K7S5A Rev. 4 Thunderbird 1.4 133/133 Issues
Jerry Jones
1.4gHz AMD Athlon processors at 133/133 as explained by the following
PDF file.
http://prohardver.index.hu/download/ecs/k7s5a_corruption_fix.pdf
1. Rev. 4 and higher boards do not correctly program the AMD bus
driver impedance on the fastest Thunderbird CPUs.
(ECS - according to the PDF above - has changed the rev. 4 boards and
higher to program 60 oHms to the CPU while the AMD revision guide from
April 2001 calls for 40 oHms. Athlon MP and Athlon XP processors can
work at either 40 or 60 at 133/133 while the Thunderbirds cannot.)
2. The problem has been *fixed* on the CBMB04 boards sold at
http://www.samsclub.com or http://www.cpubuilders.com
3. The problem does not affect AMD Athlon MP/XP processors
4. Rev. 4 and higher boards can be "fixed" to work with Athlon 1.4s,
according to the PDF above, by way of a complicated soldering
modification (which I have no intention of trying)
Personal experience:
1. Flashed to the latest bios
2. EG465P-VE power supply (431 watts)
http://www.enermax.com.tw/products/eg-365p-ve.htm#s
3. Corsair PC2100 DDR
4. Windows 2000 SP2 + all sound/video driver updates
5. 1.4gHz Athlon Thunderbird
No problems with cold boots or any other booting issue.
Computer seems to run fine at 133/133 until I attempt to encode large
.avi files to MPEG-2 video file format using Ligos encoder.
First thought I had fixed the problem by setting memory timing to
"SAFE" instead of "NORMAL" as I was able to encode more than 2 hours
without a system restart, but later the system restarted when I
attempt to encode a larger file.
Succeeded in encoding more than four hours of video (twice) at the
100/100 setting. No restarts whatsoever. Completely stable operation.
So the problem is at the 133/133 setting.
Now buying a replacement processor: XP2000+.
Will report back on results.
Shep©
jones...@netzero.net (Jerry Jones) wrote:
>It seems Rev. 4 and higher boards do, indeed, have troubles with
>1.4gHz AMD Athlon processors at 133/133 as explained by the following
>PDF file.
As previously stated by ME many times.Don't need a PDF to tell the
truth |-)
Pablo
Probably way off the original thread, but I have not figured out how to
set the markers in my palm for newgroups yet. Just picked up another K7S5A,
this one with the number 1 in the corner....all previous that I have had
were number 4. What a diff....easiest and most stable by far. Tolerated a
DDR that the other boards didnt like, had not hiccup'ed once...but here is
the funniest thing. All the four series boards averaged memory thru scores
of like low 1200's, with Ultra settings. This board, with a 1.1G and a
generic 2100DDR running 100/100 and the same settings is doing close to
1400. It has no probs with my 1.1, or the 1.1 that I modded to change the
multiplier on. The modded CPU wouldnt even start in the series 4 boards.
This one came with 12/15/2001 bios loaded, and is just all around faster. Is
the series one board better for some reason? I called and checked, and I can
get more from the same place...they have a few versions, but they checked
and had a number of the #1 boards on hand. Sure has been different for me.
"ShepŠ" <sh...@gmx.co.uk> wrote in message
news:o8ui7u0oi4pgk7uve...@4ax.com...
Shep©
cube "Pablo" <goat...@trough.com> wrote:
>Hey Shep.
> Probably way off the original thread, but I have not figured out how to
>set the markers in my palm for newgroups yet. Just picked up another K7S5A,
>this one with the number 1 in the corner....all previous that I have had
>were number 4. What a diff....easiest and most stable by far. Tolerated a
>DDR that the other boards didnt like, had not hiccup'ed once...but here is
>the funniest thing. All the four series boards averaged memory thru scores
>of like low 1200's, with Ultra settings. This board, with a 1.1G and a
>generic 2100DDR running 100/100 and the same settings is doing close to
>1400. It has no probs with my 1.1, or the 1.1 that I modded to change the
>multiplier on. The modded CPU wouldnt even start in the series 4 boards.
>This one came with 12/15/2001 bios loaded, and is just all around faster. Is
>the series one board better for some reason? I called and checked, and I can
>get more from the same place...they have a few versions, but they checked
>and had a number of the #1 boards on hand. Sure has been different for me.
All the KT7S5A boards are OEM
They are not meant for individual public consumption<grin>
They have been designed for people who run a business
building a reasonable quality low priced system to sell to the public
and the builders know their limitations and work within that spec.
They have not been designed for people to over clock or use older
hardware devices.
I got mine for 50 quid which is way below what I paid for my Hyped
Abit KT7 that died after 14 months and was not worth the money/drivers
or hassle.
If you bought your board separate then you are just like the rest of
us on here.Find a fix or return it and by the time you do the OEMS
will have built another thousand or so with no problem :)
I know this is not much help,but it's the truth :/
PS
Best 50 quid I've spent in a while :D
Martin Eriksson
news:33a73c5c.02022...@posting.google.com...
> It seems Rev. 4 and higher boards do, indeed, have troubles with
> 1.4gHz AMD Athlon processors at 133/133 as explained by the following
> PDF file.
>
> http://prohardver.index.hu/download/ecs/k7s5a_corruption_fix.pdf
>
Yup.. just fixed one for a buddy. unsoldered two 150 ohm resistors from a
scrap board I had and put them over those two in the PDF. His board works
like a charm now.
--
_____________________________________________________
| Martin Eriksson <nit...@giron.wox.org>
| MSc CSE student, department of Computing Science
| Umeå University, Sweden
- MSI K7TP2A(3.1) (KT133 + 686B) - Athlon(B) 1GHz
- 256MB/PC133/C3 Hyundai - 256MB/PC133/C3 Apacer
- IBM 46/7200/U100 P/M - IBM 10/5400/U33 S/M - LG CDROM S/S
- CLGF2 GTS @ 200/350 - SB512 - RTL8139 - USB Mouse - PS2 KB
- Win 2k ProSP2 - Det 2311 - 4in1 437 - AMD patch
- BIOS: Optimal Settings, No sound
- ABIT BP6(RU) - 2xCeleron 400 - 128MB/PC100/C2 Acer
- Maxtor 10/5400/U33 HPT P/M - Seagate 6/5400/DMA2 HPT S/M
- 2xDE-530TX - 1xTulip - Linux 2.4.17+ide+preempt
ecsboard
somewhere a few mths ago?
"Jerry Jones" <jones...@netzero.net> wrote in message
news:33a73c5c.02022...@posting.google.com...
Jerry Jones
> Don't need a PDF to tell the truth |-)
The PDF is needed for those who wish to attempt a repair.
Too bad Elite Group hasn't acknowledged the problem, publicly.
That would be a responsible business practice.
This board - in the spite of the problems with T-Birds - is a very
good value overall. ECS does nobody any favors by ignoring a valid
tech support issue.
For me, it *seems* to run very stable even at 133/133 using a 1.4
T-Bird... until I do long-term MPEG video encoding.
Had it not been for the complete information in the PDF, I might have
been tempted to replace the Corsair ram stick with another brand at
additional expense.
One thing is certain, upgrading the latest bios does not fix the
problem as the PDF above confirms.
Jerry Jones
Kyle Brant
Bring on the lawsuit, as a lawyer, I'll be more than glad to make a
complete expose of the entire k7s5a data corruption scenario. I
documented my test findings with great care, and can readily revert my
board back to "factory" form by removing the single resistor tacked
onto my Zp resistor for the demonstrations to a trier of fact of the
data corruption problem.
--
Best regards,
Kyle
cut.s.p.a.m.m.e.n.o.t. from email address to reply
"ecsboard" <ecsb...@nospam.hotmail.com> wrote in message
news:a5c7g1$5i2$1...@violet.singnet.com.sg...
TMack
"ecsboard" <ecsb...@nospam.hotmail.com> wrote in message
news:a5c7g1$5i2$1...@violet.singnet.com.sg...
> somewhere a few mths ago?
I don't think so - unless they have been listening to people's
conversations. Slander is a spoken defamation. They could try suing for
libel - but then they would have to prove that what has been posted in
newsgroups and on the web is untrue - and that is where the case would fall
apart........
Tony
Rich Webb
wrote:
[snip...snip...]
>4. Rev. 4 and higher boards can be "fixed" to work with Athlon 1.4s,
>according to the PDF above, by way of a complicated soldering
>modification (which I have no intention of trying)
Complicated? Ahhh, it consists of just two teeny dots of solder. If
you're hand-working surface mount components all day, you'd probably
want (and already have) the microscopes, paste bench, and the rest of
the kit. That does make it seem complicated (and expensive).
However, for a one-time shot just some solder paste, masking tape, a
fine-tipped iron, and a little time and patience is quite sufficient.
It's probably easier than bridging an XP.
--
Rich Webb Norfolk, VA
Rich Webb
<ecsb...@nospam.hotmail.com> wrote:
>ECS replied with the right to sue for slander.. did you see that posting
>somewhere a few mths ago?
Why oh why is it encouraged to make modifications to CPU's, mainboards,
BIOS settings, and all the other components in the quest for speed --
that is, after all, the whole reason for the existence of the
"Overclocker's Workbench" isn't it? -- but a simple modification for the
sake of stability brings out threats of lawsuits?
Are you making this representation on the behalf of Elitegroup, as your
posting alias implies?
Martin Eriksson
news:hfml7usu7i4fghduj...@4ax.com...
The best you can do is to killfile this guy... he is totally ignorant of
facts and he propably does not represent anything "ECS".
I got much less aggrevated when I did this.
JJones
to do it because I've never worked with
soldering irons of any kind before and I'm
afraid I'll make a mistake.
I suppose you're right... for people who've
worked with soldering irons it isn't that
complicated.
...anyway...
I'm buying an XP processor, the 2000+, to
see if that'll work for me instead.
Supposedly the Athlon XPs are - for the most part -
immune to the rev. 4 and above problems.
I'm quite disappointed in EliteGroup for not
issuing a "recall" on these boards.
My 1.4gHz Thunderbird *seemed* to be running
fine for two months at the 133/133 setting.
It wasn't until I attempted to encode .avi video
to MPEG video that my computer would self-boot...
...and that's when I searched the Net and discovered
the problem between rev. 4 boards and higher when
combined with Athlon Thunderbird 1.4s.
In my case, the RMA option expired about two months ago.
So - for me - it never was a *true* option.
Jerry
Rich Webb
>Well, I suppose I just don't have confidence
>to do it because I've never worked with
>soldering irons of any kind before and I'm
>afraid I'll make a mistake.
Heh ... Fair enough. I'd describe my own first attempt with soldering
electronic components but I'm afraid of the effects on innocent eyes.
Suffice it to say that the only soldering experience I'd had to that
point was in sweating copper water pipes, and the tool I used was a
rather high wattage soldering gun. Not a pretty sight...
You might find this link interesting. I'm afraid I've forgotten where I
originally saw it posted, so I can't give proper credit to the
discoverer. http://www.soltek.de/others/Athlonxp.htm
Kyle Brant
required on the iron to solder the SMD components, and a decent
magnifying glass is also needed (or it could be I'm getting old and
need glasses, naw, that can't be the problem). Finally, a good set of
tweezers helps a bunch, and needle nose pliers may be a bit too big.
I've plenty of soldering experience over the years, and it took me
more time than I thought it should to accomplish the job. Best tip I
can offer is to add a bit of solder to the top of the existing
resistor, then tack another resistor on top of the existing Zp
resistor.
--
Best regards,
Kyle
cut.s.p.a.m.m.e.n.o.t. from email address to reply
"JJones" <jfj...@cityofboise.org> wrote in message
news:cd98bbfd.02022...@posting.google.com...
Jerry Jones
Errors didn't appear until test #5 when running at 133/133.
(512mb of Corsair PC2100 DDR).
I've checked and double-checked everything else...
a. 431 watt power suppply
b. processor runs cool with Swiftech heatsink/Sanyo Denki fan
c. replaced chipset heatsink thermal tape with Arctic Silver
rev. 4 - yes
XP sticker - yes
I just hope the XP2000+ passes the test at 133/133.
If not, I doubt I'll bother with the board any longer...
Jerry Jones
Kyle Brant
seen 2 boards that only had errors in test 5 with a 1.4g Tbird,
lowering Zp below 48 ohms fixes the problem.
--
Best regards,
Kyle
cut.s.p.a.m.m.e.n.o.t. from email address to reply
"Jerry Jones" <jones...@netzero.net> wrote in message
news:33a73c5c.02022...@posting.google.com...
Jerry Jones
My Athlon 1.4gHz Thunderbird refused to be stable on the board at
133/133.
The 1.4gHz Thunderbird does, indeed, have issues with the rev. 4
boards, in my opinion.
Thanks to "Mr. Athlon"...
www.geocities.com/mrathlon2000/zp.html
...who discovered the "Zp, Zn" issue with rev. 4 boards...
Thanks to him, I learned my true options were to either:
1. perform a solder fix on the board so that it would work with the
1.4gHz Thunderbird: www.geocities.com/mrathlo...ix_2.1.pdf
or
2. simply upgrade the processor to an XP.
I did so tonight.
I am delighted to report Mr. Athlon's suggestion has worked
beautifully.
The XP2000+ is running happily at 133/133.
I've run the Memtest - 86 with none of the errors that showed up with
the 1.4gHz Athlon Thunderbird.
For details, about Mr. Athlon's findings, visit this forum:
pub65.ezboard.com/bk7s5amotherboardforum
...better yet...
...read this:
www.geocities.com/mrathlon2000/
I'm very, very glad I did *NOT* follow the so-called "advice" given on
the heavily-censored OC Workbench forum where any mention of Mr.
Athlon and his fix is brutally removed from public review.
Had I listened to the prevailing view of the OC Workbench crowd, I
would have wasted in the neighborhood of $250 as the prevailing
opinion on that forum is that Athlon 1.4gHz problems are mostly caused
by inadequate power supplies or bad ram. Even though I explained I was
using an Enermax 431 watt power supply and my CPU temperatures were
low (Swiftech heatsink) and I was using high quality Corsair DDR
PC2100 memory, the individuals on the OC WorkBench forum kept
suggesting that I had either a "heat" problem or a "power" problem.
I'm glad I did not listen to them. Instead, I listened to Mr. Athlon
and saved $250 and I have a rock-solid-stable PC.
To the folks at OC Workbench: Due to your practice of censoring posts
related to this fix, I do not plan to ever visit your Web site again
unless you make a sincere and contrite public apology.
Meanwhile, my heartfelt thanks to Mr. Athlon. My K7S5A is performing
flawlessly now thanks to your advice.
Jerry Jones
JJones
Using my new AMD Athlon XP2000+, I encoded THREE HOURS of MPEG-2 video
from a DV .avi source file WITHOUT SINGLE RESTART.
Formerly, I was using the AMD Athon 1.4gHz Thunderbird and the system
would restart during the MPEG-2 video encoding, which gave me the
first hint that something was unstable.
So - again - simply switching from a Thunderbird to an XP processor
has made all the difference in the world in my 133/133 stability using
this Rev. 4 K7S5A.
Also, I changed my bios setting to "ultra" instead of normal and ran
the Mem-test 86.
Again - no errors. ROCK SOLID STABLE.
Thanks again to Mr. Athlon for discovering this issue with 1.4gHz
T-Birds and the rev. 4 boards.
Jerry Jones
P.S. I noticed the link to his PDF file that I posted doesn't seem to
work so here's another link:
http://prohardver.index.hu/download/ecs/k7s5a_corruption_fix.pdf
Allen Crider
>
> Also, I changed my bios setting to "ultra" instead of normal and ran
> the Mem-test 86.
>
> Again - no errors. ROCK SOLID STABLE.
>
Hmm... this IS good news!