Puzzling Problems -- Pulling Hair

I just RMA'ed an ASUS motherboard from the same type of response that you
have received. I confirmed the results with Microsofts Memory testing tool -
I had to use extended testing to detect the errors.

When using Memtest, be certain that USB Legacy support is disabled in the
BIOS




"Bill Anderson" <> wrote in message
news:e-WdnS-...
> Kent_Diego wrote:
>>>> ....
>>> I went into BIOS and made a change. Now look:
>>>
>>> http://paintrock.net/cpu-z_memory.jpg
>>> http://paintrock.net/cpu-z_spd.jpg
>>>
>>> So did I accomplish anything? Am I better off? Memtest still gives
>>> exactly the same readout, apparently still showing errors. Docmem shows
>>> no errors. Wish I could take a screen shot and show you.
>>>
>>> I made the change in the "Jumperfree Configuration" section of BIOS.
>>> When I set everything up back in August I'd read that if I wanted to get
>>> a full 1066 DRAM frequency from my Crucial Ballistix memory I needed to
>>> go into BIOS and set the DRAM frequency to 1066 MHz. The closest thing
>>> I could find was 1067 MHz, so I made the change. Everything seemed to
>>> be working fine until a couple of weeks ago. Was I "overclocking?"
>>> Whatever -- I have now changed the setting back to "Auto," and you can
>>> see the result in the links above. Everything in the Jumperfree
>>> Configuration section of BIOS is now set to Auto.
>>>
>>> Do you think I'm where I need to be?
>>> .....
>> Well that is 400 MHz. I bet ASUS has a newer BIOS since August. So
>> Memtest86 is giving errors? Which test number is failing? Normally
>> Memtest86 will just run and the status window will be empty.
>> http://www.memtest.org/#screen
>> Maybe you can describe the errors and when they occur a littler better.
>> This is what errors look like:
>> http://www.flickr.com/photos/jonasdionisio/1229794358/
>> Each one is a line in red.
>>
>>
>
> Well stupid me. I should have figured out on my own that if you can't do
> a screen capture, you can always just snap a picture.
>
> Here's what Memtest looks like when it's starting. Sorry it's off-center,
> but I think this one's not important:
>
> http://paintrock.net/memtest-1.jpg
>
> Here's what it looks like on my screen for a total of about one second.
> This photo has given me my first actual opportunity to read this screen:
>
> http://paintrock.net/memtest-2.jpg
>
> When the previous screen disappears after the one second it exists, I get
> this:
>
> http://paintrock.net/memtest3.jpg
>
> So you see I get nothing at all in red. I'm not sure just what Memtest is
> saying to me. When Docmem began going crazy with all the errors it was
> finding when the one Ballistix stick was plugged in, I knew I was on to
> something. But Memtest just seems so casual about everything that it's
> almost like I'm seeing something normal, even though it seems to be
> showing errors at the same time. Can you decode it? Thanks.
>
> --
> Bill Anderson
>
> I am the Mighty Favog

Keith wrote:
> I just RMA'ed an ASUS motherboard from the same type of response that you
> have received. I confirmed the results with Microsofts Memory testing tool -
> I had to use extended testing to detect the errors.
>
> When using Memtest, be certain that USB Legacy support is disabled in the
> BIOS
>
>

Thanks. As it happens, I had USB legacy support disabled all along -
but not because I want it that way. The only disappointment I've had
with this P5K Deluxe WiFi-Ap board is that I can't have any USB hub
plugged in at boot unless USB legacy support is disabled. This means I
get USB support fine once the OS has loaded, but I can't do anything via
USB prior to having an OS in place. Thus, I can't flash bios with a
memory stick unless I unplug my hub and re-set USB legacy support in
BIOS. Doesn't come up often, but still an annoyance.

So, what is there about my Memtest results that makes you think I may
need a new motherboard? I thought I was detecting errors in the memory
stick, not in the motherboard. I need to re-think?

--
Bill Anderson

I am the Mighty Favog

Bill Anderson wrote:
> Kent_Diego wrote:
>>>> ....
>>> I went into BIOS and made a change. Now look:
>>>
>>> http://paintrock.net/cpu-z_memory.jpg
>>> http://paintrock.net/cpu-z_spd.jpg
>>>
>>> So did I accomplish anything? Am I better off? Memtest still gives
>>> exactly the same readout, apparently still showing errors. Docmem
>>> shows no errors. Wish I could take a screen shot and show you.
>>>
>>> I made the change in the "Jumperfree Configuration" section of
>>> BIOS. When I set everything up back in August I'd read that if I
>>> wanted to get a full 1066 DRAM frequency from my Crucial Ballistix
>>> memory I needed to go into BIOS and set the DRAM frequency to 1066
>>> MHz. The closest thing I could find was 1067 MHz, so I made the
>>> change. Everything seemed to be working fine until a couple of
>>> weeks ago. Was I "overclocking?" Whatever -- I have now changed
>>> the setting back to "Auto," and you can see the result in the links
>>> above. Everything in the Jumperfree Configuration section of BIOS
>>> is now set to Auto. Do you think I'm where I need to be?
>>> .....
>> Well that is 400 MHz. I bet ASUS has a newer BIOS since August. So
>> Memtest86 is giving errors? Which test number is failing? Normally
>> Memtest86 will just run and the status window will be empty.
>> http://www.memtest.org/#screen
>> Maybe you can describe the errors and when they occur a littler
>> better. This is what errors look like:
>> http://www.flickr.com/photos/jonasdionisio/1229794358/
>> Each one is a line in red.
>>
>>
>
> Well stupid me. I should have figured out on my own that if you can't
> do a screen capture, you can always just snap a picture.
>
> Here's what Memtest looks like when it's starting. Sorry it's
> off-center, but I think this one's not important:
>
> http://paintrock.net/memtest-1.jpg
>
> Here's what it looks like on my screen for a total of about one
> second. This photo has given me my first actual opportunity to read
> this screen:
> http://paintrock.net/memtest-2.jpg
>
> When the previous screen disappears after the one second it exists, I
> get this:
>
> http://paintrock.net/memtest3.jpg
>
> So you see I get nothing at all in red. I'm not sure just what
> Memtest is saying to me. When Docmem began going crazy with all the
> errors it was finding when the one Ballistix stick was plugged in, I
> knew I was on to something. But Memtest just seems so casual about
> everything that it's almost like I'm seeing something normal, even
> though it seems to be showing errors at the same time. Can you
> decode it? Thanks.

First, I like Memtest86+, but that's not important. It looks like there
is bad memory in the lowest memory addresses. The picture shows that
errors occur starting right at the lowest possible address, 00000000,
etc. Memtest needs some memory to run in, and it sets aside the first
MB, or whatever it is (can't remember) for it to run in. If that memory
is bad, Memtest can't actually run and that's why it immediately comes
up with those errors. AFAIK, that seems to be what's happening.

Does it do this with each stick installed by itself? If so, try
increasing Vdimm (memory voltage) by a step or two and see if that
helps. If both show errors like that, and increasing voltage doesn't
make a difference, then the MB may be bad.

Another thing to try is to set the memory timings manually and set it to
DDR-800, or 667: something lower than what it's rated. Are there errors
still?

RobV wrote:
> Bill Anderson wrote:
>> Kent_Diego wrote:
>>>>> ....
>>>> I went into BIOS and made a change. Now look:
>>>>
>>>> http://paintrock.net/cpu-z_memory.jpg
>>>> http://paintrock.net/cpu-z_spd.jpg
>>>>
>>>> So did I accomplish anything? Am I better off? Memtest still gives
>>>> exactly the same readout, apparently still showing errors. Docmem
>>>> shows no errors. Wish I could take a screen shot and show you.
>>>>
>>>> I made the change in the "Jumperfree Configuration" section of
>>>> BIOS. When I set everything up back in August I'd read that if I
>>>> wanted to get a full 1066 DRAM frequency from my Crucial Ballistix
>>>> memory I needed to go into BIOS and set the DRAM frequency to 1066
>>>> MHz. The closest thing I could find was 1067 MHz, so I made the
>>>> change. Everything seemed to be working fine until a couple of
>>>> weeks ago. Was I "overclocking?" Whatever -- I have now changed
>>>> the setting back to "Auto," and you can see the result in the links
>>>> above. Everything in the Jumperfree Configuration section of BIOS
>>>> is now set to Auto. Do you think I'm where I need to be?
>>>> .....
>>> Well that is 400 MHz. I bet ASUS has a newer BIOS since August. So
>>> Memtest86 is giving errors? Which test number is failing? Normally
>>> Memtest86 will just run and the status window will be empty.
>>> http://www.memtest.org/#screen
>>> Maybe you can describe the errors and when they occur a littler
>>> better. This is what errors look like:
>>> http://www.flickr.com/photos/jonasdionisio/1229794358/
>>> Each one is a line in red.
>>>
>>>
>> Well stupid me. I should have figured out on my own that if you can't
>> do a screen capture, you can always just snap a picture.
>>
>> Here's what Memtest looks like when it's starting. Sorry it's
>> off-center, but I think this one's not important:
>>
>> http://paintrock.net/memtest-1.jpg
>>
>> Here's what it looks like on my screen for a total of about one
>> second. This photo has given me my first actual opportunity to read
>> this screen:
>> http://paintrock.net/memtest-2.jpg
>>
>> When the previous screen disappears after the one second it exists, I
>> get this:
>>
>> http://paintrock.net/memtest3.jpg
>>
>> So you see I get nothing at all in red. I'm not sure just what
>> Memtest is saying to me. When Docmem began going crazy with all the
>> errors it was finding when the one Ballistix stick was plugged in, I
>> knew I was on to something. But Memtest just seems so casual about
>> everything that it's almost like I'm seeing something normal, even
>> though it seems to be showing errors at the same time. Can you
>> decode it? Thanks.
>
> First, I like Memtest86+, but that's not important. It looks like there
> is bad memory in the lowest memory addresses. The picture shows that
> errors occur starting right at the lowest possible address, 00000000,
> etc. Memtest needs some memory to run in, and it sets aside the first
> MB, or whatever it is (can't remember) for it to run in. If that memory
> is bad, Memtest can't actually run and that's why it immediately comes
> up with those errors. AFAIK, that seems to be what's happening.
>
> Does it do this with each stick installed by itself? If so, try
> increasing Vdimm (memory voltage) by a step or two and see if that
> helps. If both show errors like that, and increasing voltage doesn't
> make a difference, then the MB may be bad.
>
> Another thing to try is to set the memory timings manually and set it to
> DDR-800, or 667: something lower than what it's rated. Are there errors
> still?
>
>

I think there's a mismatch between the version of Memtest I've been
using and my computer. A little while ago I noticed there must be two
versions of Memtest. So I located and downloaded Memtest86+. I let it
run for two full passes and it never found an error. Nothing in red.
Nothing below the line. No errors at all. This version of Memtest made
much more sense to me than the one I'd tried earlier. I'd post a
picture of the Memtest64+ display, but I think there's no need. It
found no errors. I think that when Crucial sends me a replacement stick
of Ballistix memory I'll be back in business. Things are running OK
dependably now, but noticeably slower. I'll take dependability over
speed any day.

I really appreciate all the help you folks are giving me.

--
Bill Anderson

I am the Mighty Favog

Bill Anderson wrote:
> RobV wrote:
>> Bill Anderson wrote:
>>> Kent_Diego wrote:
>>>>>> ....
>>>>> I went into BIOS and made a change. Now look:
>>>>>
>>>>> http://paintrock.net/cpu-z_memory.jpg
>>>>> http://paintrock.net/cpu-z_spd.jpg
>>>>>
>>>>> So did I accomplish anything? Am I better off? Memtest still
>>>>> gives exactly the same readout, apparently still showing errors.
>>>>> Docmem shows no errors. Wish I could take a screen shot and show
>>>>> you. I made the change in the "Jumperfree Configuration" section
>>>>> of
>>>>> BIOS. When I set everything up back in August I'd read that if I
>>>>> wanted to get a full 1066 DRAM frequency from my Crucial Ballistix
>>>>> memory I needed to go into BIOS and set the DRAM frequency to 1066
>>>>> MHz. The closest thing I could find was 1067 MHz, so I made the
>>>>> change. Everything seemed to be working fine until a couple of
>>>>> weeks ago. Was I "overclocking?" Whatever -- I have now changed
>>>>> the setting back to "Auto," and you can see the result in the
>>>>> links above. Everything in the Jumperfree Configuration section
>>>>> of BIOS is now set to Auto. Do you think I'm where I need to be?
>>>>> .....
>>>> Well that is 400 MHz. I bet ASUS has a newer BIOS since August. So
>>>> Memtest86 is giving errors? Which test number is failing? Normally
>>>> Memtest86 will just run and the status window will be empty.
>>>> http://www.memtest.org/#screen
>>>> Maybe you can describe the errors and when they occur a littler
>>>> better. This is what errors look like:
>>>> http://www.flickr.com/photos/jonasdionisio/1229794358/
>>>> Each one is a line in red.
>>>>
>>>>
>>> Well stupid me. I should have figured out on my own that if you
>>> can't do a screen capture, you can always just snap a picture.
>>>
>>> Here's what Memtest looks like when it's starting. Sorry it's
>>> off-center, but I think this one's not important:
>>>
>>> http://paintrock.net/memtest-1.jpg
>>>
>>> Here's what it looks like on my screen for a total of about one
>>> second. This photo has given me my first actual opportunity to read
>>> this screen:
>>> http://paintrock.net/memtest-2.jpg
>>>
>>> When the previous screen disappears after the one second it exists,
>>> I get this:
>>>
>>> http://paintrock.net/memtest3.jpg
>>>
>>> So you see I get nothing at all in red. I'm not sure just what
>>> Memtest is saying to me. When Docmem began going crazy with all the
>>> errors it was finding when the one Ballistix stick was plugged in, I
>>> knew I was on to something. But Memtest just seems so casual about
>>> everything that it's almost like I'm seeing something normal, even
>>> though it seems to be showing errors at the same time. Can you
>>> decode it? Thanks.
>>
>> First, I like Memtest86+, but that's not important. It looks like
>> there is bad memory in the lowest memory addresses. The picture
>> shows that errors occur starting right at the lowest possible
>> address, 00000000, etc. Memtest needs some memory to run in, and it
>> sets aside the first MB, or whatever it is (can't remember) for it
>> to run in. If that memory is bad, Memtest can't actually run and
>> that's why it immediately comes up with those errors. AFAIK, that
>> seems to be what's happening. Does it do this with each stick
>> installed by itself? If so, try
>> increasing Vdimm (memory voltage) by a step or two and see if that
>> helps. If both show errors like that, and increasing voltage doesn't
>> make a difference, then the MB may be bad.
>>
>> Another thing to try is to set the memory timings manually and set
>> it to DDR-800, or 667: something lower than what it's rated. Are
>> there errors still?
>>
>>
>
> I think there's a mismatch between the version of Memtest I've been
> using and my computer. A little while ago I noticed there must be two
> versions of Memtest. So I located and downloaded Memtest86+. I let
> it run for two full passes and it never found an error. Nothing in
> red. Nothing below the line. No errors at all. This version of
> Memtest made much more sense to me than the one I'd tried earlier. I'd
> post a picture of the Memtest64+ display, but I think there's no
> need. It found no errors. I think that when Crucial sends me a
> replacement stick of Ballistix memory I'll be back in business. Things
> are running OK dependably now, but noticeably slower. I'll
> take dependability over speed any day.
>
> I really appreciate all the help you folks are giving me.

See, I told you Memtest 86+ was better. ;-) 86+ shows more information
that 86 does and I use 86+ for that reason.

Sounds as if you got it, with one bad stick. Hopefully, that's all it
is and I'm sure it is. I had a problem similar to yours years ago with
OCZ memory. Both sticks showed errors in Memtest86 (this was before
86+), but they gave me a really hard time about exchanging them. They
finally did in the end and the new ones have no errors and work fine in
the other computer.

Merry Christmas, Happy New Year, Happy Holidays, or whatever else works.
;-)

I my case, the erros occurred with 3 pairs of RAM at all standard settings -
suggesting a bad memory controlled. If your errors follow a single stick,
then you are correct to conclude it is the stick.

I apologize for my prior lack of clarity.






"Bill Anderson" <> wrote in message
news:...
> Keith wrote:
>> I just RMA'ed an ASUS motherboard from the same type of response that you
>> have received. I confirmed the results with Microsofts Memory testing
>> tool - I had to use extended testing to detect the errors.
>>
>> When using Memtest, be certain that USB Legacy support is disabled in the
>> BIOS
>>
>>
>
> Thanks. As it happens, I had USB legacy support disabled all along - but
> not because I want it that way. The only disappointment I've had with
> this P5K Deluxe WiFi-Ap board is that I can't have any USB hub plugged in
> at boot unless USB legacy support is disabled. This means I get USB
> support fine once the OS has loaded, but I can't do anything via USB prior
> to having an OS in place. Thus, I can't flash bios with a memory stick
> unless I unplug my hub and re-set USB legacy support in BIOS. Doesn't
> come up often, but still an annoyance.
>
> So, what is there about my Memtest results that makes you think I may need
> a new motherboard? I thought I was detecting errors in the memory stick,
> not in the motherboard. I need to re-think?
>
> --
> Bill Anderson
>
> I am the Mighty Favog

Kent_Diego wrote:
> Nothing is getting too hot? I would start with RAM. I have seen bad RAM
> corrupt files on hard drive and cause many problems. Run MemTest86. Try
> increasing voltage or slowing speed and timings to troubleshoot. Next likely
> is mother board then power supply. You can trouble shoot to see if video
> card is problem by running ATI Tool and undrclocking GPU.
>
> Troubleshooting a problem like this is often a matter of replacing parts
> until fixed.
>
>


I'm baaaaack. Here was the original problem:

Motherboard: ASUS P5K DELUXE/WIFI-AP LGA 775 Intel P35 ATX
Processor: Intel Core 2 Quad Q6600 Kentsfield 2.4GHz 2 x 4MB L2 Cache
LGA 775
Memory: Crucial 2GB kit (1GBx2), Ballistix 240-pin DIMM, DDR2
PC2-8500
Video Card: SAPPHIRE 100210L Radeon HD 2600XT 256MB 128-bit GDDR4 PCI
Express x16
Power Supply: PC Power and Cooling ULTRA-QUIET PSU: SILENCER(R) 610
EPS12V
Video capture: ATI VisionTek TV Wonder PCI Express (added within the
past month)
Storage: Four 500 gigabyte hard drives -- three Western Digital and
one Maxtor.

I'm running a triple-boot configuration with these OSs in separate
partitions on a single 500 gigabyte drive:

C: WinXP
D: Vista
E: WinXP testbed (a small partition where I try out things before
installing them on C

As I explained here in early December, the system had gone flaky, with
odd problems cropping up and Vista refusing to run at all and total
unpredictability. The solution was to run Memtest86+ and learn that one
of my two 1-gig memory sticks appeared to be bad. I removed it, RMAed
it to Crucial, ran for about a week with 1-gig RAM, and installed the
new stick from Crucial when it arrived. All has been working apparently
fine since then.

This morning I was typing a Usenet post in Thunderbird when suddenly my
mouse froze. Then I realized everything had frozen -- no mouse, no
Windows Task Manager -- nothing.

So I pressed the reset button and the system wouldn't even POST! What
to do? I removed the case cover and noticed that when I'd try to power
up the fancy blinky lights on the Crucial Ballistix memory sticks were
dark. Both sticks were dark. Hey, maybe the blinky lights would come
in handy for something after all. After letting the system run for
about 30 seconds, the memory LEDs would actually light up, but they'd
never blink as normal.

So I unplugged power and removed both sticks from DIMM A1 and DIMM B1
where I had them. I then put one of them in the DIMM B2 slot. Don't
ask why, I just decided to do that. When I rebooted, the stick lit up
like a hyperactive Christmas tree and I heard the post beep and I was
back in business as though nothing had happened. I did get a BIOS error
when booting -- seems "overclocking" had failed. So when I landed in
BIOS, I reset the memory DRAM frequency from its advertised DDR2-1066 to
Auto.

Everything has been running fine since, with 1-gig RAM. I have company
coming tomorrow and they'll stay until Tuesday afternoon and I want my
computer to be working fine while they're here. So I'm not going to
touch a thing until Tuesday night. It's working now and I have no plans
to fix it yet.

Still -- am I making a mistake trying to make my memory sticks run at
1066? Crucial advertises that they will. Should I just give that up?
Ideas for me to try on Tuesday?

--
Bill Anderson

I am the Mighty Favog

Bill Anderson wrote:
> Kent_Diego wrote:
>> Nothing is getting too hot? I would start with RAM. I have seen bad
>> RAM corrupt files on hard drive and cause many problems. Run
>> MemTest86. Try increasing voltage or slowing speed and timings to
>> troubleshoot. Next likely is mother board then power supply. You can
>> trouble shoot to see if video card is problem by running ATI Tool and
>> undrclocking GPU.
>>
>> Troubleshooting a problem like this is often a matter of replacing
>> parts until fixed.
>>
>
>
> I'm baaaaack. Here was the original problem:
>
> Motherboard: ASUS P5K DELUXE/WIFI-AP LGA 775 Intel P35 ATX
> Processor: Intel Core 2 Quad Q6600 Kentsfield 2.4GHz 2 x 4MB L2 Cache
> LGA 775
> Memory: Crucial 2GB kit (1GBx2), Ballistix 240-pin DIMM, DDR2
> PC2-8500
> Video Card: SAPPHIRE 100210L Radeon HD 2600XT 256MB 128-bit GDDR4 PCI
> Express x16
> Power Supply: PC Power and Cooling ULTRA-QUIET PSU: SILENCER(R) 610
> EPS12V
> Video capture: ATI VisionTek TV Wonder PCI Express (added within the
> past month)
> Storage: Four 500 gigabyte hard drives -- three Western Digital and
> one Maxtor.
>
> I'm running a triple-boot configuration with these OSs in separate
> partitions on a single 500 gigabyte drive:
>
> C: WinXP
> D: Vista
> E: WinXP testbed (a small partition where I try out things before
> installing them on C
>
> As I explained here in early December, the system had gone flaky, with
> odd problems cropping up and Vista refusing to run at all and total
> unpredictability. The solution was to run Memtest86+ and learn that one
> of my two 1-gig memory sticks appeared to be bad. I removed it, RMAed
> it to Crucial, ran for about a week with 1-gig RAM, and installed the
> new stick from Crucial when it arrived. All has been working apparently
> fine since then.
>
> This morning I was typing a Usenet post in Thunderbird when suddenly my
> mouse froze. Then I realized everything had frozen -- no mouse, no
> Windows Task Manager -- nothing.
>
> So I pressed the reset button and the system wouldn't even POST! What
> to do? I removed the case cover and noticed that when I'd try to power
> up the fancy blinky lights on the Crucial Ballistix memory sticks were
> dark. Both sticks were dark. Hey, maybe the blinky lights would come
> in handy for something after all. After letting the system run for
> about 30 seconds, the memory LEDs would actually light up, but they'd
> never blink as normal.
>
> So I unplugged power and removed both sticks from DIMM A1 and DIMM B1
> where I had them. I then put one of them in the DIMM B2 slot. Don't
> ask why, I just decided to do that. When I rebooted, the stick lit up
> like a hyperactive Christmas tree and I heard the post beep and I was
> back in business as though nothing had happened. I did get a BIOS error
> when booting -- seems "overclocking" had failed. So when I landed in
> BIOS, I reset the memory DRAM frequency from its advertised DDR2-1066 to
> Auto.
>
> Everything has been running fine since, with 1-gig RAM. I have company
> coming tomorrow and they'll stay until Tuesday afternoon and I want my
> computer to be working fine while they're here. So I'm not going to
> touch a thing until Tuesday night. It's working now and I have no plans
> to fix it yet.
>
> Still -- am I making a mistake trying to make my memory sticks run at
> 1066? Crucial advertises that they will. Should I just give that up?
> Ideas for me to try on Tuesday?
>

Is Vdimm raised, to help the memory work at DDR2-1066 ? I don't think
any DDR2 wants to do that with only 1.8V applied. It probably takes
more than that.

Paul

Paul wrote:
> Bill Anderson wrote:

>>
>> Still -- am I making a mistake trying to make my memory sticks run at
>> 1066? Crucial advertises that they will. Should I just give that up?
>> Ideas for me to try on Tuesday?
>>
>
> Is Vdimm raised, to help the memory work at DDR2-1066 ? I don't think
> any DDR2 wants to do that with only 1.8V applied. It probably takes
> more than that.
>

I haven't run CPU-Z today. Here's what it looked like earlier this
month when I was having the original problems:

http://paintrock.net/cpu-z_spd.jpg

So, is that 1.8V or 2.2V?


--
Bill Anderson

I am the Mighty Favog

Bill Anderson wrote:
> Paul wrote:
>> Bill Anderson wrote:
>
>>>
>>> Still -- am I making a mistake trying to make my memory sticks run at
>>> 1066? Crucial advertises that they will. Should I just give that
>>> up? Ideas for me to try on Tuesday?
>>>
>>
>> Is Vdimm raised, to help the memory work at DDR2-1066 ? I don't think
>> any DDR2 wants to do that with only 1.8V applied. It probably takes
>> more than that.
>>
>
> I haven't run CPU-Z today. Here's what it looked like earlier this
> month when I was having the original problems:
>
> http://paintrock.net/cpu-z_spd.jpg
>
> So, is that 1.8V or 2.2V?
>
>

Cool. I've never seen an EPP displayed in CPUZ before. Yes,
that looks like 2.2V to run DDR2-1066 (right most column).
And the command rate is set to 2T as well (command/address
cycles last for two clocks).

Column 1 shows DDR2-667 with no sweat. Column 2 versus
column 3 shows that added voltage allows the timing to
be tightened to 4-4-4. A little looser timing in column 4,
allows the clock to be raised to DDR2-1066.

There is probably a manufacturer provided maximum voltage
value, but that might not be shown in CPUZ. In other words,
more than 2.2V may be allowed, but it would be a good idea
to check what the value is first. For some companies, you
have to download their datasheet to find out what the limit
is.

Paul