945GCT-M/1333 Graphics
Pete Rynas
compatable so i wrote back asking them for the " List " of incompatable
cards - Duh!! it looks like paul and i are the only people w/any input
anymore on this site?
anyway, i was wondering if perhaps this samsung 932b monitor of mine needs
a video cable the digital one with every pin in it?? i have a few with the
few pins missing in a couple of spots but no full d-pinout plug - but that
is what came w/ the monitor was the cable w/a few missing pins so that along
with the fact that the analog doesn't work any better suggests that the
cable isn't the case. ECS should list incompatable cards!!! i have a Fry's
brand EVGA geforce 7200GS and a 7600GT , they work flawlessly once windows
is running , but you have to reset, reboot , unplug , turn power switch in
rear of machine off several times just to initialize the card( get video
even for Bios)?? i have been at this comp building for years so i'm not
stupid but i will admit i learn every day and may be missing something
simple or obvious or ECS better come up with a "list" or refund / advance
RMA this pile to something that will truely support PCI-E slotted card
Paul
Maybe there is a forum somewhere, where all the ECS fun takes place :-)
And you're missing out :-)
I don't think there is a particular reason to suppose the 945GT is
incompatible. Like I said before, all 16 lanes look to be wired,
judging by the pictures. And other companies have used that chip,
and I bet they don't have the same problems.
Have you been able to run any tests, after you get into Windows ?
Does everything work under load ?
Does the board pass memtest86+ without errors ?
Can you run Prime95 while in Windows ?
Once the video card driver is in place, you have DirectX installed
and so on, does a version of 3DMark run OK on the video card ?
If you're in Windows, and do a "restart", do the odds improve of
the motherboard starting properly ?
It sounds like there is some kind of initialization problem,
almost as if the BIOS is checking the video card before it
is ready, or there is just a problem with the BIOS design.
Since you say you've built a number of systems before, is
there a chance of you popping a PCI Port 80 display card in
the first PCI slot ? Those cards have a 2 digit hex display
and codes are written to the display by the BIOS. The codes
are not error codes - they are progress codes, indicating
when a certain activity is occurring.
The thing about the Port 80 display card, is figuring out
what is happening from the port card, is virtually impossible.
The progress codes tend to bundle several activities
together, so you don't know which one is at fault. And
in any case, the description of the activity is obscure
enough, that it doesn't mean much of anything.
But having said so many bad things about it, it does have
the benefit of being an "activity" indicator. In a system
where the motherboard starts inconsistently, you'd want to see
what the port code display looks like, for the working and
non-working cases.
Another example of an activity indicator, is some people
own RAM that has some kind of LED activity display on the
edge of the DIMM. Some people have used to good effect,
noting that when a BIOS won't run, the activity LEDs never
flash. That is almost as effective as the Port 80 card.
OK, so where does that leave us ?
When the BIOS starts, in your case, you have built-in graphics
and you have a PCI Express video card slot. (I'd have to look
it up, but it is also possible the video card slot also supports
the Intel ADD2 display card, so the PCI Express lanes have more
than one definition. With the ADD2 card, a digital stream is
passed to the ADD2 card, and it has things like a TMDS transmitter
chip on it, to drive a DVI connector.)
The BIOS should check the video card slot, to see if something
is plugged in. If a video card is detected, then at least on
some motherboards, the BIOS would disable the built-in 945GC
graphics, and rely on the video card instead. The BIOS should
load the code stored in the video BIOS chip on the video card.
On one of my FX5200 cards, the video BIOS actually displays
a banner prompt on the screen, when the video BIOS is loaded.
It is the first text to be displayed on the screen. After
that, the normal BIOS display appears. For the video card
to work normally, I would have expected that loading the
BIOS would have been an early activity. Not all video BIOS
display something on the screen, so you cannot rely on that
as a debugging feature. I have two FX5200 cards, one has
the banner, and the other one doesn't.
The video BIOS should define some standard display modes
(VESA?). A couple of those are "standard", and for the
BIOS to be able to display, the appropriate modes have
to exist. Matrox had a couple cards, that failed for some
reason, to meet that requirement, and the old Matrox forum
had a number of reports about it. But I cannot say that I've
read of ATI or Nvidia having problems like that.
Your problem could be a BIOS problem or it could be
a motherboard problem of some sort. Or even a power
supply problem for that matter. As I suggested before,
you can partially eliminate some of those possibilities,
like testing the video cards in another system. Swapping
power supplies doesn't guarantee anything, as there have
been cases where a particular model and brand of power
supply, doesn't work with a bunch of different motherboards.
(Antec comes to mind.) In the Antec case, it almost looked
like a timing problem, like the power supply output was slow
to reach full voltage or something.
At some point, you're going to have to move on, and try
another motherboard. Perhaps your ECS board will make
new friends on Ebay...
I found a page for 932B here. It is a 19" with 1280x1024
resolution. That should not cause any grief, for any interface.
That resolution can be handled by a "single link" DVI digital
output (half the pins populated).
http://en.wikipedia.org/wiki/Dvi
Paul
Pete Rynas
like you said the board will have to take a hike pete p.s. i am at a
loss not being able to test the cards elsewhere - bummer , am tempted to
trash this newsgroup on a regular basis until they RMA my board or refund
my money - just a thought but they should have posted or made accessable
literature about issues w/some cards
"Paul" <nos...@needed.com> wrote in message
news:g0gih4$5tc$1...@registered.motzarella.org...
Paul
> thanks again paul , the battle continues , i'll try a few more things then ,
> like you said the board will have to take a hike pete p.s. i am at a
> loss not being able to test the cards elsewhere - bummer , am tempted to
> trash this newsgroup on a regular basis until they RMA my board or refund
> my money - just a thought but they should have posted or made accessable
> literature about issues w/some cards
Some similar video results are mentioned here. Monitor does not "catch"
until reset is pressed.
"...successful Leopard install on ECS 945GCT-M/1333"
http://forum.insanelymac.com/index.php?showtopic=82928&st=120
In the Newegg reviews for your board, there is a phone number for
tech support :-)
"510-226-7333 (option 6)"
http://www.newegg.com/Product/ProductReview.aspx?Item=N82E16813135068
The post at the bottom of the Newegg reviews (1/1/2008 10:46:34 AM), notes
a weirdness with the power LED, followed by video not initializing properly.
There have been motherboard problems in the past, where parts of a motherboard
don't properly recognize that the motherboard is shut down - then on the
next reset sequence, things don't initialize properly.
Paul
Pete Rynas
goofing around and shutting off power switch computer shuts down and green
led ( power on) stays lit when comp is off!!! also as i have stated before
hittting the reset reset switch sometimes brings up a display - so i do have
something in common with others now i have to gather my thoughts and find
the time to figure the rest out , you seem to have many hrs. in this , i
still work a 50 hr. wk , own home , 3 kids ( grown but ) busyy! thanks
again i'll check into some of those - still a tad irate at ECS for no
warning on this board and the PM 800CE 7401 which would not run the Pata and
Satather as advertized and no fix was ever given out ( bios etc. )news:g0jep1$sdn$1...@registered.motzarella.org...
Pete Rynas
computer started ok....with the 7200GS .... on goes the battlenews:g0jep1$sdn$1...@registered.motzarella.org...
Pete Rynas
NOT come up for several reboots...on it goes
"Pete Rynas" <r5...@comcast.net> wrote in message
news:P6CdnR5mbpBTjKvV...@comcast.com...
Pete Rynas
945 GCT-M/1333 , it will not run @ 1333 FSB WITHOUT overclocking the PCI-E
video slot , hence the periodic,at best ,initialization of the 7600GT and
the 7200GS. seems as if they (ECS)should RMA this board and or replace
w/something better for the false advertizing and monster frustration and
time i have (paul also) put into this..............???????news:h7mdnec5KNjawqrV...@comcast.com...
Paul
> finally figured it out - due to false advertizing on the retail box for the
> 945 GCT-M/1333 , it will not run @ 1333 FSB WITHOUT overclocking the PCI-E
> video slot , hence the periodic,at best ,initialization of the 7600GT and
> the 7200GS. seems as if they (ECS)should RMA this board and or replace
> w/something better for the false advertizing and monster frustration and
> time i have (paul also) put into this..............???????
That is some good detective work there :-) It doesn't have to be that
way. All it should take, is a clock generator chip, with relatively
independent clock output reference for PCI Express. On some boards,
they go cheap, and either the PCI Express frequency is wrong
or the SATA bus clock can also be screwed up. Since this has been
a solved problem for some time, I wouldn't even have thought of
it.
Paul
Pete Rynas
did it , pci-e cards just won't take it on startup most of the time. ECS
won't talk, but i am going to press the issue and try to get a board from
them that will work as advertized or a refund??
"Paul" <nos...@needed.com> wrote in message news:g1hk5v$a6$1...@aioe.org...
Paul
> yes, sync/async is very old and i never thought of it - can't believe they
> did it , pci-e cards just won't take it on startup most of the time. ECS
> won't talk, but i am going to press the issue and try to get a board from
> them that will work as advertized or a refund??
You never know - you might find someone at ECS local office,
sympathetic to your situation. But if they did something altruistic,
you can imagine they'd be inundated with requests from other
customers they screwed over. And acknowledgement of mistakes,
is a sure way to get a class action lawsuit filed by the
customers.
Paul
Pete Rynas
PM800 CE where SATA - PATA couldn't run at same time - not as advertized ,
but then again you get what you pay for , but why did they even try it? a
25% overclock of the video bus??!!
"Paul" <nos...@needed.com> wrote in message news:g1qcek$3qu$1...@aioe.org...
Pete Rynas
tried/are doing the same thing - automatically overclocking to 1333 - BUT
some at least have the sync/async thing figured out with a
different(good)bios
"Paul" <nos...@needed.com> wrote in message news:g1qcek$3qu$1...@aioe.org...