ECS K7S5A 133mhz FSB unstable?

[MeloManiac]

Back with more ECS trouble! OK so I thought I had found the problem being a 512MB memory module not liked by the board. Maybe it still has something to do with it, since I had to put it back in (this time in slot 2 so it gets used later) because of a project I am working on. So I worked for 1 hour or so and freeeeeeeeeeeeeze. I then rebooted, worked for 15 seconds. Freeeeeeeeeze. OK so maybe it's not the RAM. Then I remember someone mentioning trouble running 133mhz FSB on the board. I changed to 100mhz, and since then I ran my project in loop for a couple of hours without any trouble.

I will still buy some DDR RAM to change this SDRAM with ASAP, but I wonder if anyone else had to run 100Mhz FSB on a rev 3.1 of this board (which mine is)? My CPU temp is 53 degrees, so it shouldnt be the reason.

 

[osage]

a bit more info on your system if you could.....cpu,are you overclocking etc:

temp is a bit on the high side, but within spec.

 

[novalogic88]

thats temp is too high,

1. replace your HSF for the cpu

2. Leave the side case open

3. Replace the chipset heatsink with some thermal grease


i just leave my side case open, its a little noisier but its not too bad, i have a volcano 6 cu (not plus), it keeps my 1.4 athlon C (133fsb) at aroudn about 45 under load.

 

[MeloManiac]

Damnit I have a Silverado which is supposed to be good!! 53 shouldnt be a problem - dont AMD say 60 or EIGHTY? My system temp is 32 degrees, so it shouldnt be necessary to touch the chipset sink - or will that be hotter when I run 133mhz FSB compared to 100mhz too?

 

[AA0]

that temp is fine, as long as it doesn't get higher. Not the temp which makes that board unstable.

 

[osage]

what cpu are you running ?

 

[DHL]

How about the PSU?

I once had a similar problem with my k7v not being stable when the memory bus was set to 133, (.25 Athlon = power hungry) and I changed the power supply and it run fine

Still, more info on ur system will definitely help

DHL

 

[Buz2b]

It could well be your SDRAM also. What brand and type is it? Possibly it is heating up. Either way, it sounds like some sort of heat issue. Especially since it froze again shortly after rebooting; no time to cool down. Try leaving the case open with a small fan directed towards it and see if that helps. It's a simple and effective test.

 

[MeloManiac]

I ran this system without any instability problems on an MSI K7T Pro2a motherboard and a TBird 1ghz before. Same PSU, same RAM, no problems. I have two SDR Ram modules, one Apacer 256mb, one noname 512mb (which I suspect is one of the sources of trouble..). 300w 'lownoise' PSU (pretty noisy though). Now the CPU is an Athlon XP 1600 (1400 mhz).

 

[MeloManiac]

OK I tried Motherboard Monitor 5... it gives me somewhat higher readings than the BIOS after I had the crash. I also assume the temp can go down quite rapidly after a crash. The MBM indicates variable temperatures from high 50's pushing more and more into the 70's. Crashed at 71, but had been up to 76.

I don't know how accurate those readings are though. But temperature was clearly rising, starting out at high 40's.

I do really harddrive intensive work on the computer, and the HD's ARE placed very close to the SD RAM on this motherboard - could that be a contributing factor? CPU stress is only around 20-50%, whereas my harddrives are pushed to the max much of the time.

 

[MeloManiac]

... interesting, when I run it at 100mhz FSB (1050mhz) I dont get temp readings above 50. How can there be more than 25 degrees C in difference from 1050 mhz to 1400mhz?????!!

 

[RaynorWolfcastle]

Definetly reinstall the heatsink, those temps are waaaaayyy too high. I'm running at ~38C idle ~45C load for my 1600+ with a retail heatsink on an ECS K7S5A

-Ice

 

[Buz2b]

<< I ran this system without any instability problems on an MSI K7T Pro2a motherboard and a TBird 1ghz before. >>

I'm sorry but that is an example of the infamous "famous last words". The ECS board is a completely different animal. For that matter any two boards from differring manufacturers will have separate requirements and quirks about them; what works on one may have trouble on another. However..........
Your temps are way too high. I'd look strongly at the HSF and it's placement and/or abilities. Start by removing, cleaning and reapplying it. Make sure to use some good AS2 or comparable paste and double check the position after you install it. Make sure it is not "tipped" on an angle over the die. Yes, the HDD's can effect the temps of the RAM when positioned close by. Any modifications you could do there would help. Any way you can move them around a bit or at least move one a bit further away. You might also want to try running just the 512 MB stick by itself and then the 256 MB stick by itself and see if there is any difference as far as freeze ups.
No.1 though is get that CPU temp down though. An Athlon can theoretically take up to about 90C, but I wouldn't want to try it. You are pushing the envelope there. You need to get below the 55-57C range at all times, even under load.

 

[MeloManiac]

So, based on the fact that i get 70+C from 1400mhz and only 50C from 1050mhz, you wouldn't say that the AMD CPU is faulty and I should take it back to get a new one?

 

[wasamicron]

Novalogic 88...u are right!!!! i just had same prob. with my shuttle....took off heatsink, added thermal grease, slapped it back on, and all is well. my temps were around 53 before (just checked..Sandra says 38)...oh, how sweet!!

 

[Buz2b]

<< So, based on the fact that i get 70+C from 1400mhz and only 50C from 1050mhz, you wouldn't say that the AMD CPU is faulty and I should take it back to get a new one? >>


Correct, with qualifications. First we are assuming that this is indeed a 1.4 GHz CPU, not a 1.0 GHz that is being overclocked. Second, a better way to view this is that you need to reapply the HSF again very carefully to see if that will bring the temps back down to something reasonable. If it doesn't, and again, assuming that your HS is in reasonable shape and you have a good strong fan installed on it, then OK; it is possible that your CPU is faulty. If you are worried that this might be the case and you still have warranty left, then it would be to your advantage to eliminate the HSF and/or the installation of the HSF as the culprit. The Silverado is a decent HS but you really need a strong fan on it for a 1.4.

 

[MeloManiac]

Thanks for your help. The Silverado is running at 2x2500 rpm, that should do the trick. But I'll try another cooler on it. I did apply thermal grease, but I noticed it looked rather odd and full of 'cracks' on the CPU core after I had the Silverado just sitting there without mounting it properly. The grease might be too old or something. How do you get old grease off a CPU. You cant use acetone can you?

 

[Boisean]

One possible cause for unstable 133/133 1.4gHz Athlon Thunderbird behavior with the EliteGroup K7S5A is a subtle change made by ECS in anticipation of XP processors, i.e. the so-called "resistor" issue documented here:

http://prohardver.index.hu/download/ecs/k7s5a_corruption_fix.pdf

The rev. 4 version of the K7S%A is the most notorious for unstable 133/133 operation with 1.4gHz Thunderbirds.

But it has been noted on other revisions as well.

Basically three options are available:

1. upgrade to an XP processor (which I've done and it completely fixed the instability at 133/133)
2. RMA the board (available for one year directly through ECS)
3. perform the solder fix documented in the above PDF file

Rev. 1 and 2 versions of this board had no trouble with the 1.4ghz Thunderbirds.

But the resistor problem crops up quite often in later revisions.

Here's another Web site with information:

http://www.geocities.com/mrathlon2000/zp.html

I should add this is a great board when it's matched to the correct processor.

I'm running the LAN version with the latest bios with an AMD Athlon XP2000+ and it's ROCK SOLID STABLE. (I could not get stability at 133/133 with my Athon Thunderbird 1.4.)

Good luck,

Jerry Jones

 

[Dually]

I know two people with that board that have 133MHz just fine. Make sure the ram is seated right, latest bios and the right bios memory settings and 133MHz ram.

 

[MeloManiac]

Thanks Jerry, but I AM using an XP chip... I wrote that earlier in the thread, an XP 1400mhz (1600), so that cant be the issue.

 

[Buz2b]

<< How do you get old grease off a CPU. You cant use acetone can you? >>

Use Isopropyl alcohol and a lint-free cloth if possible to remove the grease. I wouldn't use the Acetone.