Gigabyte GA-7VAXP Recap

[Birdyman]

First ... thank you BADCAPS.NET and all the posters

Second ... I am entering these posts "post-fix", but will try to do it as if I had been posting as I worked through the issues. Hopefully someone can get some use out this thread this way.

Third ... I vow to take this thread to completion. I will not only post the symptoms and problems leading up to recap, but also to post follow-up documenting what the recap solved, and what it didn't solve.

Cheers for now.

[Birdyman]

Re: Gigabyte GA-7VAXP Recap

The following is why I ended up on BADCAPS.NET. This was a perfect system from 2/03 until 6/06. I keep it ultra clean and up-to-date - tested software only - no useless bells and whistles when I can control it.

The system:

• Windows 2000 SP4 w/ all security, program, and drivers up-to-date
• Norton Internet Security 2005 (which is another topic of its own)
• Gigabyte GA-7VAXP w/ VIA KT400
• AMD Athlon XP "Thoroughbred" 1800+
• 256 (x2) MB Kingston PC3300
• Promise RAID (onboard) w/ 2 WD Caviar 40GB in RAID 1
• ATI Radeon 7500 All-in-Wonder
• 2 Sony CD/DVD
• Sparkle Power ATX-300GT

THE PROBLEMS (most of which are random):

• Blue screen with and without automatic shutdown - depends on STOP code. Some common STOP codes are 0XD1 and 0X1E. Tend to be more common when HD activity is high, but not always the case.
• Problem with Internet Explorer "has encountered an error" resulting in a restart of IE. This error is usually in MSHTML.DLL or UNKNOWN module.
• Problem with Media Player 9 "has encountered an error" resulting in a rrestart of WMP. This error is usually in the INDIVBOX.KEY module and won't let me play licensed music. I also get this when I try to rebuild my music library.
• Problem with Explorer.exe encountering error. Sometimes it recovers, and other times it may hose system.
• Problem with Quicken encountering error and requiring restart.
• Problem with Symantec Network Proxy service (ccProxy.exe) stopping and causing network to be unavailable. FYI, this was the original issue I had and dates all the way back to 6/06 - took a while to track down. Network was restored by simply restarting service. Thought Norton was to blame for all this mess, but not so sure now.
• Broad range of errors logged in sytem event log, as well as a few errors in the application event log.
• System voltages fluctuating more than normal. CPU voltage was 1.5v +/- 5% . DDR Vtt and other voltages showed similar fluctation.
• On my 3rd power supply. First one lasted 2.5 years.

Will post on troubleshooting next.

[Birdyman]

Re: Gigabyte GA-7VAXP Recap

OK, here's the condensed troubleshooting list.

• New power supply. Helped the problem with dead power supply, but DID NOT help the other problems.
• Reinstall Norton Internet Security (for 3rd time). DID NOT help.
• Cleaned system thoroughly and reapplied thermal grease to CPU heatsink. DID NOT help.
• Ran burnin style tests on CPU, MB, RAM, and HDD. All OK. DID NOT help.
• Inspected MB closely. HELPED! Found all four 3300uF GSC crapacitors were bulging and leaking. No other caps showed sign of damage.
• Shut down several services including the RAID event log service. DID NOT HELP.
• BACK rev'd most of the peripheral drivers and BIOS - audio, graphics, RAID, LAN, etc. DID NOT help.
• UP rev'd the peripheral drivers and BIOS to version before last step - audio, graphics, RAID, LAN, etc. DID NOT help.
• Disconnected system fan - it was getting loud, and my system and CPU temps were 42C idling without it. HELPED! The number of blue screens dropped dramatically. The other errors also became less frequent, though they still appeared. HMMMM ... definitely going to suspect the capacitors now!
• Ran Windows System File Checker (SFC). Found quite a few problems with version mismatches on system files. HELPED a little. Starting to indicate problem with HD. Original problems still remained, but new problems cropped up.
• Had to reinstall Internet Explorer, Outlook Express, DirectX, and one or two more because of SFC modifications to system files.
• Used Microsoft Baseline Security Analyzer to go back and update system files that SFC modified. HELPED to eliminate the problems that appeared after running SFC, but DID NOT help the original problems.
• Getting desperate, I started analyzing files, folders, registry entries, etc. DID NOT help.

Next, it's time to make some decisions.

[Birdyman]

Re: Gigabyte GA-7VAXP Recap

OK, I can't justify doing a reinstall of Windows until I get the hardware fixed. And if I'm going to pull this system apart to replace the caps, I might as well do some other stuff while I'm at it.

Decided on Digi-Key, because it's one stop shopping. Mouser, Newark, and Arrow didn't have all the components I needed in stock. From Digi-Key's capacitor offerings, I chose to use the Panasonic FM series for 3300uF, 1500uF, and 1200uF. The FM series specs are excellent, and they are on the GOOD LIST. For the 1000uF, I chose the Chemi-con KY series because they are the only 8mm 1000uF caps Digi-Key offered that are on the GOOD LIST. I had to have 8mm because the 1000uF caps go in some tight spots.

Digi-Key Shopping List:

• 10 3300uF Panasonic FM series
• 10 1500uF Panasonic FM series
• 10 1200uF Panasonic FM series
• 25 United Chemi-con KY series
• Desolder braid
• No-clean solder
• No-clean flux

NewEgg Shopping List:

• Arctic Cool AF8 80mm system fan (quiet)
• MassCool Socket A 80mm CPU cooler (semi-quiet)
• Zalman ZM-NB47J Fanless Northbridge cooler
• OCZ Ultra 5+ Silver Thermal Compound

Next, the recap.

[Birdyman]

Re: Gigabyte GA-7VAXP Recap

Man, Digi-Key and UPS rock. Had order in less than 48 hours using UPS 3-day shipping!

Below are some pics of the board before and after recapping. I started the soldering iron about 8pm. I used a Radio Shack 40w left over from college, although I had to give the tip some love before I could use it. The 3300uF Panasonic caps were 12.5mm whereas the original GSC's were 10mm. I had to do some bending on the leads to get them to fit the space correctly, but it turned out pretty good.

I heated the cap leads with a "wet" tip from the back side of board and gently wiggled old caps out. No real problems doing this. Trimmed the leads of new caps to something more reasonable before I installed them. Once I got new caps in, using the heat-and-rock procedure, I trimmed excess from leads leaving about 1.5-2.0mm protruding from board. I fluxed the solder area, then touch "wet" solder tip to the tip of the cap lead. The solder flowed down over lead and onto pad. Any excess solder on tip of cap lead was trimmed off. Once done, I viewed backside of board (from the side/profile) to make sure that none of the new cap leads extended farther from board than any other component leads. Cap lead touching chassis = BAD!

To finish, I gave each cap a wiggle and closely examined the solder joints for movement. Next I closely examined the areas around the new solder joint to ensure that no solder had gotten onto adjacent areas - good thing I did this - I found 2 spots that I had to use the desolder braid to remove some solder. Next I examined the adjacent areas to ensure that I had not removed solder from any other part of the board - by the way, this is something you should REALLY be careful of when you do your solder job. Though I used no-clean solder and flux, I went ahead and used some board cleaner and lightly scrubbed each of the new solder joints. Finally, I put machine back together.

1:30am ... Houston, we have POST!

Next, will post more on the initial performance of system after recap.

[Birdyman]

Re: Gigabyte GA-7VAXP Recap

As stated, here are some performance notes on the recapped board. The machine has been running Windows for about 8.5 hours at this point. The values were taken from a hardware monitoring utility that came with board. All peripherals are installed and working, and case is buttoned up.

Voltages:

• Vcore before: 1.532 +/- 3.0% after: 1.500 +/- 0.0% (perfect)
• DDR Vtt before: 1.260 +/- 1.6% after: 1.240 +/- 0.6%
• All other before: noticeable fluctation after: none to little fluctuation

Temperature:

• CPU before: 42C idle, 52C active after: 31C idle, 35C active
• SYS before: 42C idle, 46C active after: 39C idle, 39C active

Windows/System Error or Issues:
None


So far the machine is running great ... and quiet. Before recapping, it's almost a given that I would have encountered at least one IE or ccProxy error at this point in the day ... but so far it's stable! Will post more as I know more.

Next, the Gigabyte GA-7VKMLS board in the kids computer.

[willawake]

Re: Gigabyte GA-7VAXP Recap

excellent writeup. welcome!

[Birdyman]

Re: Gigabyte GA-7VAXP Recap

Well, the machine has been running continuously for around 72 hours now, and I've been doing A LOT OF WORK on it. So, I think I can post an update of the issue resolution/outcome with a good degree of confidence now.

ISSUES SUMMARY:

• RESOLVED! Blue screen with and without automatic shutdown - depends on STOP code. Not one bluescreen or shutdown since recap - normally would have seen 6-10 by now.
• RESOLVED! Problem with Internet Explorer "has encountered an error" resulting in a restart of IE. This error is usually in MSHTML.DLL or UNKNOWN module. Due to file corruption (probably from random reboots), I had to reinstall IE and all updates. No errors since reinstall.
• RESOLVED! Problem with Media Player 9 "has encountered an error" resulting in a restart of WMP. This error is usually in the INDIVBOX.KEY module and won't let me play licensed music. I also get this when I try to rebuild my music library. Due to file corruption (probably from random reboots), I had to reinstall WMP9 and all updates. Additionally, I had to hack on the Digital Rights Management files, essentially deleting several corrupt DLLs and reinstalling from KB891122 (I had to figure this one out on my own as Microsoft doesn't publish much on the DRM files systems). No errors since reinstall and I can now play my licensed music files.
• RESOLVED! Problem with Explorer.exe encountering error. Sometimes it recovers, and other times it may hose system. No errors seen so far.
• RESOLVED! Problem with Quicken encountering error and requiring restart. No errors seen so far.
• RESOLVED! Problem with Symantec Network Proxy service (ccProxy.exe) stopping and causing network to be unavailable. FYI, this was the original issue I had and dates all the way back to 6/06 - took a while to track down. Network was restored by simply restarting service. Thought Norton was to blame for all this mess, but not so sure now. No errors seen so far.
• SOMEWHAT RESOLVED!Broad range of errors logged in sytem event log, as well as a few errors in the application event log. Still see one network related error coming up, but honestly I think it's been around for years.
• RESOLVED!System voltages fluctuating more than normal. CPU voltage was 1.5v +/- 5% . DDR Vtt and other voltages showed similar fluctation. Still stable.
• ONLY TIME WILL TELL On my 3rd power supply. First one lasted 2.5 years.

So who would have thought that bad capacitors would have screwed up a machine so badly?!?!?!

[rkchkr]

Re: Gigabyte GA-7VAXP Recap

[starfury1]

Re: Gigabyte GA-7VAXP Recap

Excellent write up I'll seconded that.... welcome too

Good to see another MB saved from possible scrapping

Yes

So who would have thought that bad capacitors would have screwed up a machine so badly?!?!?!

Not as many as should be me thinks

I had a Gigabyte MB that was playing up and yeah, was thinking it was software issues...then I decided to have a close look at the pcb.....yep Caps were the problem.
That MB is running fine now after recap, Windows stable ...ok well as stable as windows is

That was a GA-7ZXE

(Willawake did a post on recapping this MB) yep GSC caps were the problem

Thing is you are not expecting something 2~3 years old to have stuffed caps
(in a home situation)

Well I wasn't then but do now!

It does make me wonder how many people are blaming windows for the problems when it in reality a HARDWARE problem causing corruption of windows install.

How many software people have been abused for not fixing the problem when it in reality its not a software issue.

So if you are having constant reboots and BOSD etc, check your Hardware for badcaps....then fix/ reinstall windows

It may save you some bad Karma if you are trying to resolve computer problems for somebody else

This box has a 7VAXP ULTRA but no GSC to be seen, the last time I looked the caps looked ok
not sure of brand thought off hand

This site is a fine example of why you should check your hardware and a Great Voice of this BADCAP debacle..
Letting the world know

Thanks TC, Mods also the forum members for making this So!

Humm maybe M$ helpdesk should have this site as a link on their reply emails may save them some abuse,
I am sure that they would have got some for something that isn't their fault


Just my thoughts on it

Cheers

[Birdyman]

Re: Gigabyte GA-7VAXP Recap

It does make me wonder how many people are blaming windows for the problems when it in reality a HARDWARE problem causing corruption of windows install.

How many software people have been abused for not fixing the problem when it in reality its not a software issue.

Yeh, I have a pretty good idea of how many. I spent several years with Dell then several more with Arrow Electronics. I was always more of a systems engineer than a board level guru ... but I have done it all at one point or another. I can't tell you how many times I've had to give a "sales pitch" to a colleague, customer, or exec trying to convince them that the hardware was at fault - ESD damage, board revision change, poor quality - you name it , I've seen it. Hardware is a hard argument to make, simply because you can't connect to the web, click the mouse a few times and get an instant hardware upgrade to resolve the issue and prove your point! Besides, the average person thinks the programmers should have thought about some ingenious way to get the code to execute, even with non-functional hardware!

Of course, I understand the difficulty some people have buying the hardware argument, as hardware and software are so tightly intertwined. It's hard to distinguish bewteen the two sometimes. But, from my standoint, when you have a problem that slowly gets worse over time, with the software setup remaining realtively constant, hardware is generally the first suspect. Hence, this is precisely the reason I'm so anal about throwing SW on this machine. Interestingly, that's what made this particular problem so tough, because I made over 20 SW mods about the same time period the problems started ... but NOT ONE SW mod was eventually attributable to the problem ... just a few caps ... WOW, huh?

Man, the SW vs. HW argument is one of my favorites, and I have spent years trying to perfect it. Here's my take. Software, in reality, is conceptual. It is based on a programmed flow that depends on the value of a single bit, and dependent upon hardware for life. In a simplistic point of view, when an unexpected change is made to just one bit of executable code, the execution is interrupted, and it usually results in a complete break. As an example, make one "unexpected" change to just one bit of "non-data" code and see what happens. Will you see an adverse affect immediately? If the piece of code that got modified never makes it into the CPU, then you probably wouldn't notice it. But, when that piece of code DOES hit the CPU, you know what's gonna happen. CRASH! Look at it like this ... Did we make the smallest possible increment of change (i.e., one bit) to an execution component? YES. Is the component in use? YES. Did the system break? YES. Will it ALWAYS break with this modification? YES ... it is ABSOLUTE - it is in the program. Your code would require redundancy PROGRAMMED IN to circumvent the change.

Hardware on the other hand is different. It's PHYSICAL, and inherently has a range of operation. I could go tweak the voltage on a board without breaking it. Let's say you lower the voltage from 12.0V to 11.999V. Did we make a very small change to an "execution" component? YES. Is the component in use? YES. Did the system break? NO. Will it ever break? Maybe, but that's a function of probability and is based on the DEGREE OF CHANGE ... it's not absolute until a RANGE LIMIT is violated. It's the probability of error occurence that kicks our butt when diagnosing early HW failures.

I'll drop it here before the moderators kick me off. I gotta get some sleep anyway, as I haven't seen the bed much lately.

[starfury1]

Re: Gigabyte GA-7VAXP Recap

Well put the above

I can see why you went down the SW path as you had made many changes and somethings wont play together well.

I would have too

Seems the badcap issue is so wide spread over the last few years that more then likely this will be the cause and not software in a lot of cases.

So thats why I think it is an important first step.

Hopefully now with this half baked electrolytic formula behind us and the use
of the new breed of caps these issues will fade

Not much you can do about el cheapo made caps or low quality Mfr, except not buy it.

RoHs seems to be bringing its own set of problems for hardware from what I read here on the forum

Sounds like you have had your fair share of problems over the years and an extensive customer contact.

Although I have been involved with electronics over the past 30 odd years At this point in time I am not active in the electronics field work wise now.

Which is why I say "not the voice of experience here"...cause I know a lot forum members here work with this stuff on a day by day basis and have a vast hands on knowledge and experience of todays stuff

Cheers

[Birdyman]

Re: Gigabyte GA-7VAXP Recap

Yes, you have to love BADCAPS.NET because they're making one helluva HW argument. Not just talking about it, but arming the average person with what they need to actually get up and get the job done. Great stuff IMHO - almost like a grass roots kinda thing for the computer industry.

You know, board mods really aren't much more effort than major SW mods ... just takes a few tools (hey ... maybe that's it ... nobody owns tools anymore), and a little knowledge of hardware (that society has decided to quit teaching). I remember being 15 years old (1982ish) and deciding to install my first car stereo system. A few chats with a girlfriend's uncle, a few trips to Radio Shack, a little reading, a scratch pad, a good Saturday afternoon, and I was done. That's the first electrical project I ever tackled, and it actually worked! Sad, really. I really miss Radio Shack

And yes, I think that some of the industry leaders should make reference to BADCAPS.NET seeing how widespread the problems appear to be. But IMHO that would interfere with the underlying politics that keeps the "upgrade" money flowing in the industry. Even the hassle that SW makers like MS may put up with is offset by the fact that when the customer finds that the [out of warranty] HW is at fault, they just go buy a new PC ... which of course comes with a fresh new MS OEM license. Ingenious really from a business perspective - don't waste money directing people to fix their HW, make money selling them new SW on a new piece of HW. But, in the end, that's what makes for a good movement ... standing up to THE MAN and letting your voice be heard

I'll sign off now, before I start pulling out the poster board and making signs.