I think lightning fried my mobo

[taomation]

We had a close strike in an electrical storm. Lost several consumer electronic gadgets. in addition my homebuilt computer re-booted, ran about 30 sec then went off and into a endless re-boot cycle. It will post and enumerate devices, you can enter the bios, but it won't boot. Single beep after post, then restarts.

System= GA-p35-ds3p board, (rev 2.0), 4x1gb corsair 1066 ddr2 RAM, int q6600 quad psu, 3 x 250 gb segate SATA HD's, one boot drive, (no raid), 650 watt 4-rail HEC PS, MSI 260GTX graphics card, Creative audigy zs2, ATI 650 TV card, Floppy (??) lite-on sata DVD-RW_DL, sony DVD-ROM, Various USB, devices, (wireless mouse, keyboards, air-mouse etc.) Win XP pro, SP3.

Diagnostics= clear CMOS and repost, then testing each stick of ram, one by one, then the system was stripped down removing PCI cards and other devices including non boot drives, floppy and DVD's. Next swapped HEC 650 wat power supply with new HEC 650 watt PS, Next, swapped PSU with working pentium D, next, swapped video card with working GA-7300gt, next reconnect DVD and set bios to boot from CD. System would not boot to win XP CD. same symptom, single beep folowed by restart.

conclusion=fried mobo, valid or not??

Unfortunately, I did not see the exact replacement board available and ordered a GA-EP45-Ud3r which will allow me to use the same 1066 RAM.

I also note that since SP3 and IE8 are installed I cannot do a repair inst., but I would really not like to have to start from scratch. I now I can take data of my drives by hooking them to another machine in an external case, and I intend to do that first, but I would really like to use the current build; lots of little reasons. Obviously I cannot prepare by uninstalling old board related hardware. I guess the biggest concern is the ICH10 south bridge vs my old ICH9R, particularly, in how the boot drive floppy and DVDs’will be recognized by the IDE controller etc. They are very similar, but not the same.

Plan = attempt to boot (in safe mode) get into device manager and uninstall old devices, and get the new devices installed. I realize this may take more than one restart. Again I probably see the IDE controller as being a priority.

Should I keep the system stripped down, just the boot drive board and graphic card?
Or should I make the system look as close to the last hardware configuration?

Is there a better plan?

 

[hunter315]

Take the hard drive and put it in another machine to see if you can access data off of it, if not then it may have gotten cooked. Since you seem able to pass post and make it to bios it makes the motherboard seem less likely to me. I would guess that your PSU got a bit cooked or your HDD got toasted by the strike to give you the ability to get into bios but not to boot the rest of the way.

 

[taomation]

I backed data from the boot disk and that worked fine, mounted in a different pc. Ran chkdsk first and didn't see any problem. Data copied without incident. I don’t plan on including the other HDDs in any start up adventure, so I’ll deal with them later. Just for giggles, I bread-boarded the system, popped it back in and powered it up. It did the same thing, (re-start cycle) except after one or two, a small fragment was ejected from the board attended by a spark and a bit of smoke. Incidentally, the board continued to re-start and I was still able to enter the bios. At this point sensibility outweighed curiosity and I shut it down. The board is definitely fried now. On inspection, there did not seem to be any ancillary, reasons for the board to short, loose screws, dust bunnies etc, other than previously being subjected to a lightning surge.

 

[Paperdoc]

First, I agree with your assessment that the likely culprit is the mobo, and I see you already have its replacement on its way.

I am not clear why you say Repair Install is not possible. I do understand that IE8 is a problem, and I thought it could be "solved" by simply un-installing IE8, probably necessitating Safe Mode since many of your device drivers for the new mobo won't be there. (Of course, you'd have to re-install IE8 later.) But why is SP3 a problem? I gather your original XP Install disk did not include SP3, so after a successful Repair Install you might have to re-do the upgrade to SP3. But it sounds like you believe the Repair Install can't be done without an SP3 Install disk. I have not run into this before, but I also am not disagreeing with you. I'm curious to know whether this really is a problem for pre-SP3 installations that have been upgraded.

Your alternative of using Safe Mode to delete all device drivers, then installing new ones from CD's etc, probably is a good route if the Repair Install can't be done. It is more time-consuming, of course, than having some software utility tools do all the work for you, but it should produce the proper end result - a fully functioning updated system on a new mobo.

 

[taomation]

Well if the board wasn’t the root cause earlier, it certainly is gone now.

No, my original disk did not contain SP3, and I never had any issues with SP3. I'm sure I could do a repair instal it will leave me with a nonfunctional IE6, or so I've heard, unless I first unistall it.

To do that I need to boot up the computer. I wish I could! If so, I could also uninstall all the old hardware too, but booting up and doing any of this seems not to be an option.

Actually, you could use the recovery console to "disable [device_driver_name]", especially the IDE controller. I suppose windows would load generic drivers. However, I have partially allayed my apprehension. It turns out that the driver issued by Gigabyte for the ICH10 is labeled ICH9R/ICH10, so I guess both controllers use same driver. I’m just going to try it. Hopefully, with the new mobo, the system will boot. Then, it will just be a matter of cleaning up in device manager. I just hope that there isn’t some other underlying problem that actually caused all this.

 

[rockyjohn]

Just curious - assuming lightening to be the likely cause - what kind of circuit protection device were you using? Did it happen to include insurance?

I wish they had exploding board parties at conventions - let you bring in an old mobo or video card, connect it up to some huge transfomer, test how much power it will take before failure (to gain some valuable research on limits and to compare different board), and then goose the power and watch it blow (the fun part of the research - and maybe mildly therapeutic).

 

[taomation]

Well not enough, but factually it was a Furman SS-6B. Insurance? Hum. Well, since I also lost a couple of other gagets too. I'll probably make a claim on my homeowner policy.

The funny thing is the power strip seems to still work just fine, at least if you consider that it doesn't really do much in the first place.

 

[rockyjohn]

Did the circuit breaker on the Furman trip?

It looks pretty strong - but the listing, even on the Furman Sound sight did not even list the joules of protection it provides. I am guessing it might not be very heavy duty.

You might want to consider getting a stronger unit.

 

[taomation]

Well about the Furman; I don't know that it did. But it appear it did not. Or at least that's an opinion I have adopted.

So, an update:

I replaced the board, and it will attempt to boot, and of course I get the NTLDR not found. I'm thinking that the partition on the drive is being recognized. I was tempted to think in terms of the IDE controller, but that is a virtual device, doesn't really exist until windows boots up.

Therefore my choices are:

1 repair install-but I would have an issue with IE8
2 use the repair console to fix the MBR and the boot partition
3 Or reinstall windows

I'm thinking that 2 will probably end up at 3 anyway, so I might as well so the experiment.

But wait, it gets worse. I notice the 260gtx works but dosen't fill the screen, just a frame inside it. Popped it out and put in a 7300gt and it worked fine. The 260 gtx did the same reduced size frame on a different machine. So, unless there's some secret reset option for the card, I guess it goes to the insurance co too. I will wait for the new video card before trying no. 2.

I guess this will teach me not take storms for granted. Just because nothing usually happens doesn't mean it can't and it a genuine pain in the A.

 

[rockyjohn]

I am not an electronics expert, but it certainly seems strange that lightening might be the cause if your PC was plugged into the power saver and its circuit breaker never tripped. Even more so that it would damage the mobo and maybe the video card without affecting the PSU unless the case itself took a direct lightening hit.

But you indicated that this happened right after the strike and you said other electical gadgets were lost - which strongly point to lightening as the cause. A conumdrum.

Perhaps you should test the PSU to make sure it is not damaged. It would be a shame to get new components and then have them damaged by a bad PSU.

 

[taomation]

Well, I have only one piece of data which might indicate the status of the CPU at this point, other than it's ability to operate in the bios environment. It was recognized correctly by name, but posted as with a frequency multiplier of 8 or 2.13GHz when put on auto detect. It should be 2.4GHz. This may just be a bios snafu. The board has bios F5 loaded, however Gigabyte has F9 available currently. Bios operation is not a very stressful environment for a multi-core processor. All I can really think to do at this point is flash the newer bios, and see how this is recognized. Truth is I really don't know how to interpret this. When I get it loaded it working in an operational environment I can render some vireo or something to gauge its performance.

I'm not really aware of a way to evaluate a CPU in a bios environment; is there???? Hey, all these things to can be replaced. It's not even difficult. It's just the time it takes to sort it out piece by piece. I guess that's why they have reference systems. I don't have another build, I can pop this cpu into currently.

Not really sure what the circuit breaker did, if anything.

 

[taomation]

I have found the experiences had by others, and posted to these boards of considerable use and wanted to provide a follow-up on how this concluded.

I ended up executing my original plan which was basically:

1 Replace mobo
2 Load all other hardware so that the configuration would be otherwise the same.
3 start in safe mode and try to swap out hardware using device manager.

In addition, I first loaded the most recent bios, in this case flashing F9 from F5. Typically, there is a concern flashing ahead more than 2 to 3 in bios issues, but they tend to result in the type of problems that I already likely had in abundance. IOW, here it wouldn’t matter.

I loaded the fail-safe bios settings and allowed it to continue to boot to safe mode (F8 while booting) with all hardware PCI, devices hard drives etc in place.

The boot was successful, however the PC seemed unstable, i.e. the mouse was stuttering and UI windows were leaving trails when dragged across the screen, etc., but no blue screens. Hardware was being found in abundance. I first started just closing these, but there seems and endless parade, so I just lest the next one come up, ignored it and went into device manager.

I set device manager to show all hidden hardware and there was an alarm by a good number of things. Notably, the IDE controller was working. Obviously, I was communicating with the drive having had loaded windows. Also, all the PCI devices showed faults, even though the proper drivers were loaded previously, ah, but the PCI logic for that board must be a little different. The two bridges actually are slightly different in voltage too, I think by 0.5v, I looked it up, but don’t recall specifically. Anyway, the combination of factors was enough for all devices to malfunction.

The SM bus was also alarmed and was being recognized as “? other device”. This is a big concern because this regulates a lot of timing and power features on the board as well packet error checking. I wanted to get what I needed done get out. Interestingly enough USB devices seemed relatively impervious. I would suggest that if you couldn’t even get this far trying to boot you drive as a USB device through a enclosure.

I basically stuck to the plan, I uninstalled the IDE controllers, and few other devices and rebooted to safe mode, ignoring calls to load drivers. I then used the mobo installation disk, but the weird thing is the install program indicated that many of the drivers were already installed. So I used the install program to uninstall most features and rebooted, this time letting windows start normally. Then using the install program I installed the mobo hardware and it worked pretty well, the SM bus was back and the parade of VD’s under it were all showing ok.

Networking was a problem, for some reason the network adapter was clear in device manager, but I could not get a network address. I had to again uninstall the driver and load it back, but it was Ok.

The rest of it was pretty routine business, getting devices installed and updating drivers. Worked out great lost no data and the system seem pretty stable. There was a fair amount of registry clean-up that followed broken shortcuts, etc., but nothing unexpected. I was able to place load on the system, game, video etc. I over-clocked the processer a bit and it continues to run smoothly.

That’s pretty much it.

 

[taomation]

I would also add that configuring the system to look like it did with old board was unproductive. It just added a buch of gagets that were not properly recognized by the system and were at best a nussance and potentially a source of instability.

 

[Paperdoc]

Glad you got your system cleaned up and working again. Now you should go back to check your Furman power strip. On many the device for protection against high-voltage spikes is a MOV connected across the Hot to Neutral leads. Up to some point (say, 500 V) they do nothing. But over that, they act almost like a direct short and allow the current to flow heavily through them, reducing the strip's voltage delivered to the load to safe levels. However, in doing this the MOV becomes the sacrificial component. It is destroyed by this action. The result is that there is no longer any protection device across the Hot and Neutral lines so you have no protection system now, but it still seems to be working because your computer is getting power.

I am not sure exactly how you go about testing the power strip. In some cases if you open it up and look at the MOV's - they sometimes are disk-like devices with two wires coming out, connected across the incoming power supply leads - there is obvious damage like exploded or burned appearance. But if you don't see that, I'm not sure how you check for hidden damage or verify that there is no damage. Maybe check with the unit's manufacturer.

 

[taomation]

I think the best way to test the power strip surge protector is with the trash can.
Funnily enough, I just finished placing an order for a replacement. This time I went for a model that will also protect the co-ax cable and the RJ-45 ether net. I don't remember if I mentioned that a bunch of other assorted gadgets around the house were affected. On my wife’s computer, the ethernet module of her mother board died. I replaced it with a PCI card. The DSL-modem continued to work but the router portion failed. I just replaced it.

I have a few choice words to say about ATT NON SERVICE. Lets just say networking is not my strong suite and when I told them I had bought a router other than 2-wire, from the ATT e-store, they basically just said have a nice day, but we can't help you with the settings. I got it right after a few tries, but really!

Anyway, these days when we have computers hooked up to lots of other things, you have to consider all the ways a rogue surge can get into your system. The coincidence of the modem my wife’s ethernet and my problems and the fact that my power supply and CPU seem to be OK seems to indicate that maybe the fault isn’t with surge protector at all, but with the lack of protection for ancillary connections. Maybe I shouldn’t through the Furman away, but I should replace it with a unit that can handle the other connections as well.

 

[agentbb007]

Typically, there is a concern flashing ahead more than 2 to 3 in bios issues, but they tend to result in the type of problems that I already likely had in abundance. IOW, here it wouldn’t matter.

Really? I've never read anything about flashing the latest bios from an old bios will cause problems. Anyone else read that you shouldn't update from an early bios version to the latest?

 

[taomation]

This was copied from Gigabytes Q-flash utility guide. I'm not qualified to comment, but I recollect hearing this recommendation from others also.

With Q-Flash™ utility, please do not cross too many BIOS versions at one time if
your current BIOS version is rather old. For example, do not cross from F1 to
F12, but to flash from F1 to F4 and F4 to F8 and so on.

 

[agentbb007]

Ah thanks Youngster that's good to know, I've never heard that before.

 

[taomation]

Ok, I'm just a hobbyist. Don't want to wade in to deep. Love to read the treatise though.