the mobo is AGP 4x max from the manual, and hear that its only 2x max from the previous owner).
EDIT: I’m surprised that no-one's commented on the idle HDD in the bottom of the tower in the photos :crazy:
-
Hodgkinson wrote:
Thanks for the replies.
If it were solely interference related, I'd of expected to see it before I started beefing-up the system, whereas im my case it only started appearing afterwards, which is why im suspecting the PSU so much.
Beefing up = new graphics card? I'd be inclined to agree, really, unless this also coincided with the start of the air conditioning season. I'm also from NTSC-land, so maybe my idea of a line of static is slightly different...
The screen is a Sony Trinitron MultiScan E500 CRT, and with using either the DVI output (Via adaptor to VGA) or the straight VGA output yields the same problem (Granted, I haven't actually tried DVI yet, but I'd have to yoink a LCD off another system for a good couple of days to test it).
A CRT rules out any LCD analog-input quirks. It could still be susceptible to power transients, I guess.
Is the 'line' a simple black or white line/near-imperceptible 'shudder' of the display? Or is it clearly multicolored (or noticably colored) garbage?
If it's colorful and really looks like garbage of some form or another, I'd immediately suspect the graphics memory/access to it. If it's more subtle, well.. monitor issues are ruled out by trying another monitor.
Re voltages, the card has both the 1.5v and 3.3v keys, so I assume it should work quite happily (I read that the mobo is AGP 4x max from the manual, and hear that its only 2x max from the previous owner).
EDIT: I’m surprised that no-one's commented on the idle HDD in the bottom of the tower in the photos :crazy:
I think I missed any link to the photos. There's not one actually showing the symptom, is there?
If it really seems to be the video memory... I'm loathe to overvolt, but I have to admit, I recently went through some horror with system memory on a DDR2 machine where the regulation was biased just enough for vDIMM to be too weak for certain portions of a DIMM.
That was hell to figure out, and video cards should have their own regulators on-board that aren't meant to be user-tweakable. But if any of your voltages read lower than they should (4.999 vs 5, 3.299 vs 3.3) and your BIOS has tunable regulation, you could try giving them the slightest bump. Or you could RMA it and hope you get one that works within the tolerances of your system. (Or it could quite easily be hosed in a way that power can't help...)
It might help to look for video memory testing utilities. I have no recommendations and I'm not sure if any of them produce truly useful output (the ones I'm finding on Google seem to bill themselves as 'stress tests'), but they do seem to exist.
-
Here's the links (Large-ish images):
http://www.booni.info/downloads/PC_Wiring.jpg
http://www.booni.info/downloads/PC_Wiring_Close.jpg
Nope, there not of the symptom - Just of the new PSU installed and how the fans are rigged up (The only way to capture the problem would be to have a camera running for a good couple of hours on a tripod).
By "Beefing-Up" I mean new GFX card, new HDD, extra fans (Numerous, depending on where you look...), internal modem, and swapping an old CD-Writer for a DVD-Writer. Quite a bit of stuff, actually.
It hasn't been that hot here this summer (And neither as wet as last year, god forbid). The air-con machine we have hasn't been used these last couple of years, mainly 'cause it uses that much power. I doubt many other people would be using one at the minute either, and the PC/Monitor/Stuff runs via a surge-protected strip anyway.
The line flashes for an instant on the screen, and appears to consist of what could be described as white noise. I only used to notice it at the same part of the screen, but the other day I saw it further down.
Heh, I checked the rail voltages of the new PSU in the BIOS and with a multimeter. The BIOS readouts are utterly useless. I mean, the +12v was 12.05v on a multimeter, and the BIOS said that it was more than half a volt out! Sometimes, you can see the value jumping between one and another, and the last couple of digits are completely meaningless due to the magnitude of the quantisation error. Not only that, but I went out and bought a new fan just so that I could make use of the case fan RPM sense function of the motherboard. Turns out that the fan runs too slow for the bios to register (As you might be able to read down the edge of the clear fan in the photos). That also took some figuring out and a lot a messing around with spare CPU coolers and oscilloscopes.
If the new power supply doesn't fix the problem, I'll be sure to have a look around for a test utility. I actually asked the manufacturers if they had one and they never answered that part of the question. Sigh. :roll:
-
@ Floid - ok I screwed up on the voltages - my memory must be as bad as Hodg's video card. :lol:
@Hodg' from the photos it looks like you've got a pretty old (and basic) cheapo motherboard.
Try updating the BIOS to see if that helps.
Failing that - it could just be a hardware compatibility issue. If so, I'd seriously consider getting a newer motherboard - at least one that has AGP x8.
By the way - what card did you have originally installed?
-
You see those big clumps of caps near the processor? If I remember rightly the reason why we were given the board was because the original caps were leaking. That's our repair job (Including the rather more concerning point that all the caps that we used there are not low-ESR - The only ones we had - Whereas the originals were low-ESR).
My dad recently bought a new dual-core mobo and tried to get 32-Bit XP to work on the 64-Bit processor (Using something called a HAL Refresh). Now all he's got left of the system is a ghost image from a few months back. I ain't trying that trick again :crazy:
On a similar note, I’m afraid I’m similarly inclined with regards to BIOS updates - From what we've heard there a last-minute resort if nothing else works, and are prone to ruining systems (Well, the last firmware update - As included on the CD and instructed in the manual - Made my MP4 player revert to Spanish menu's every time it was turned on. Fortunately I could take that one back for a refund)
BTW, I think the original GFX card was a GeForce MX440 64MB.
Heh, on the subject of stuff not working, whenever you use the onboard USB, the machine never powers-off the PSU. Windows shuts down and there's a click, but the PSU stays on - Every single time. Don’t use the on-board USB, and it powers-off fine (Guessing it’s also a BIOS issue…).
On the other hand, the USB2.0 card (A £6 board + front panel package) in there seems to be prone to errors. Normal documents it handles fine, but if you transfer a video, occasionally half of the odd frame might turn white or you get clicks in the audio track.
-
Hodgkinson wrote:
You see those big clumps of caps near the processor? If I remember rightly the reason why we were given the board was because the original caps were leaking. That's our repair job (Including the rather more concerning point that all the caps that we used there are not low-ESR - The only ones we had - Whereas the originals were low-ESR).
Hmm, the pictures were sort of informative, maybe not in the way you were hoping.
That's a Via KT133A-based board; I had one too not too long ago (no real problems with the Matrox G200 installed in it, heh). It probably has the 686B southbridge with the obscure issue that can potentially cause major headaches with PCI peripherals. That might be the issue with your USB card.
Or, depending on the chipset of the USB card, you could be running into a hardware quirk with the USB chipset... by awkward coincidence, Via is the manufacturer with a quirk that I'd expect to find on an add-on card, NECs seem to be the baseline for 'known good.' [I'd also assume the vendor drivers for Windows address or avoid the quirks, since the issues I'm thinking of appear to have been 'discovered' with *NIX... so that'd be more reason to question the PCI business first.]
On a similar note, I’m afraid I’m similarly inclined with regards to BIOS updates - From what we've heard there a last-minute resort if nothing else works, and are prone to ruining systems (Well, the last firmware update - As included on the CD and instructed in the manual - Made my MP4 player revert to Spanish menu's every time it was turned on. Fortunately I could take that one back for a refund)
Dodgy MP3 players aside, most BIOS updates just include patches pushed by (and tested by) the chipset vendor, CPUIDs for compatibility with newer processors, and fixes for any braindeadness in the configuration utility.
That board is worth about US$1 now, the final BIOS release probably happened 3 years ago, and some of the 686B's quirks could be mitigated in device configuration registers that get set by the BIOS. You might as well apply that last update and see if anything (particularly data integrity re: the USB card) magically improves.
Heh, on the subject of stuff not working, whenever you use the onboard USB, the machine never powers-off the PSU. Windows shuts down and there's a click, but the PSU stays on - Every single time. Don’t use the on-board USB, and it powers-off fine (Guessing it’s also a BIOS issue…).
See above; that is just plain odd, though. I'm not of a mind to try to figure out how the capacitor swap might contribute.
Anyhow, back to the graphics card:
I found some vague references to KT133A AGP VDDQ issues, but they aren't really clear. You'd also think a more modern 8x card that has to accept down to .8v would be tolerant in that regard.
It never hurts to turn the "Spread Spectrum" BIOS options off, if they aren't, particularly any relating to the AGP bus while you're questioning an AGP card.
As to the voltage question... I admire your tenacity, but if you're going that far, have you taken readings directly off the pins of the monitoring chip? I used to think the monitors had a huge margin of error as well, but after my little DDR2 adventure (admittedly on a much newer board, where such features might have developed much tighter tolerances) I'm inclined to suspect they might actually accurately reflect conditions at their point on the PCB. (Which, if a bias is being introduced somewhere, might be the reality for the other components on that PCB...)
Along with all that, I'd still try to find some way to test whether it actually is the card's graphics memory. Good luck with that, let us know if you ever find a way. :-?
-
Thanks for the informative post!
Oh dear :-( The 686B issue could explain a lot of things - Including games being unreliable on my miggy, as a matter of fact...
(PCI Network card in this PC...), as well as a dud network transfer the other day...
Any idea whether a list of affected VIA 686B's exists?
BTW, The USB card is an ALi chip, M5273 A1.
I've checked the MSI (The manufacturer) BIOS Updates for the mobo and they don't mention anything about the VIA issue. Mind you, from what I've read the BIOS only becomes an issue if the manufacturers tried to make their own work-around for the problem.
Although im afraid that im still not inclined to perform a BIOS update (It’s not the value of the board that’s the problem, its the fact that it would all have to be rebuilt if something went wrong...) I could try a full Ghost backup of the system (Remember that new HDD? It's mounted in a caddy and is intended for this kinda thing) followed by a VIA drivers update. Might fix the slow UDMA mode speeds (3 At most...) at the same time...
Re the voltages, true, I wasn't measuring them at the monitoring chip.
I'll see if I can run a Ghost backup tomorrow and see if the VIA driver set helps.
Ah, this video RAM test looks pretty good:
http://mikelab.kiev.ua/index_en.php?page=PROGRAMS/vmt_en
Might give that a try, too.
Hodgkinson.
-
I've installed the up-to-date motherboard drivers from the MSI website. I’m not quite sure what they've done yet (Though a few new things have appeared in device manager)...I'll have to wait and find out (Just spotted a flash USB-test utility on the mikelab website, that could be useful...).
As for the RAM test, everything seems OK. I might give the self-booting variant a try.
Hodgkinson.
-
The RAM test (In windows) showed no errors ;-) I’ve not spotted any of these flickers (Yet) with the new PSU.
I've also tried the USB flash test utility from the site - Although I've only experienced the file transfer errors with my USB camera on the USB2.0 card, as you'd probably imagine I don't want to wear out the camera just yet so I've been using a standard Toshiba 2GB memory stick for the tests. Also, the camera (Windows?) doesn’t allow files to be moved back onto it.
No errors showed up in that test, and I've also just moved a 10MB AVI video on and off the stick a 1000 times using a batch file and the video seems to be fine. I’m now running a ~20MB video that previously suffered from corruption once when being transferred at USB2.0 from the camera through the batch file test routine.
I guess I really should of remembered to run these tests before updating the drivers, I suppose :-(
EDIT: 20MB Video test also fine. Ran file compare on all files and no errors reported (Execpt for when I compared a corrupted file I saved from a while back, that did show differences :-D )
Hodgkinson.