K7S5A quirks...L O N G and convoluted post

[krazefinn]

As I'm slowly recuperating after latest round of spinal fusion (dont ask, it wasnt real fun!!) I've pulled out a couple of older but stalwart hobby motherboards still lurking within the dark confines of my shop to timker on...and ran into an interesting challenge. I had used this mobo for near 3 years until I updated last year, running an XP 2200 tbred at 147/147 using the collaborative Honey/OcX 03/8/11 bios with a tweaked ME install. At first I fired it up, and all worked as it should, from the ATI AIWonder, Hercules crystalsound FortissimoII audio, OB LAN, USB2 card, PCI iee1394 firewire card, onboard Intel CNR modem. I first pulled out the TV tuner equipped AIW and rplaced with an MX200. Suddenly board would give the 2 short beep, followed by 8 more short beeps, but of course no boot or bios. Tried with 2 other video cards, and a pci card, and all gave the same symptoms. OK, clear bios again, try again, replace AIW...VOILA! It still worked. Remove video driver, remove display card, put in MX420. Uhoh, same 2 beep=8 beep code. Usually (w AMIbios) 2 beeps means low cmos 2032 battery or improper settings, and 8 beeps indicates one of several display card problems (usually failing video memory, loose card or plug, blown agp slot, etc).Awrighty then, lets startt from scratch, clear cmos, fully dissasemble, install HPT 370A RocketRaid100, 2 new maxtor ATA100's, one stick 256 DDR2700 and floppy/cdrom. Format drives, upgrade clean install XP, etc. All sorts of problems installing any of 3 different sound cards, would either not install driver (code 10) or repeat the now-too-familiar beep pattern. After disabling OB sound (again),USB,CNR,LAN , prevent irq assignation to VGA, disabling serial ports, switching LPT port off and everything else non-critical to strripped boot, I finally got driveers loaded for bios and the display card fully functional (MX420). Start installing the usb2 card, firewire card, divorced TV tuner card, etc, and its back to the same 2 beep-8 beep pattern.

All of the cards in question were checked out on another pc, and all worked, with updated drivers obtained. Cmos battery replaced, PSU tested and TCO within stated parameters...board inspected and no questionable caps were found. Replacing the AIW, the original ME hardrive, etc, and the board works fine. In the new configuration, cannot seem to ascertain why it will not function properly. Now, i know although that pc-chitz red-headed stepchild ECS motherboard had its share of quirks and esoteric requirements, but this conundrum is is very challenging. Performance wise, this board will still hold its own with the unified SiS 735 chipset, and since it was stable, fast and reliable for me for several years, and still works with original configuration, I would really like to obtain the differential diagnosis and ressurect her. I suppose the PSU "could" be suspect, so as soon as I can recover enough to wiggle under the desk to lift the case out, remove/replace the psu, and perform a more thorough visual inspection for any possible bad cap indications.

I'm at a loss beyond that however, and don't believe either issue is relevant, and am wondering if anyone else here (HAVE EXTENSIVELY SEARCHED FOR SIMILAR ISSUES AT a couple of SiS board-specific forums, unfortunately all without success.
Am hoping/wondering if any of y'all have any ideas you'd care to share here regarding the 2-beep/8 beep bios beep pattern, and what to try next. I could simply just get another mobo and start from scratch, but with all this time on my hands presently (doc tells me I will not be cleared to push, pull, reach, twist, bend or even sit still for extended periods for up to 5-6 months...), and willing to take on the challenge of trying to figure this out... throw me a bone here! Alll ideas dulu considered and appreciated...thanx in advance, and I know this board has another life left in it...

[Rocketmech]

Have you cleared ESCD ?

[krazefinn]

Hey Rock, good gambit!

Better to ask "how many times did I short the rtlc contatcs?"

Initially upon the first vga swap, several random efforts between, and finally upon batt replacement.

[Rocketmech]

Maybe the bios , or the extended OC damaged something , perhaps the N.B. ... but I'm making a WAG .

[krazefinn]

Any wag is cool!as far as the OC bios,I set it within specs to default 133/133 immediately when I decided to clean install XP on the RAID/2 drive array. Have found sometimes that even a "negligible OC" can corrupt/defeat an install...really don't know why. And the board and chip both work fine when I reinstall the HD with the older WinME OS. Go figure...
all WAG's welcome,
challenge becoming more interesting... since it works with one drive on the old me, and since I have yet another spare drive, will install xp to it, and try xp without the RAID card (but its ALS)

[chuckbiskut]

I may sound misguided but. is there a newer version of the bios? I didnt see if you had done any updates in your message. Just my 2 cp.

Chuck

[krazefinn]

Yes, there are newer OC bios, from Honey, OcX, cheepoman, ECS, overkookyverkbench and others. The one presently in use, a collaborative effort between 2 of the pre-eminent writers, is widely regarded as the best choice for just about any OS and type of OC, or even to run at default or underclock.

It is id'd as the H/OCx3-8-11 (or something like that)

[genesound]

I did understand your bios info, but I think it still might be a good place to look. Try a stock bios for big changes in hardware and then go from there ;)

[krazefinn]

I fortunately just b4 retiring this board downloaded all available stock and OC bios, updated drivers for XP for all my hardware, and burned to cd. Just added the latest ECS bios updates to that cd, and discovered that one of the ocbios I like was now an "official" one-and apparently this bios is one of the very few that "might" allow fsb and memory settings of 166/166. I may try to use that as a last resort. Have in past "ruined" malfunctioning board by inopportune reflash-but since its already problematic, don't have that much to lose if it gets hosed. The hardware changes are not that dramatic really, as I'm reusing nearly all of the original stuff. And the video card (MX420 NVidia) is era-accurate, and somewhat of a downgrade from the AIW (as well as less power consuming).

We'll have to see today, as I should be able to swap in the old single ME harddrive, and will probably have to reinstall ME (original year-old install went corrupt after one year of amazing-for-ME-stability) I never did find out why, and couldnt even use an ERD, sysrestore or Norton SysWorks premiere to fix it. Kind of curious now that I think about it, that OS hosedmwhen I first tried to simply swap vieo cards from AIW to MX200. Guessing here-suppose that just swapping video cards caused the AGP slot to fail, would/could that corrupt the OS?? EBD scandisk returned tons of errors, and scanreg/restore was unable to be run in dos ostensibly due to degree of corruption.

Advice will be checked out, thanx Gene. Right after trying the single disc/ME first. Truthfully don't really feel its probable esp because hardware is all being reused and what is changed is nothing real dramatic ,but any degree of possible will suffice...my WAbestG is either badcaps in psu or motherboard, or a motherboard agp slot, pci slot, unified chipset fault.

I have spare 350 w psu in another case, but still unable to even sit for 5 minutes, can't pull it for test yet...maybe by mid next week. Let me tell ya, I don't reccomend vertebral fusion except as last resort. The operation by itself is utterly barbaric. Guess its no wonder hurts like the dickens...he was in there with hammer and chisel, saw and powerdrill.

I'm somewhat surprised no one has chimed in yet who has board-specific experience and might actually have encountered the 2/8 beep bios code...after near 60 views...
thanx all, wish me luck.

[krazefinn]

Wanted to also mention I have already also attempted to diagnose the problem disabling quickboot option in bios, as this platform sometimes responded well to the better hardware/bios communication afforded by full boot. Unfortunately no effect.

[Peter M]

The problem might very well be a simple power supply problem - caused by a dying PSU, or by dying capacitors in the CPU supply circuitry on the board. Clogged heatsinks or bad fan motors also cause the kind of random boot failure and OS problems you describe.

Note that it also helps when you first figure out what's actually wrong and THEN badmouth the manufacturers - instead of just parrotting the "usual" opinions. Pisses people off, that.

[krazefinn]

Hey Peter, wondered when you might chime in with some ecs knowhow. Can assure it is neither failing nor unmaintained heatsinks. Not only do I blow and clean out case , sinks and such regularily, sleeve fans get the requisite drop of 3 in 1 oil. My first suspicion is either motherbard or power supply caps (preferably psu-its still under warranty, and is the 3rd psu in this rig). As mentioned, have carefully visually tried to ascertain bad caps, but have not yet opened psu (don't think I will break the sealtape) nor removed mobo from case for much better inspection. Was hoping to hear some more diagnostic tips from users who may have encountered a similar consistent 2 beep/8 beep code, failure to load drivers or install/change any pci card (actually, agp AND cnr affected also: its totally consistent)

Sure hope nobody takes a motherboard that personally esp if, quoting now, "ECS motherboard had its share of quirks ", is construed to be particularily " :D badmouthing"; or for that matter since shared opinion was formed via personal experience, :confused: is it neither "parrotting". Hey, if it offends ya, not too :rolleyes: big to say "sorry". Now back to the problem...

[Peter M]

Y'know, I'm a BIOS engineer, so I'm kind of a creator of mainboards (a totally different kind than desktop, though) - so please figure out what's wrong first then shout at "us" if, and only if, it was actually "our" mistake ;)

Regarding bad capacitors ... this usually affects only those in the CPU supply circuit. On the K7S5A(pro), that'll be inbetween the CPU and the I/O connector faceplate, nine tall capacitors. If these bulge (check undersides too) or leak, then there's your stability issues.

Eight beeps would be a VGA RAM test failure. The two beeps in front of it just mean that the (AMI) system BIOS is prompting the user for a decision - only that you don't get to see this.

With that kind of random component failure, I'd suspect a PSU problem too. First look always goes to fans, heatsinks and capacitors, though.

Oh, and DON'T open the PSU unless you really really know what you are doing. Lethal voltages inside, even after days of non-use!

[krazefinn]

Thanx for the id of potential badcaps. Cursory visual didnt detect any badcaps anywhere, but noiw that I know which ones to check more thoroughly. I'v also seen bad caps inside psu's, although am not sure if these involved the industrial espionage/incorrect formula and cheap taiwanese caps of late. Good words regarding PSU safety and NOT opening them casually. As in the case of TV's, or anywhere else a large curent is stored inside components, there are very specific grounding or stored current dissipating techniques to render them safe, as well as certain techniques (old one was one hand in pocket, only one hand working to avoid the very lethal cross-chest electrocution).

The amibios generally (at least on this board) indicates 2 shorts as low cmos battery (which for some reason this mobo often reports, even with brand new 2032). Can also elicit the 8 beps by slightly unseating the vga card, or not plugging the monitor plug all the way in. I didn't know of any specific signifigance when the codes consistently present together.

As soon as I get cleared by the doc to even turn a screw driver, bend under the desk, or lift more than 2 lbs, (which presently I cannot entertain trying) I will install a spare psu I have. I also realize these mobos often had some very specific power supply requirements, which led to wild early conjecture about about seemingly disparate, unrelated symptoms. Not exactly sure why this is the case also, but using a psu with sufficient tco on the 3 main rails, also using a good UPS, could circumvent many build issues early users reported at some forums. This unified chip platform always impressed me with its stability and performance, and I have used this one for 3 different build iterations, starting with sdram and an old 700 duron, to a tbird with ddr, and finally to the xp2200 running oc at 147/147 fsb/mem.
It sat about one year as I used a p4 as spare, and last year built an NF2/mobile OC to 3200+. Since I sold the spare P4 and the ATI AIWonder from the ecs, I tried to resurrect her as working spare. Weird aint it, when you keep 2 pc's always operational, you have few problems and seems nothing fails, but try to get by with only one PC and suddenly viruses, hardware failures, or software install/uninstalls cause all kinds of data loss, down time, and no email/web access!

Thanx again for the pointers, and sure cures an awful lot of cabin fever...but I can't wait to be up on my own 2 feet again. I know I've needed this spinal fusion surgery for 9 years now (very serious accident), but the doctors wanted to exhaust all conservative therapies first, and first two procedure were not very good as permanent cures. Still cant sit up more than a few minutes at a time, and still taking something like the equivalent of 76 or 80 percocettes DAILY. And that doesnt touch the neural pain..oh well, its GOT to get better, the way I see it...

[Peter M]

Two beeps on CMOS loss again is just the prompt notifier for the F1, F2 keypress decision.

K7S5A is known to lose CMOS at powerup with wimpy power supplies that suffer a brownout on +5Vstandby in the moment of firing up the main power rails. Seen it happen - totally reproducable on a certain kind of case/PSU combination, and just as totally cured by using a different PSU.

[rmanet]

Gawd it's a good thing you have time on your hands - sorry to hear about the operation.....you seem to have done everything but I searched too and educated guess would be to reflash (BIOS eg chb1001?), replace cmos battery, replace PSU - your posts are long but I think you're still using the same stick of RAM (?) - try another, and then I've had some video cards that under XP simply won't work with some mobos even though they're supposed too - maddening.

Sure trying to do some ambitious things with that new rig - may not be able to do it all? Good luck.

[krazefinn]

Thanx for the kind words rmanet, I have twice installed new batery, also flshed with latest official bios(just abit nervous as i wasnt sure that it would be stable enough to run through it, however with the ups, and totally stripped configuration, as well as removing the 769 memory (256 + 512) and use of only the known good 256 in slot1, I did flash successfully. Alas, to no avail...about the long posts, just trying to include as much relevant information as possible....thanx for bearing with me.

I see the doc on the 14th, when he'll either ease my restrictions alittle when I should be able to tinker with my hands alittle more, as well as ride/drive. Presently cannot even ride in a car except for short distance to MD/pharmacy and other necessary journeys. And believe you me, being in medical field myself and understanding the risks, potential complications and implications impacting rest of my life, I MUST heed hid advice 100%. Even stopped smoking 9NOW 21 DAYS WITH no relapse, JUST 2 PACKS OF 4MG NICORETTE GUM! Smoking implicated as cause of 4 times greater risk of developing pseudo-arthroses (basically non-union of bone fusion), and since I have no desire to repeat this surgery again....I HAD to quit. Will install new psu at that time if I can, and remove mobo from case and perform out-of-case-stripped boot.

Its funny, as I recall now, when I assebled this latest reincarnation of this venerable survivor, I had issues immediately, in that the new rig refused to boot or even go to bios. Seems there was some power supply/20 pin atx connector, or motherboard connector issue, which i was able to correct by removing/wiggling the atx connector repeteadly until it did boot. This problem manifested itself only once or twice subsequently, again necessitating removal/wiggling/reinsertion atc connector to mobo.

Just a WAG, but that could tie in to this, as the same occurred when I first tried to boot. Possible scenarios include short'/mobo standoff/board connector/psu wiring or connector. Replacemnt of psu will tell alot....

again appreciate all who have stuck through long/ tedious, complicated posts penned under the influence of vast quantities powerful CII narcotic analgaesics...its a wonder I can be even semi coherent considering....just can't wait to recover and get off all this poison....and get some semblance of my past life back.

[krazefinn]

more news, not necessarily good news. PSU tested ok at vendors, replaced into case. Removed mobo, examined all caps, absolutely no evidence of leakage, swelling, discoloration, etc. All power supply components on motherboard appear ok. Place one known good ram, replace motherboard. replace cmos battery yet again. No boot. Short board power pins, no boot. Check atx connector at board, when pushed on, keyboard lights flash, but no boot.

Good thing I have nothing BUT time on my hands at this point. WAG ythat mobo may be toast, whats the chance the xp2200 chip is gone? I still think it "may" be intermitent psu problem given symptoms, but at this point willing to entertain any guesses.

As any Finn will tell ya, never give up (SISU!), but if I have to, will give this thing the proper sendoff, cremation included! Fiurst got to confirm it though, so removing board as we speak, and will try stripped, out-of-case boot. If there happened to be a "board/case" short around standoff etc, would it present like this? How about boot without ram or proc?

Thanx to all guesses, just point me in some direction!

[krazefinn]

Bare out of case boot, nada. very strange. Retest psu on bench, checks out fine both with the 6-led pocket diagnostic as well as powering another board. Mobo also tested, and boots.

Purchased new Antec psu, I know the 350 w codegen that was in it last 2 years did not have the 185 eatts tco on the 3.3/5/12 volt rails, as well low on amps. Antec specs does exceed the finicky ecs requirements. Gonna try an out of case boot with the antec, but sure don't (yet) know why its not posting.

Its got to be some strange little gremlin...sure hope the xp2200 is not toast...I know it never ran hot. With custom cooling scheme, 100% loadtest did not exceed 44 celsius at an ambient 28 degrees. Ran at 32-33 degrees consistently with average use.

I suppose its possible that one of the other cables is defective, possibly ide or molex, anyones guess...or some other board short-although bare out of case boot "should" have picked that up.

[krazefinn]

Well, after having the mobo tested and post on bench at local builder/vendor, I've come to the conclusion it is the cpu. 3 yr warranty retail chip, unfortunately I gave the stock hs/fan away...DUH, needed for RMA.

Anyone with a spare hs/fan for xp2200 they'd like to get rid of? I cant think of anything else it might be...am going to test with my mobile 2500. Of all the things, never expected it to be the processor...and rather a strange way to fail. I still have sneaky suspicion its related to a weak psu...could be wrong though.