Re: N6800 vid not able to use full 3 GB MEM on A7V333 but works for 2.5GB or lower

In article <1151989470.414347.324160@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>,
"Liquid" <liquidtek555@xxxxxxxxx> wrote:

Hi all,

I having a frustrating problem with upgrading my video card to an ASUS
N6800 (Geforce 6800 512MB of ram) on an ASUS A7V333 motherboard. At
first I couldn't get it to post ASUS tier 3 tech support said to RMA or
exchange it.

I finally got it to work after giving up on bios changes. I tried just
1GB DIMM and it worked! Problem is I can only get 2 1GB DIMM's to
work. Or 2 1GB DIMM's and one old 512MB DIMM. The new memory is 3GB of
Corsair XMS PC3200 part number CMX1024-3200C2PT for the 1GB sticks.
The old memory is 512MB Corsair XMS part number CMX512-3000C2.

I can get all 3GB of memory to work with my old video card ASUS V8440
(Geforce 4 4400 ultra) no errors with memtest86. I'm unsure as what is
the culprit. I'm thinking the vid card, motherboard or power supply
(which I have a gut feeling for). The power supply is an Antec 430
true power came with case wanted to get better one but was free. Am I
correct that my board uses a 20 pin power cable for the board?

My 3.3 voltage goes from 3.24-3.26 is this the voltage the RAM runs
under? Is this the problem? Do I maybe need a bigger/better power
supply? I'm gonna run memtest86 tonight to see if it is stable with
2.5GB. Will post any news. I would really like to max my board (3
slots) to 3GB. Voltage for everything is stable line in ASUS PC Probe
but 12v and 3.3v go up and down slightly. I just bought an APC smart
ups 500 watt so I'll try connecting that too but I don't think I have
"dirty" power.

Also, when I called back tier 3 tech support with my finding they
advised to run all memory of same size and manufacturer and part
number. If memtest86 checks out will it be rock solid if I have to run
a 512MB in the mix (getting same Corsair XMS series with free
restocking if not needed). The 1GB sticks are same part number but 1
is a different version only bought less then one week apart from
newegg. On original vid card I had to put two of the same version in
slots 1 and 3 instead of 1 and 2 to get no errors in memtest86 did get
errors in test 5 on memtest86 moving blocks where the 2 goes to 3 slot
I think. But OS didn't lockup was just checking memory didn't run OS
that much before testing or stress it...

Well, any help or insight to diagnosing this or advice would be GREATLY
appreciated. This is buggin me I want to use my full 3GB and don't
want to use my old vid card.

Thanks in advance for any help,
Dan

Three sticks of CAS2 memory should work as well as could be expected.
With that loading, of three double sided sticks, downclocking the
memory is a likely solution to whatever stability or memory errors
you are getting.

Motherboards tend to re-regulate the voltages coming from the
power supply. DIMMs tend to have a small linear or switching
regulator, for the memory voltage and for the termination
voltage. Internally regulated motherboard voltages are not
monitored on the hardware monitor chip, so the only way to
verify what is actually being used, is with a voltmeter.
In this case, the supply voltage is intended to be in the
2.5-2.6V range, and does not come directly from 3.3V.

In the fifth post down, there is a table of JP1/JP2 jumper
position, versus DIMM supply voltage.

http://forums.pcper.com/printthread.php?t=113092&pp=40

According to this article, the default setting is already
providing 2.78 - 2.85V, which is as far as I would go with
memory that is not known to overvolt well. If you had some
old BH5 based memory, then virtually any jumper setting
would be valid. I certainly wouldn't run my Ballistix
memory in a board like that, and use all possible jumper
settings, because the high setting would fry my RAM.
So stick with the default.

http://www.lostcircuits.com/motherboard/asus_a7v333/5.shtml

Paul
.

Relevant Pages

  • Re: Supervisory Processors
    ... monitors the power supply voltage to detect a brown-out ... But maybe supervisory processors will make a comeback. ... cycle the card. ...
    (comp.arch)
  • Re: Supervisory Processors
    ... monitors the power supply voltage to detect a brown-out ... But maybe supervisory processors will make a comeback. ... cycle the card. ...
    (comp.arch)
  • Re: Supervisory Processors
    ... I recall the DEC-20 had an 11/40 console processor. ... at power-up in which the voltage rose, then fell, then ... finally rose again and passed the minimum operating power ... cycle the card. ...
    (comp.arch)
  • Supervisory Processors
    ... monitors the power supply voltage to detect a brown-out ... But maybe supervisory processors will make a comeback. ... cycle the card. ...
    (comp.arch)
  • Re: Hunter Alignment Machine
    ...  Possibly just a DC power supply issue. ... okay, but no beep from the CPU board on power up, and no response to ... boards and cleaned card edges. ... If no voltage, remove cards one by one, power ...
    (rec.crafts.metalworking)