A7V333: Memory voltage settings and how I stabilized this board.
Moe Howard
two days ago, because I ran into some problems I could not solve until I
experimented a bit with the undocumented jumper settings, JP1 & JP2, which
control the voltage into the DDR memory.
For starters, I had been monitoring the Asus NG and performing Google searches
in preparation of the installation. There seems to be a lot of confusion out
there as to what exactly are the correct settings for the jumpers (ie, JP1 and
JP2), with conflicting information coming from sources like Mushkin, Asus, and
this NG. I'm still not sure what to believe at this point, but I will detail
what worked for me.
According to Mushkin, the default settings for JP1 & JP2 are wrong. You can
read about it here:
http://www.mushkin.com/cgi-bin/Mushkin.filereader?3cc5e7931d2076f02740c0a801020
6ad+EN/products/990880#asusa7v333
[watch out for wrapping of URL, you'll have to cut-n-paste]
My mobo, which is Rev. 1.02, came with the same default settings, which are:
JP1 JP2
. o
o o
o .
That is, JP1==>pins 1,2 jumped. JP2==>pins 2,3 jumped.
On the advice of some in this NG who couldn't get their mobos to work and also
after reading the Mushkin write-up, I changed the jumpers to give what I
thought would be 2.5V to my PC2700 DDR ram (512MB module, Kingston brand).
This then would mean that I jumped the pins to the following:
JP1==>1,2 JP2==>1,2
According to Mushkin, this would give me 2.5V and the most stable configuration
for the memory. Also, because these two jumpers are undocumented in the Asus
manual, I found out from this NG that they are documented in the A7V266-E
manual, which I downloaded and looked at. Low and behold the 266 manual shows
that pins 1,2 on both JP1 and JP2 have to be closed to give 2.5V, but comes
from the factory set at 2.65V by default (JP1==>2,3 JP2==>1,2).
Now on my A7V333, rev 1.02, the board came from the factory with the default
settings of JP1==>1,2 and JP2==>2,3. According to the Asus 266 manual this
would mean that my board came jumpered by default set to 2.75V !! But this is
assuming, of course, that the jumper blocks and pinouts for the A7V266-E are
identical with those of the A7V333. I think they are not, and here's why.
When I got my machine ready to go with just the basic components (Athlon XP
1900+, 512MB DDR,cdrom, floppy, an older Fujitsu 10GB IDE HD, and a spare TNT2
M64 AGP graphics card), I fired it up and it booted just fine. I went into the
bios (v. 1006) and loaded setup defaults, saved, rebooted, and all was fine.
For testing purposes, I then rebooted with a copy of MemTest86, and proceeded
with testing the ram. After about an hour the program started showing memory
errors all over the place, across most of the chips built on the ram module.
My first thought was, "Shit, I've got a bad stick of RAM." After some thought,
I decided to return the jumper settings in JP1&JP2 to where they were from the
factory: JP1==>1,2; JP2==>2,3
Upon restarting the machine and running Memtest86 again, this time NO errors
were produced. I let the machine run for 30 hours under Memtest, and not ONE
single error. Conclusion: My particular memory module is only happy with the
Asus default setting, whatever that is in Volts because it's not fricken
documented in the manual. Also, I don't know what to think about the Mushkin
write-up, because my machine clearly would not work under their suggested
settings. According to their diagram, I have my boarded jumpered to deliver
2.78 - 2.83V. I find this hard to believe, and I have to wonder if a revision
change by Asus has changed the meaning and values of jumpers JP1 & JP2.
Does anyone else with this mobo have anything to add? What are your jumper
settings, and what is your mobo revision number? BTW, I am not OC'ing my
board, and have it setup using jumpers and dip switches as opposed to "Jumper
Free." Thanks everyone for taking the time to read this [totally to long]
post.
Regards,
Moe
ntra
I had the same experience. I was led to believe, as you were, that
altering (or, in my case, simply removing) the jumpers was
advantagous. My system wasn't even close to stable with them removed,
and I received bluescreens regularly, randomly self-quitting
applications, general instability. Memtest showed errors as well.
Putting the jumpers back on as they came (JP1/JP2 = 1-2/2-3) returned
my system to health...and I've regained the ability to use a slight
overclock (fsb=138) at the same time, which I had lost when removing
the jumpers.
My tentative conclusion, as yours seems to be as well, is that the
meaning of the jumpers has changed between the A7V266 and A7V333.
Asus should have either documented the jumpers, or even just labelled
them something different so nobody assumed that they were functionally
identical to the 266 board's JP1/JP2.
It's also disturbing that Mushkin, and other sources, seem to have
made inferences about these jumpers and issued what looked like
authoritative information about them, when from what people are seeing
"on the street", that information was simply incorrect.
-ntra
On Fri, 21 Jun 2002 13:17:05 +0200, "Moe Howard" <m...@ibm.org> wrote:
>[snip]
Sadar
jumper
> settings, and what is your mobo revision number? BTW, I am not OC'ing my
> board, and have it setup using jumpers and dip switches as opposed to
"Jumper
> Free." Thanks everyone for taking the time to read this [totally to
long]
> post.
>
> Regards,
>
> Moe
>
Nice write up Moe. I myself am using Samsung PC2700 with factory default my
machine erratically crashed/locked up. After setting to the 1-2 1-2 jumper
setting everything is peachy.
So in the end it comes down to checking the specs of the DDR used & a
serious slap for ASUS for omitting this vital info.
Sadar
---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.371 / Virus Database: 206 - Release Date: 13/06/2002
Davide
> Asus default setting, whatever that is in Volts because it's not fricken
> documented in the manual. Also, I don't know what to think about the
Mushkin
> write-up, because my machine clearly would not work under their suggested
> settings. According to their diagram, I have my boarded jumpered to
deliver
> 2.78 - 2.83V. I find this hard to believe, and I have to wonder if a
revision
> change by Asus has changed the meaning and values of jumpers JP1 & JP2.
I am in the same situation. I made all the tests that you made and even
for me the conclusion is that my memory is only happy with Asus default
setting. My config is :
Asus A7v333
Athlon XP 1800
512Mb = 256x2 Mb PC2700 Nanya 2.5 (1 & 3 memory Slot)
Ati Radeon 8500
HD 40Gb Maxtor + HD 30Gb IBM
Asus CD-ROM 52x
Plextor 12/10/32
WinXP professional
If I change Asus defaults 1-2 e 2-3 (Jp1 e Jp2) WinXP starts but
it's impossible to work : for example programs like 3dmark doesn't
work, the system crash ecc. blue windows etc
With 1-2 e 2-3 all OK and I can mantain fsb 141 without problems.
Hi Davide
TDK
"Moe Howard" <m...@ibm.org> wrote in message
news:aev215$ab3aj$1...@ID-127331.news.dfncis.de...
> Does anyone else with this mobo have anything to add? What are your
jumper
> settings, and what is your mobo revision number?
Hi Moe, running an A7V333 board here rev1.02 , athlon xp 1800, 512 megs
Samsung PC2700 Ram.
JP1 and JP2 are both set to 1-2.
Bios setup is even set to TURBO.
I have had no stability problems whatsoever running Windows XP home edition.
:)
TDK
AL
<snip>
Here's the deal, JDEC spec is 2.5V +/- 10% so that makes the outside
limit 2.75V. However, we frequently go over that here in the labs and
test as high as 2.8V.
<snip>
Taken from:
http://www.houseofhelp.com/v2/showthread.php?s=&threadid=5998&highlight=frequently
You guys are blowing this VID 1 & 2 thing way out of proportion. Asus
PUT them that way for a reason: STABILITY. And as you can see, a major
RAM manufacturer is saying the same thing. 2.75 - 2.8 V isn't going to
kill your RAM. And if for some reason it does, WHO CARES, that's what
the LIFETIME warranty is for.
I'm running ROCK SOLID STABLE at 183MHz (1:1), but ONLY because I'm
using the Default Asus jumper settings and not jumping on the "OH MY
GOD THAT TO MUCH VOLTAGE" bandwagon. Because frankly it won't even
BOOT with those jumpers removed.....
Moe Howard
"AL" <j...@mama.com> wrote in message
news:d207hugeq129u5eos...@4ax.com
: Here is what Corsair says about memory voltages.
:
: <snip>
: Here's the deal, JDEC spec is 2.5V +/- 10% so that makes the outside
: limit 2.75V. However, we frequently go over that here in the labs and
: test as high as 2.8V.
: <snip>
Well, opinions are a lot like a**holes, everybody's got one. ;-) To be a
little more precise, the Jedec spec actually calls the spec as follows: VDDq:
+2.5V +/-0.2V; VDD: +2.5V +/-0.2V
So that's actually about +/- 8%, so in reality DDR memory should NOT be seeing
more than +2.7V being applied. The specs were written for a reason! You can
read all about at
http://www.jedec.org , look for a pdf called "JESD79."
:
: Taken from:
:
:
http://www.houseofhelp.com/v2/showthread.php?s=&threadid=5998&highlight=frequen
tly
:
:
: You guys are blowing this VID 1 & 2 thing way out of proportion. Asus
No I'm [we're] not, I'm simply pointing out inconsistencies from several
sources. Also, I assume you're referring to JP1&JP2, not VID.
: PUT them that way for a reason: STABILITY.
I don't follow your logic. The Jedec spec says don't run DDR past the
established tolerances of 2.7V, and you say Asus has jumpered it's board for
2.75=2.8V by default for **stability**? WTF?
: And as you can see, a major
: RAM manufacturer is saying the same thing. 2.75 - 2.8 V isn't going to
: kill your RAM. And if for some reason it does, WHO CARES, that's what
: the LIFETIME warranty is for.
No insult intended, but that's a lot of crap. Specs are specs, and the numbers
put forth by Jedec were done so for a reason. It's also a line of crap when
manufacturer 'A' (Corsair) says, "Don't worry, be happy. You can run your DDR
**past** the established limitations without problems", while manufacturer 'B'
(Mushkin) says, "Hey wait a minute, don't do it!" What is Joe Consumer
supposed to believe? For that matter, why in the hell won't the Asus board run
properly when jumpered for +2.5V, as PER the Jedec spec!!?? Finally, lifetime
warranties are fine, meanwhile your data is being corrupted up to the point of
RAM failure....does this make any fricken sense???
:
: I'm running ROCK SOLID STABLE at 183MHz (1:1), but ONLY because I'm
: using the Default Asus jumper settings and not jumping on the "OH MY
: GOD THAT TO MUCH VOLTAGE" bandwagon. Because frankly it won't even
: BOOT with those jumpers removed.....
No bandwagon to jump on here, Bucko. I paid a lot of money for this board and
my memory. The **least** I expect is that it runs within established specs
problem free. No, I don't OC and crap like that, I just want a stable machine.
**Stability** dude, is the name of the game. If Asus is supplying what looks
like +2.75-2.85V to my DDR right out of the gate, I have a **serious** problem
with that, because down the road I expect I will have problems with the memory
and data corruption. This, IMHO is entirely bogus.
Regards,
Moe
EODx2
fine - even o/c my 1.2 CPU up to 1.4 with a 140 FSB. I'm using a single
Samsung 2700 512 MB in slot 1.
"Moe Howard" <m...@ibm.org> wrote in message
news:aev215$ab3aj$1...@ID-127331.news.dfncis.de...
RazorFace
Ok, is anyone else confused by the responses to Moe's post? Moe says
he has his MB working with JP1(1,2) & JP2(2,3).
>My first thought was, "Shit, I've got a bad stick of RAM." After some thought,
>I decided to return the jumper settings in JP1&JP2 to where they were from the
>factory: JP1==>1,2; JP2==>2,3
>
>Upon restarting the machine and running Memtest86 again, this time NO errors
>were produced. I let the machine run for 30 hours under Memtest, and not ONE
>single error. Conclusion: My particular memory module is only happy with the
>Asus default setting, whatever that is in Volts because it's not fricken
>documented in the manual.
ntra's post concurs with Moe:
>Putting the jumpers back on as they came (JP1/JP2 = 1-2/2-3) returned
>my system to health...and I've regained the ability to use a slight
>overclock (fsb=138) at the same time, which I had lost when removing
>the jumpers.
Sadar, however, does not concur:
>Nice write up Moe. I myself am using Samsung PC2700 with factory
>default my machine erratically crashed/locked up. After setting to the
>1-2 1-2 jumper setting everything is peachy.
Davide agrees with Moe & ntra:
>If I change Asus defaults 1-2 e 2-3 (Jp1 e Jp2) WinXP starts but
>it's impossible to work : for example programs like 3dmark doesn't
>work, the system crash ecc. blue windows etc
>With 1-2 e 2-3 all OK and I can mantain fsb 141 without problems.
And TDK is in agreement with Sadar:
>Hi Moe, running an A7V333 board here rev1.02 , athlon xp 1800, 512 megs Samsung PC2700 Ram.
>JP1 and JP2 are both set to 1-2.
>Bios setup is even set to TURBO.
>I have had no stability problems whatsoever running Windows XP home edition.
I've yet to build my A7V333 system and this problem is not getting any
clearer. Can anyone smarter than I draw a conclusion from this
dichotomy?
RazorFace
---
RazorFace
AL
to speak for themselves, here are the facts surrounding my system:
System Behavior with JP1-2 REMOVED:
1.) My system gets MAJOR file corruption with JP1-2 REMOVED
2.) I get random reboots and BSOD's with JP1-2 REMOVED
3.) Memtest86 gives me assloads of errors with JP1-2 REMOVED
4.) Highest stable bus speed: 172MHz
System Behavior with JP1-2 at ASUS FACTORY DEFAULTS:
1.) NO file corruption at all.
2.) NO Reboots or BSOD's AT ALL
3.) Memtest86 can run for 24 HOURS+ with 0 errors
4.) Highest stable bus speed: 186MHz
So, these are the facts with my system. These facts lead ME to say,
"The HELL with JDEC specs." I KNOW what works and what doesn't.
And apparently ASUS AND CORSAIR do as well. ASUS by leaving those
jumpers as they are, even on Rev. 1.02, and Corsair by saying it is OK
to run THEIR ram at 2.8V.
So, as you can see from the above facts. I'm going to keep running the
DEFAULT Jumper positions, like MANY other people in this group. So you
go on talking about Whitepapers, SPECS and how things "are supposed to
be". And I'll stick to the facts of how things ARE.
scott c.
board?
I just picked up an A7V333 rev. 1.02 and a couple sticks of Corsair
CMX512-2700C2 XMS2700V1.1. Haven't put it together yet, still reading up on
it
TIA,
Scott
"AL" <j...@mama.com> wrote in message
news:d2c9hu4sqm8rfms62...@4ax.com...
AL
I have a A7V333 Rev. 1.01. I'm running 2 sticks of Corsair XMS3000C2
V1.1. I have it running at 183Mhz FSB (1:1) 366DDR @ CAS (2,2,3,5,1T)
It is VERY stable and have had no file corruption since putting the
jumpers JP1-JP2 back to the factory default settings.
FYI: Corsair also told me if you're running more than one stick, like
I am, they definitely recommend a bit more voltage than JDEC
specifies.
john
i doubt default would be overclocking. its probably < 2.75V . its still
disturbing that its not documented properly. why cant board manufacturers
tell more about their jumpers and specs and features rather than trying to
keep it minimal. i dont know about you but i want to know about every
circuit in the board. so why not put all that info out in .pdf or something
on their web site. a lean printed manual must $ave them money. whats worse
is that support for each board ends as soon as its made. then its obsolete
and theyre on to the next one.
"Moe Howard" <m...@ibm.org> wrote in message
news:af1np7$ap02u$1...@ID-127331.news.dfncis.de...
AL
Here are the settings for RAM Voltage on the A7V333:
ii 2.59 - 2.63V (same with one or both jumpers removed completely)
i! 2.78 - 2.85V (default setting, out of spec according to JEDEC)
!i 2.88 - 2.96V
!! 2.95 - 3.06V
As you can see, the defaut setting IS indeed out of spec. However,
depending on what type, brand, speed, etc. RAM you are using, the
default seems to work best. In addition, most quality sticks of RAM
have built in Voltage Regulators that will not let you get too high on
the voltage.
ntra
I'm using the same RAM you're using (Corsair CMX512-2700C2 parts), and
am one of the folks unable to run stable at 1,2/1,2.
With the jumpers in the from-the-factory 1,2/2,3 positions, this
Corsair memory runs just fine in my setup; I'd advise you to test in
that configuration and change only if you experience problems.
-ntra
On Sat, 22 Jun 2002 18:33:34 GMT, "scott c." <ca_...@yahoo.com>
wrote:
>AL, just curious, what kind of memory are you running and rev. level is your
Maurits
news:n0k9hu4dlhc0skga9...@4ax.com...
> John:
>
> Here are the settings for RAM Voltage on the A7V333:
>
> ii 2.59 - 2.63V (same with one or both jumpers removed completely)
> i! 2.78 - 2.85V (default setting, out of spec according to JEDEC)
> !i 2.88 - 2.96V
> !! 2.95 - 3.06V
>
> As you can see, the defaut setting IS indeed out of spec. However,
> depending on what type, brand, speed, etc. RAM you are using, the
> default seems to work best. In addition, most quality sticks of RAM
> have built in Voltage Regulators that will not let you get too high on
> the voltage.
Wrong. RAM has *never* built in regulators.
Maurits
AL
Corsair. They say their DIMM's have built in regulators.
ntra
Maurits,
Before making declarations about someone being wrong, perhaps it would
be more contructive to ask where they'd received their information?
Many of us have been in contact with Corsair over this very issue, and
the word from them is that regulators on their parts are the only
component of the DIMM that sees this "out of spec" voltage.
Since I'm running Corsair parts - as are many watching this issue in
various forums - and since Corsair has indicated that the
from-the-factory 1,2/2,3 setting should be fine, I'm inclined to
believe them, and their assertions about voltage regulators you claim
are "never" present.
-ntra
On Sun, 23 Jun 2002 01:17:23 +0200, "Maurits"
<m.raven@_nospam_chello.nl> wrote:
alhaz
after running memtest..512 mushkin pc2100news:d207hugeq129u5eos...@4ax.com...
Rex Imperator
says...
> have it setup using jumpers and dip switches as opposed to "Jumper
> Free." Thanks everyone for taking the time to read this [totally to long]
>
>
I am planning to set up the A7V333 with jumpers.
Kindly let me have, as many of your jumper and dip switch settings as
your patience will allow.
At least sysclk, jen, romsip, vid1-4 vid, jp1-2 please.
Many thanks
Rex.
BGanger109
<m...@ibm.org> writes:
>My mobo, which is Rev. 1.02, came with the same default settings, which are:
>JP1 JP2
>. o
>o o
>o .
>
>That is, JP1==>pins 1,2 jumped. JP2==>pins 2,3 jumped.
The above is how my A7V333 was jumpered from the factory. I am running 2-256
megs DDR333 from Crucial and running slightly overclocked (room is hot in
summer).
My PC runs just fine - super stable. For grins I removed the jumpers
completely and I could not get it to boot to the log in screen (I am using W2K
Pro SP2). Put the jumper back and it boots with no problems.....
Board revision is 1.021
-------------------------------------
Brian
CylonRed on OnlineRacing.net