wombatman

I'm trying to put together a new computer here at work (engineering lab at U of South Carolina working on fuel cells), and it is being a righteous pain in the ass. I'm getting a beep code that consists of one long beep, a pause, one long beep, a pause, and so on. I called the Chaintech technical support (the mobo is a Chaintech S1689...Socket 939 Athlon 64) and was told that this meant there was a problem reading the memory. I've unplugged everything and tried both sticks of RAM (Corsair Value Series 2x512 dual channel) one at a time in each of the 4 slots for memory. Same beeping pattern every time.

Part Summary:
Chaintech S1689 - Socket 939 Athlon 64
GeForce MX4000 video card
2 x 512MB Corsair Value Series Dual Channel (and apparently recommended by Chaintech as working well with their boards)

Beep pattern = one long beep, pause, one long beep, pause, long beep, pause....ad infinitum

Any help is greatly appreciated.

God of Thunder

Sounds like memory to me as well.

Unplug everything else on the board and re-seat the processor.

Also, see if there is a jumper for memory speed. That may be a culprit.

__________________
I reject your reality, and substitute my own

-- Adam Savage

wombatman

Have literally everything unplugged from the motherboard (except for the needed power connectors) except for the processor 3-pin fan connection and the memory sticks. Still the same beep codes. Also, no jumper to determine memory speed. The manual says the motherboard can handle DDR266, 333, or 400. The memory being used is DDR400.

shakran

I'm guessing it's the mobo itself. Chaintech is, shall we say, not famous for quality.

wombatman

Had a thought. Is it possible that the Athlon 64 requires ECC memory? Or something else like that? I ask because I just tested the motherboard again this morning with some DDR333 from home that I know works and got the same beep code. This means that either the mobo isn't working, or that I need to use some other special type of memory.

shakran

the A64 does not require ECC memory nor should you use it unless you have a mission-critical application that you need 100% reliability for. ECC is slower since it spends time. .well .. error checking.

wombatman

Sorry to dredge up an older thread, but I'm still having troubles with this. I RMA'ed the original board and just got a new one back yesterday. Unfortunately, I'm still getting the exact same repetitive beep mentioned in the first post. I'm going to lunch soon and I plan to get some memory I know to be good and see if it is the memory that is bad. The basic problem is that the computer doesn't really start up. The fans all turn on and spin when I push the power button on the front, but nothing shows up on the monitor. I'm at a loss on this.

SiN

:/

I feel your pain, I build computers at my job, and the 'it runs, but there's no picture' problem is one I've oft dealt with. it can suck, tracking down the error.

things to consider:

you've an award bios, here's the beep codes
http://bioscentral.com/beepcodes/awardbeep.htm

and indeed sounds like a memory problem.

yes, test it with some good memory, that's usually the first thing I'd do in your situation.

you say it turn/stays on ... do you have to hold the power button for a few seconds to get it to turn off?

and, did you reseat the proc? ... just because, it's not a bad idea to try that too if you're stumped.

__________________

wombatman

Ok, just got done trying it with the memory that I know should be working (Crucial DDR333 I brought from home--used it for probably 2 years with no problems). Tried both sticks of that in all 4 slots on the motherboard. Same repetitive beeping. The beeping would seem to indicate a memory problem, but I've double-checked the seating of it and it's in there very well.

One other suggestion I found was to flash the bios to an updated version. However, doesn't this imply that you can actually get to a BIOS screen? If not, then perhaps I can try that. And as I said before, this is supposed to be a new working board after RMA'ing the last one, which was doing the exact same thing.

When I talk about turning it on, I hit the power button on the front of the case, and the back 120mm fan and the processor fan both start up (and stay running). I then have to either hold the power button on the front for a few seconds or cut the power in the back to turn it off when the beeping starts. The processor was reseated in the new motherboard and is locked down tightly with the retail HSF, so I don't think that's the problem. Also, I'd really like to not have to take out the processor again, because when I did it the first time, the HSF wouldn't separate, and the whole thing (HSF and processor) came out at once, resulting in a bent pin. I was able to straighten the pin such that it slipped back in on the new board, but I'd rather not mess with that again if possible.

Thanks for the help.

wombatman

Some more stuff I've tried:

cleared the CMOS
unplugged the front USB (read about USB stuff maybe causing a problem)
made sure the motherboard was screwed tightly onto the risers

Interesting note (at least to me) was that I turned the computer on while the CMOS jumper was set to the "Clear CMOS" setting (on pins 2 and 3 vs 1 and 2 for normal operation) and it turned on, with fans running and nothing on the monitor, but without any beeping. Any significance to that?

wombatman

SiN, anybody? I really don't have any idea what to do about this other than RMA for a refund and get a different board. But that would need to be a last resort since I've been working on this for probably going on 3 weeks (had trouble procuring a 24-female to 20-male converter).

SiN

shit, dude ...I've been hoping someone else would come along, smarter than I.

in your situ i'd just switch out all parts with another working computer of as close to the same hw setup .. but you probably don't have that luxury.

I'm still as stumped as you are, unfortunately

__________________

cyrnel

I'm hesitant about jumping into a well-oiled discussion.

SiN's right about isolating. Put the parts next to a working box and carefully swap junk until you find the culprit.

Specifics that come to mind, simple stuff first:

- Standard (non-USB) keyboard & mouse attached? I've seen too many low-end boards that didn't seem to include headless operation in their test suites.

- Different video card?

- The 24->20pin power supply adapter is a failure point. Do you have another PC you can sit on the bench to confirm the supply/adapter isn't a factor?

- Don't worry about the fans running during your lights-on-nobody-home tests. It doesn't point to anything specific beyond that the basics of your p/s are working. Also, don't let it just run with the clear CMOS jumper in position. Depending on mb design, continual power with the jumper can harm some components. When clearing, use the minimum jumper duration required to restore the POST.

- Don't worry about flashing the BIOS. You're right, it implies getting further than you have.

- The "bent pin" processor experience sounds ominous. Could be terrible or not a big deal. Can't know from here. If you're confident then don't mess with it until you're out of options. Still, given the number of other things you've tried, it would be haunting me for a test in a known-good box.

- Have you tried with the motherboard removed from the chassis? (on a non-conductive of course)

BTW, you're discharging the p/s before installing or removing components, yes? It's almost difficult to hurt things this way these days but I hate to see someone repeatedly blow away parts.

Crap, I just had a copy/paste moment. Think it's all back now...

From this point I resort to my remote-trouble-shooting voodoo dance. Not pretty.

__________________
There are a vast number of people who are uninformed and heavily propagandized, but fundamentally decent. The propaganda that inundates them is effective when unchallenged, but much of it goes only skin deep. If they can be brought to raise questions and apply their decent instincts and basic intelligence, many people quickly escape the confines of the doctrinal system and are willing to do something to help others who are really suffering and oppressed." -Manufacturing Consent: Noam Chomsky and the Media, p. 195

FloydianOne

You must be a patient man, I would have definately returned this board and got a more reliable (read: more expensive) board by now.

__________________
Fight apathy! ..... or dont.

wombatman

Let's go down your list Cyrnel:

-Standard keyboard and mouse attached (USB to PS2 adapter for the mouse)

-Don't really have a spare video card lying around, and there's no onboard video for this mobo. I'll try the card in a different system to make sure it's working correctly.

-I have another 24->20 pin adapter currently in use in a different machine that is operating perfectly. Also, when I cut off the power supply (via the switch in the back), I hear a distinct sound that I believe is the motherboard losing it's charge. No LED on the motherboard to confirm this, but I definitely hear the "I've lost all power" noise.

-Good to know about the CMOS thing.

-Also good to know about the BIOS, although if I ever get it running, I'll probably update the BIOS anyway.

-The pin was not bent to start with, and I was having the same problem with the original board, so I do not believe that it is a problem. As I said, I was able to straighten the pin such that processor slipped right into its place as it is supposed to. Also, some genius (not me) threw away the boxes that all the parts came in, so I will be a little harder pressed to be able to return it.

-I have not tried removing the motherboard onto something like cardboard, although I have considered it. I was mainly worried about cables and cords being able to reach far enough. Any suggestions on a best way to do this?

-When I've been doing stuff like changing the memory into different slots, I've not been discharging the motherboard (that is, cutting the power to the power supply and listening for the noise mentioned above). I have been doing the discharging when moving major pieces around, especially with stuff like the CMOS jumper, pulling switches, that sort of thing.

Sadly, I don't have a box with which I can switch parts as the other one we got is up and running a fuel cell system (hooray for researching on the cutting edge! Go U. of South Carolina's National Fuel Cell Research Center!)

Thanks to both you and SiN for your suggestions and help so far. Hope this response to your list helps maybe narrow this stuff down. I won't be able to work on this computer until Monday when I go back to work, but feel free to throw out any other ideas if you have them.

Lasereth

I would have shitcanned the SOB by now. I'd order a completely new motherboard. I feel your pain though...computers can be incredibly annoying to put together if something like this happens. I just thought of something:

I think it could be your processor. I fixed a PC last summer that was doing the exact same thing...I put my extra processor in it and it fixed it immediately. That's how I fix all of my computers...I trade out every single damned part in the system until it works, even if it means trying another motherboard. You said you've tried two motherboards and two RAM sticks and it still does it. That points to another component, and a PC not even getting into BIOS is a big symptom of the processor being fucked. That's what I'd try next. Not sure how you can try out another processor though; it sounds like you don't have many resources to work with.

-Lasereth

__________________
"A Darwinian attacks his theory, seeking to find flaws. An ID believer defends his theory, seeking to conceal flaws." -Roger Ebert

wombatman

Considering we still actively use Pentium 200Mhz CPUs on the fuel cell systems, yeah, a lack of resources would be the understatement of the year. As I mentioned before, though, I think the packaging for the processor/HSF was thrown (despite my best attempts to hang onto it all). The processor was ordered from newegg. Does anyone know how or if I could still RMA this thing? If we had another board, I would test the processor there, but as far as I know, this beast is the first Socket 939 (and most updated processor) in our group.

cyrnel

Newegg has a tight return policy for cpu's. I think it's 7days.

Find someone with a known good 939 board to test your cpu. A small computer shop, consultant, whatever. Fuel cell acumen is worth lots in the world of geek information barter.

First though, I'd detach the MB from the chassis & see what happens. Shorts are usually deader than the beeps you describe but it's worth a try. Unscrew, lift, and insulate from the chassis. Something like cereal box cardboard would work for a quick test.

If that doesn't get you anywhere, try AMD's warranty dept. A smooth operator might work out eval A64s from a geek customer service rep once they know your predicament.

__________________
There are a vast number of people who are uninformed and heavily propagandized, but fundamentally decent. The propaganda that inundates them is effective when unchallenged, but much of it goes only skin deep. If they can be brought to raise questions and apply their decent instincts and basic intelligence, many people quickly escape the confines of the doctrinal system and are willing to do something to help others who are really suffering and oppressed." -Manufacturing Consent: Noam Chomsky and the Media, p. 195

wombatman

Thanks for the advice, Cyrnel. There's some IT guys in the building I'm in, so I'll go check with them and see what they have in the way of Socket 939 boards. If no go there, I'll call up AMD. Any particular advice on how I might secure a test A64 or two? Just wondering if there are particular buttons I should look to be pushing to improve my chances with them

cyrnel

Service might be the simplest method. Should they be pitiless bastards then go for freebies.

I can't find much on eval units on their site. Used to be prominent when they were gunning for mindshare. Start with customer relations.

Buttons:
- Namedropping. Aforementioned fuel cell tidbits. It's worth points.
- "Really want to go with AMD."
- University budget constraints. Boo hoo.

Might you add them as a sponsor? Probably something for your project lead. A birdie told me project websites are a great place to solicit and support these relationships, even if it's just sponsor logo placement.

Alternately, try <a href="http://www2.amd.com/us-en/RegForms/1,,51_52_10781_12820,00.html?1117562558">their official sponsorship route.</a>

__________________
There are a vast number of people who are uninformed and heavily propagandized, but fundamentally decent. The propaganda that inundates them is effective when unchallenged, but much of it goes only skin deep. If they can be brought to raise questions and apply their decent instincts and basic intelligence, many people quickly escape the confines of the doctrinal system and are willing to do something to help others who are really suffering and oppressed." -Manufacturing Consent: Noam Chomsky and the Media, p. 195

wombatman

Well, it's on the way to them at AMD's place, to be evaluated, and I'm looking at about one and a half to 2 weeks or so. The conversation was had on Monday, and I don't really recall if I mentioned the fuel cell bits or not, but I did mention the University at some point. This was, however, just a customer service rep.

Now, how are AMDs for hard scientific work? Possibly graphical, but especially for modelling work? I ask because we also do modelling work related to fuel cell (I'm not involved directly, so I'm short on details), but do you think I could finagle that into some sort of deal with AMD to get some evals for actual evaluation and use? I know we have a small cluster at the moment that does the modelling, but it'd be pretty awesome to build another/add on to the current one with some newer stuff. If this is a possibility, am I still talking to customer relations?

wombatman

So I just got the RMA'ed processor (or at least a processor) back from AMD, and I'm still getting the exact same problem. I've now changed out the motherboard, CPU, and RAM. Presumably the processor is fine as I got it direct from AMD. I previously brought memory from home I know to work fine. How likely is it that the RMA'd motherboard is non-functioning? This is frustrating as hell.

cyrnel

Did you ever have a chance to try that video card in another box? Another video card in this box? It's highly unlikely you're dealing with a MB/CPU problem unless you have a bad habit we haven't picked up on.

BTW, by discharging the PSU I meant switching it off and then pressing the power button. (do you hear anything or see fans spin up briefly?) Designs vary, but PSU caps can hold significant power on the bus even after being switched off, potentially affecting anything you plug in.

Try another PSU yet?

Anyway, good deal on getting another processor! AMD's are very competitive. Intel puts more into helping vendors tune products to perform best on their processors but AMD generally splits the winnings. Any optimization means knowing the specific platform and exploiting advantages relative to their worth for your application. I'm pretty stale on assembly tuning, and am way out of date with the current cores. Check around for comparisons. Google should have plenty buried in the chaffe.

Can't speak to eval stuff or sponsorship now. Always a sales job, which I dislike. I'm only good at it if it hinges on something around tech with fellow geeks. Snow jobs put me to sleep.

Getting ready for a couple days at Sun River. Pack, pack, pack...

__________________
There are a vast number of people who are uninformed and heavily propagandized, but fundamentally decent. The propaganda that inundates them is effective when unchallenged, but much of it goes only skin deep. If they can be brought to raise questions and apply their decent instincts and basic intelligence, many people quickly escape the confines of the doctrinal system and are willing to do something to help others who are really suffering and oppressed." -Manufacturing Consent: Noam Chomsky and the Media, p. 195

pacaveli

can u get ur cash back on the chaintech board and just buy a new one from another vender, because at this point that is what i would do. you've tryed almost everything that u can do, replaced nearly every part, and your still having problems i would.


maybe the memory was ran across a magnet or durring shipping at one point it was run across something that could damage it's quality, thats all i could really think of....

__________________
Check my blog out. Basically me talking about video games

http://gginsights.blogspot.com/

CinimoD

Hey guys,

I just recently bought a new kit on ncix.com (AMD64 3000+ with 2x512MB Corsair Value Series Dual Channel and a Chaintech S1689 mobo), installed all the component and windows XP. Everything was going fine until I decided to flash my bios! (What a bad idea!!)

I flashed my bios with WinFlash and the latest bios version (7.0), but now when I reboot I got the same beep problem than wombatman!! I really don't know what happened, the bios has been flashed currently with no error message in Winflash...

Any idea?

wombatman

I guess I should have put a resolution on this beast. Actually ended up RMAing the next Chaintech board and buying an Asus, which worked on the first boot-up. CinimoD, I guess you could try flashing again. And double-check to make sure the bios you're trying to upgrade to is the correct one. And, you might try to find an original bios (the version the board shipped with) and try to flash to that since you know it works.

CinimoD

I Cleared the CMOS by using the JP1 jumper( as instructed in the manual
step by step)