Author Topic: ASUS mobo dead (Read 17150 times)

Carol Haynes · « **on:** September 20, 2008, 04:42 AM »

[Note I have gone through and changed the title of this thread. It was originally called "Strange problem with ImgBurn (I think)" but ultimately turned out to be a motherboard problem]

Just curious if anyone else has come across behaviour like this?

Over the last few days I have had a number of occasions where ImgBurn has failed to burn an ISO image to a DVD5 disc. In each case it has siad the device was not ready halfway through the burn. The operation cancelled but the drive kept spinning and nothing would eject the disc except for a system restart. Upon restarting the BIOS cannot detect my DVD drives !! If Windows starts there are all sorts of problems not least of which is that the system is unusable because there is obviously an interrupt hogging a large proportion of the CPU time - eg. the mouse moves only in fits and bursts. The drives don't appear in the devices list (but the IDE interface does and it says it is working normally!).

The only way to recover the system is to turn off the computer, remove power completely and leave it long enough for all capacitors in the system to discharge. When it starts up again everything is fine.

At first I thought it was a hardware problem but I can burn the same image on the same drive using Nero. On one occasion a DMA/CRC error was reported by Nero but it was handled properly and the device unlocked.

Is it possible for an instruction to a controller to block access to drives in this way until capacitors are discharged?

I don't think it can be a dying IDE controller because I have my hard discs on the same controller and have had no issues at all with them. Granted they are on channel 0 and the two DVD-R DL are both on channel 1 so maybe it is a problem with one channel. However, it is strange that I only seem to have this problem with ImgBurn. Could it just be a bug in the handling of one unusual error type? I have noticed in the past the ImgBurn doesn't handle errors totally gracefully and often takes a long time to release a disc - even just for a simple bad block error.

Any ideas?

4wd · « **Reply #1 on:** September 20, 2008, 06:16 AM »

Just curious if anyone else has come across behaviour like this?

...

The only way to recover the system is to turn off the computer, remove power completely and leave it long enough for all capacitors in the system to discharge. When it starts up again everything is fine.

Any ideas?
-Carol Haynes (September 20, 2008, 04:42 AM)

The only times I've had ImgBurn go toes up on me is when trying to:
1) write to poor quality discs.
2) write to a disc at too fast a speed.
3) some kind of ASPI layer conflict from other software.

In all the above, never has it required removing power from the system to fix. If ImgBurn didn't want to exit then killing it with TaskManager let me open the drive, (I think anyway, it's been a while since ImgBurn has failed, 2+ months and 100+ discs).

However, I have had previous faults that cause the same kind of problem, (and in almost all cases that I remember), it was caused by an interface cable coming loose, (not necessarily for the burner).

Since upgrading my system, reinstalling the OS and using only TY DVD-Rs I've not had any problems at all with ImgBurn, (and it's the only thing I use).

It used to be that any kind of optical writer was to be positioned as the master on any IDE controller. I don't know if that still holds true or not but I do know that if my BenQ DW1540 is not the master then I have pretty much no hope of writing a good disc, {was on the same bus with a Sony AW-G170A (Optiarc 7170) as master}. My old NEC 3520 and 2500 were pretty iffy as slaves also.

Nowadays I have a Pioneer 212 (SATA) and the Sony AW-G170A (IDE), the Sony is on the only IDE bus as master with a 400GB HDD as slave.

You could try changing your IDE cabling so that you have the one DVD-DL as master on each bus with the HDDs as slave.

Have you checked the System Event Log? You might find more info, (errors/warnings), in there.

Carol Haynes · « **Reply #2 on:** September 20, 2008, 08:33 AM »

Oddly there is nothing in System or Application Event logs at all.

I'll try removing ImgBurn and installing the latest build from scratch and see if something somewhere has got corrupted?

As for burning stuff - I have two Pioneer DVR-111 drives on IDE-1 (both master and slave) and they have both worked fine for well over a year now.

I use Verbatim DVD-R x 16 discs and Verbatim DVD-R DL x8 discs - both of which seem to give reasonably consistent results on my system. I never burn them at full speed (even though my drives support those speeds). I usually use 50% of rated speed for the discs.

As far as I know I haven't got any software that adds an ASPI layer.

If reinstalling ImgBurn doesn't solve the problem I will try a cable swap.

Actually I have ordered a replacement mobo just in case. I have 3 desktop systems all based on 939 technology so getting a mobo while I still can means I have some reserve plan if something goes pear shaped!

Anyone got any experience with the ASRock 939N68PV-GLAN mobo - it seems to be one of the few 939 boards that is still available. Amazon UK still had a few in stock at ~£40 - which didn't seem a bad price. Main criticisms I read were the ATX connector is poorly sites, SLI graphics cards are tight fit w.r.t. memory cards and there is no Firewire.

I can't see any of these are huge negatives (I have a number of Firewire cards sitting doing nothing so that isn't an issue), the card has integrated nVidia 7 Series graphics (with both D-Sub and DVI inputs which can run dual monitors). I don't play games so that wouldn't be an issue (and I have an SLI graphics card if needed anyway). Connectors only really have to be connected once so they are't really a huge issue AFAICS.

My current ASUS board only seems to be available second hand at ridiculous prices (i.e. higher than the price I paid when new - which was expensive).

Shades · « **Reply #3 on:** September 20, 2008, 12:10 PM »

With IDE, was it not that the slowest device dictates the speed of all devices connected to the same cable?

Anyway, when I was in the PC repair business this was a golden rule. IDE harddisks are already slow enough as they are, why cripple them even more?

Carol Haynes · « **Reply #4 on:** September 20, 2008, 01:44 PM »

I think with the old ATA66 cables that was the case but with the introduction of ATA100 I think that was less of a problem.

Having said that I prefer to keep hard discs and optical discs separate.

Stoic Joker · « **Reply #5 on:** September 20, 2008, 06:51 PM »

Only time I've seen an optical drive lock-up that tight is right before it died. I've got a Plextor DVD burner now, but before that I had a Lite-on and used CDR-Win exclusively. The Lite-on drive did what you're describing for a while right before it died. I have no issues with the Plextor/Imgburn combo now.

*Shrug* ...Just thinking out loud.

An IDE channel will still run at the speed of the slowest device.

4wd · « **Reply #6 on:** September 21, 2008, 04:22 AM »

An IDE channel will still run at the speed of the slowest device.
-Stoic Joker (September 20, 2008, 06:51 PM)

But since almost all late model writers are UDMA4 I don't see that as a problem.

And according to Wikipedia this is not correct for all modern host adapters, (not saying Wikipedia is correct but it does coincide with what I have previously found on other sites).

What you might run into though is the "one operation at a time" issue.

Carol Haynes · « **Reply #7 on:** September 24, 2008, 06:25 PM »

Oh bum - failing IDE interface on my mobo as far as I can tell.

The mobo no longer POSTS (drive LED comes on and then waits for ever for the check to complete).

Luckily I have another identical mobo so I have swapped them and now everything seems to be working fine (fingers crossed).

The knackered mobo is still under guarantee so now I have to do battle with ASUS tech support - not a happy prospect!

Stoic Joker · « **Reply #8 on:** September 25, 2008, 06:01 AM »

Sorry, I hadn't intended to be a prophet of doom ... Glad you had a spare.

4wd · « **Reply #9 on:** September 25, 2008, 07:48 AM »

The knackered mobo is still under guarantee so now I have to do battle with ASUS tech support - not a happy prospect!
-Carol Haynes (September 24, 2008, 06:25 PM)

Why do you have to deal with ASUS?

Is UK trade practices that markedly different from Australia?

ie. In Australia a sale agreement occurs between the retailer and the customer, not the manufacturer and the customer, (unless the retailer is the manufacturer of course). Therefore if the product fails within the warranty period, you return it to the retailer and it's up to them to repair, replace or refund.

Of course if they test the item as OK, you're up for technician time and postage.

Carol Haynes · « **Reply #10 on:** September 25, 2008, 09:45 AM »

In the UK the supplier handles the RMA process normally for a period of time. ASUS boards are guaranteed for 3 years but few shops/retailers are going to provide RMA support for 3 years.

I can't get the board to post any more at all - I have stripped it to only CPU, 1 memory stick (which is fine because it works faultlessly in my other system) and a graphics card - there is nothing else attached and I have tried 2 different PSUs. It doesn't even get as far as a beep code.

4wd · « **Reply #11 on:** September 25, 2008, 05:55 PM »

In the UK the supplier handles the RMA process normally for a period of time. ASUS boards are guaranteed for 3 years but few shops/retailers are going to provide RMA support for 3 years.
-Carol Haynes (September 25, 2008, 09:45 AM)

Thanks for the info Carol.

Good luck with ASUS.

Carol Haynes · « **Reply #12 on:** September 26, 2008, 06:09 PM »

Actually I went to the ASUS website and they said go to your supplier. The supplier website we only provide RMA support for 12 months, and then only for manufacturers that don't handle RMAs directly.

I decided to ask for an RMA anyway at the supplier (Overclockers UK) and they said OK. So now I have to box it up and send it back for testing. Last time I did that they couldn't find anything wrong with it. How do you ensure a motherboard doesn't work without any obvious physical damage?

4wd · « **Reply #13 on:** September 27, 2008, 05:38 AM »

I decided to ask for an RMA anyway at the supplier (Overclockers UK) and they said OK. So now I have to box it up and send it back for testing. Last time I did that they couldn't find anything wrong with it. How do you ensure a motherboard doesn't work without any obvious physical damage?
-Carol Haynes (September 26, 2008, 06:09 PM)

Put some electrostatic discharge into it

Make sure the board is not grounded in anyway, sit it on some polystyrene foam. Charge up something that you know will generate a decent ESD spark and zap it a few times. If you can't get a spark to jump, try connecting a ground wire to one of the mounting screw holes.

eg. Our vacuum cleaner has metal tube for the wand, I know if I wear the wrong socks while doing the vacuuming I can generate up to a 1/2 inch spark when I put it near a metal fixture, (bed frame, piping, etc).
Not to mention it gives me one hell of a wallop if I forget to discharge the b*gger before touching it.

If you've got a peizo gas igniter gun, use that a few times on it.

Shouldn't leave any trace but will probably kill a couple of ICs.

Carol Haynes · « **Reply #14 on:** November 01, 2008, 12:38 PM »

OK - they have admited the motherboard was faulty and suplied a replacement.

I sent them an ASUS Socket 939 (supporting Athlon64 X2 chips) board with PCIe SLI support etc.

The replacement is a Socket 478 Intel board with AGP graphics!!

They have agreed they got it wrong but say they are unlikely to be able to source the correct board for replacement. How can a company offer 3 years warranty and not have access to replacements for the period of the warranty? They now plan to offer a refund on the mobo - but then what do you do with a 939 CPU and DDR memory?

Talk about frustrating!

f0dder · « **Reply #15 on:** November 01, 2008, 12:47 PM »

Christ, that's lousy! >_<

Dunno what to do about 939 CPU and DDR memory. Aren't there any other 939 boards available for purchase? If I had a spare 939 board I would've offered sending it over, but the one I have is in a machine destined to replace my mum's P4-celeron (which is so slow that starting live messenger is an ordeal).

Carol Haynes · « **Reply #16 on:** November 01, 2008, 02:47 PM »

It's not really a big problem but it is more a matter of principle. If companies offer 3 year guarantees they should be able to honour it and if not offer a new up to date motherboard and include any necessary upgraded hardware.

What I have said to overclockers is that would accept either an AM2 motherboard with similar spec and a discount on new CPU and memory - or alternatively a low end built computer (they actualy have one that is an AM2 boards with an Athlon64 X2 CPU and 2Gb or DDR2 memory for only £20 more than the mobo I returned) that costs a simlar price to the mobo combo I purchased.

4wd · « **Reply #17 on:** November 01, 2008, 08:15 PM »

It's not really a big problem but it is more a matter of principle. If companies offer 3 year guarantees they should be able to honour it and if not offer a new up to date motherboard and include any necessary upgraded hardware.
-Carol Haynes (November 01, 2008, 02:47 PM)

Not coming down on anyone's side here but don't all normal guarantees state: "repair or refund."

They decided it's not economically viable to repair it so they are offering to refund your money - something it usually takes a company a very, very long time to offer, (usually requiring multiple returns of a faulty product and even then only reluctantly offering it).

Unfortunately, the down side is that if you have purchased items based on that one piece of hardware, you take the gamble that whatever they replace it with will be compatible or, as in your case, if they give a monetary refund that an alternative is still available. Something that's almost impossible to guarantee with the current pace of innovation.

As an alternative, try looking for an AsRock motherboard if you want to keep the existing CPU/RAM, I think the only two still available are the 939N68PV-GLAN and the 939NF6G-VSTA.

I have two Asrock motherboards and both have been very reliable, (K7S41GX and a 939SLI32-eSATA2).

AsRock are also a subsidiary of ASUS and usually implement features on their motherboards before ASUS does.

EDIT: Just noticed you mentioned the 939N68PV-GLAN up a few posts. All I can say is my experience with AsRock motherboards has been better than with ASUS, EPoX and earlier Gigabyte boards, (now using a Gigabyte MA78G-DS3H).

J-Mac · « **Reply #18 on:** November 02, 2008, 02:02 AM »

I have a similar board in my box Carol: Asus A8N32-SLI-Deluxe. Lot of trouble with the system here from day one. Most of the mysterious crashes I've had turned out to be stemming from nvata.sys. After much searching and posting all over the web I found that it has always been problemsome, most likely due to poor drivers. And Asus and nVidia have abandoned the board and any drivers for it.

And I agree that dealing with Asus support has been a big headache here too.

Good luck!

Jim

Carol Haynes · « **Reply #19 on:** November 02, 2008, 03:08 AM »

That's interesting - it is the same board that I have returned as faulty. Having said that I bought two of them and have been using them both for 2 years with no problems (except that on one of the boards one of the SATA ports has never worked) until one pretty much died on me.

I did buy one of the ASRock GLAN boards when I returned the ASUS board for replacement so in a way I don't really need a replacement but it is the principle of the thing. No motherboard stays in production for 3 years (you are lucky if they stay in production 6 months) so the guarantee system is totally meaningless. A refund on a motherboard is OK except that you don't a have a computer any more - imagine taking a car into a garage any they say sorry folks we can't supply a new distributor so we will refund the cost of the distributor .... The motherboard is fundamental to the kit you buy for a system so if they can't supply a replacement they should supply a more upto date model and compensate for the useless hardware by at least supplying a new CPU and memory of similar spec. Personally I would be happy with an equivalent spec/price AM2 motherboard as it is far cheaper and easier to soucre a new CPU and memory - and there is no reason why they could give an 'at cost' discount to compensate for the hassle. If I had been relieant on this system I would have been without it for well over a month now (and it still isn't resolved) - I can't imagine how I would have dealt with a customer if I was trying to fix their system!

f0dder · « **Reply #20 on:** November 02, 2008, 05:26 AM »

Oh, by the way - with nforce chipsets, be SURE NOT to install the nvidia firewall - it's one of the buggiest and most crash-prone drivers I've ever seen.

Carol Haynes · « **Reply #21 on:** November 02, 2008, 06:30 AM »

Yep - I discovered that the hard way!

Dormouse · « **Reply #22 on:** November 02, 2008, 06:46 AM »

Carol, is there any chance of changing the title of the thread?

ie crossing out ImgBurn and plugging in Asus. So current readers recognise the thread and new readers don't think there's a problem with ImgBurn.

f0dder · « **Reply #23 on:** November 02, 2008, 06:59 AM »

That'd probably be a good idea

Carol Haynes · « **Reply #24 on:** November 02, 2008, 07:35 AM »

Done - I have edited the whole thread.