Welcome Guest ( Log In | Register )

Bump Topic Topic Closed RSS Feed

Outline · [ Standard ] · Linear+

 WTA Graphics card problem, confusing R9 290 can't be detected

views
     
TSasyrafwersh
post Mar 20 2017, 08:45 PM, updated 8y ago

New Member
*
Junior Member
14 posts

Joined: Jan 2017


The story began as couple of days ago i bought used R9 290 Powercolor from bboyspirit by postage. i plugged in the card and no output by the card or presence of it in device manager.

assumptions:

1. gpu balls was cracked by postage and so it is DOA
2. seller is a scam . lol
3. power not sufficient
4. integrated graphics was the display output
5. different os

all the assumptions were wrong as it is running just fine by plugging in eselai comp. plus i have 2x R9 290 reference card running just fine in my desktop.

methods i've tried:

1. switched pcie lanes
2. switched display outputs
3. switched power connector
4. any connections have been assured connected properly
5. run the card primarily and one card only at a time (not psu prob)
6. run just fine in other's pc (same os)
7. tried with pcie riser with 2 other cards installed, it tried to install the driver but ended up blue screen . after that it POST
and gives black screen after windows loading. (this method i didn't investigate further, possible solution?)

the only suspect i am looking at is my motherboard. any suggestions?

My pc specification :

Motherboard Model: MSI Z68A-GD80 (B3) (MS-7672)
CPU: Intel Core i7 2700K (LGA 1155)
BIOS : American Megatrends 1/9/2013 V20.1 (Latest for mobo)
OS : Windows 7 Ultimate x64 Bit
RAM : 16 GB
PSU : Acbel 1000W
GPU : 2x R9 290 ref
Radeon Software Version : 16.12.2

R9 290 Powercolor : Attached Image
Mobo : Attached Image

My current Build 2x R9 290 Ref (working) : Attached Image

With 3rd GPU Powercolor ( 3rd GPU undetected ) : Attached Image
btw i did plugin the PCIE power connector to the gpu . same
the 3rd GPU connected via PCIE Riser with molex power . PCIE X1 to X16 via USB

Update :

Somehow the bluescreen of death is a bug caused by the program and is in loop as in the picture :

Attached Image

it happened after it tries to install the vga drive which is the undetected gpu
after i removed the gpu , start windows , and restart with the undetected gpu ,
device manager finally can detect it with problem (of course) . The error as in the picture :

Attached Image

it says that windows has stopped the device because it causes problem with Error Code 43
hence can i say that the software of the gpu is the cause of this problem? is it related with the os?




This post has been edited by asyrafwersh: Mar 23 2017, 05:07 PM
bobowyc
post Mar 20 2017, 08:53 PM

You are who you are.
******
Senior Member
1,888 posts

Joined: Apr 2013
From: Petaling Jaya


So.. Means youre running them at crossfire?
TSasyrafwersh
post Mar 20 2017, 08:59 PM

New Member
*
Junior Member
14 posts

Joined: Jan 2017


QUOTE(bobowyc @ Mar 20 2017, 08:53 PM)
So.. Means youre running them at crossfire?
*
nope . just 1 way . the other 1 for mining purpose at the same time
bobowyc
post Mar 20 2017, 10:04 PM

You are who you are.
******
Senior Member
1,888 posts

Joined: Apr 2013
From: Petaling Jaya


QUOTE(asyrafwersh @ Mar 20 2017, 08:59 PM)
nope . just 1 way . the other 1 for mining purpose at the same time
*
Are the current ones the same brand or different?
graphidz
post Mar 20 2017, 10:07 PM

F.A.T.E
******
Senior Member
1,408 posts

Joined: Jun 2009
From: land of burung kenyalang

device manager? did you check in BIOS?
TSasyrafwersh
post Mar 20 2017, 10:22 PM

New Member
*
Junior Member
14 posts

Joined: Jan 2017


QUOTE(bobowyc @ Mar 20 2017, 10:04 PM)
Are the current ones the same brand or different?
*
different brand , 1 sapphire 1 asus

here hehe : Attached Image
TSasyrafwersh
post Mar 20 2017, 10:31 PM

New Member
*
Junior Member
14 posts

Joined: Jan 2017


QUOTE(graphidz @ Mar 20 2017, 10:07 PM)
device manager? did you check in BIOS?
*
yes i did check in the device manager . no detection . it can barely detected when i connect it with pcie riser . it attempted to install the driver for the display but ended up blue screen . it can post but it give black screen after windows loading screen. i couldnt check the pcie lanes in the bios

This post has been edited by asyrafwersh: Mar 20 2017, 10:36 PM
hotwebs
post Mar 20 2017, 11:20 PM

On my way
****
Senior Member
659 posts

Joined: Jan 2003
From: JB


Show us the event viewer system log and application error log. What's the error code of blue screen?
TSasyrafwersh
post Mar 21 2017, 12:02 AM

New Member
*
Junior Member
14 posts

Joined: Jan 2017


QUOTE(hotwebs @ Mar 20 2017, 11:20 PM)
Show us the event viewer system log and application error log. What's the error code of blue screen?
*
ok i nver thought of looking for the codes for BSOD . anyway here : Attached Image
hotwebs
post Mar 21 2017, 09:09 AM

On my way
****
Senior Member
659 posts

Joined: Jan 2003
From: JB


Are you oc your processor, GPU and RAM? try to reset everything back to default setting.

The error code showed that's the power issue.

Is either OC, PSU's power connector failure or GPU power connector failure, or RAM.

How long you used your PSU? try to monitoring your PSU voltage: +3.3V, +5V, and +12V.

plus or minus within 5%, it should be stable.

Try to use other other type GPU's power connector. example below:
Attached Image

For your Information:
Windows Kernel event ID 41 error "The system has rebooted without cleanly shutting down first" in Windows

This post has been edited by hotwebs: Mar 21 2017, 09:46 AM
TSasyrafwersh
post Mar 21 2017, 03:33 PM

New Member
*
Junior Member
14 posts

Joined: Jan 2017


QUOTE(hotwebs @ Mar 21 2017, 09:09 AM)
Are you oc your processor, GPU and RAM? try to reset everything back to default setting.

The error code showed that's the power issue.

Is either OC, PSU's power connector failure or GPU power connector failure, or RAM.

How long you used your PSU? try to monitoring your PSU voltage: +3.3V, +5V, and +12V.

plus or minus within 5%, it should be stable.

Try to use other other type GPU's power connector. example below:
Attached Image

For your Information:
Windows Kernel event ID 41 error "The system has rebooted without cleanly shutting down first" in Windows
*
i've used the PSU for 4 years. bad eh?
no oc or whatsoever . can i say that it is a bug caused by the diff manufacturers and so forth?

This post has been edited by asyrafwersh: Mar 21 2017, 03:39 PM
hotwebs
post Mar 21 2017, 05:57 PM

On my way
****
Senior Member
659 posts

Joined: Jan 2003
From: JB


Eh.... correct me, if i am wrong...

PSU Calculator

Estimated voltage around 1024w, haven't put in BitCoin mining shocking.gif , and average running 8 hour per day only. you can play around to the page.

Refer photo to save 1000 words.

Attached Image

Attached Image

Attached Image


* Easier way to verify it, just get another 550w psu and direct power the R9 290. then you will get the answer.

This post has been edited by hotwebs: Mar 21 2017, 06:11 PM
TSasyrafwersh
post Mar 21 2017, 07:46 PM

New Member
*
Junior Member
14 posts

Joined: Jan 2017


QUOTE(hotwebs @ Mar 21 2017, 05:57 PM)
Eh.... correct me, if i am wrong...

PSU Calculator

Estimated voltage around 1024w, haven't put in BitCoin mining shocking.gif , and average running 8 hour per day only. you can play around to the page.

Refer photo to save 1000 words.

Attached Image

Attached Image

Attached Image
* Easier way to verify it, just get another 550w psu and direct power the R9 290. then you will get the answer.
*
Aha i do know tht it exceeded the wattage but that is only
For normal user who do crossfire and fully load the gpu with normal algorithm for gpu. But for mining purpose each R9 290 load about 200W with power efficient enabled

Like i said i did try the undetected card running just 1 way and 2 way still the same

This post has been edited by asyrafwersh: Mar 21 2017, 07:49 PM
hotwebs
post Mar 21 2017, 10:35 PM

On my way
****
Senior Member
659 posts

Joined: Jan 2003
From: JB


QUOTE(asyrafwersh @ Mar 21 2017, 07:46 PM)
Aha i do know tht it exceeded the wattage but that is only
For normal user who do crossfire and fully load the gpu with normal algorithm for gpu. But for mining purpose each R9 290 load about 200W with power efficient enabled

Like i said i did try the undetected card running just 1 way and 2 way still the same
*
try add one more PSU, no harm to give it a try. undetected the card doesn't meant not using the power and when we select the bit coin mining the wattage did increase a lot wattage.

some more your PSU is Acbel R9 900W with 100W continuously overclocking and used 4 years, but are you running 24 hours? if yes, i will suspect that might cause the power issue, due to PSU capacitor life span reduce.

This post has been edited by hotwebs: Mar 21 2017, 11:12 PM
Richard
post Mar 21 2017, 11:03 PM

Regular
******
Senior Member
1,083 posts

Joined: Jan 2003
From: Kota Kinabalu


I'm wondering since when did the tech section get this hard core..

I thought i knew a lot and I'm totally lost by this..
TSasyrafwersh
post Mar 22 2017, 01:02 AM

New Member
*
Junior Member
14 posts

Joined: Jan 2017


QUOTE(hotwebs @ Mar 21 2017, 10:35 PM)
try add one more PSU, no harm to give it a try. undetected the card doesn't meant not using the power and when we select the bit coin mining the wattage did increase a lot wattage.

some more your PSU is Acbel R9 900W with 100W continuously overclocking and used 4 years, but are you running 24 hours? if yes, i will suspect that might cause the power issue, due to PSU capacitor life span reduce.
*
haha sorry if i've offended you . but yeah i am planning on adding more psu as i might be adding two more R9 290 . and yes i do realized the psu's caps have been degrading
over the years , i am looking to replace the main psu but the psu has been utilized like 400 W only for 3 years and for almost maximum load just this week lol. Talking about
adding one more psu/replacing and adding more gpu has been at my bucks limit so yeah i am waiting for the mining profit and expenses get even. But thanks for your opinions bro
as for the power concern is important as it has been signalling the power consumption is at limit.

For now i am going to troubleshoot and investigate further this matter later on as i am on a quite hectic week. I will update about the troubleshooting and I do hope people here
gives their opinion and help me regarding the problem so that more people could solve this matter easily in future whenever they encountered this matter.

TSasyrafwersh
post Mar 22 2017, 01:05 AM

New Member
*
Junior Member
14 posts

Joined: Jan 2017


QUOTE(Richard @ Mar 21 2017, 11:03 PM)
I'm wondering since when did the tech section get this hard core..

I thought i knew a lot and I'm totally lost by this..
*
lol i couldnt agree more haha, it is indeed the most odd and confusing thing ever i've encountered about gpu installation.
goldfries
post Mar 22 2017, 01:07 AM

40K Club
Group Icon
Forum Admin
44,409 posts

Joined: Jan 2003




QUOTE(hotwebs @ Mar 21 2017, 05:57 PM)
Eh.... correct me, if i am wrong...

Estimated voltage around 1024w, haven't put in BitCoin mining shocking.gif , and average running 8 hour per day only. you can play around to the page.
Yes you're wrong. Recommended wattage is 1024w.

Actual draw of 974w or whatever the number is happens ONLY if all cards are maxed out and CPU also maxed out.

R9 390 is a power hogger, around 275W if I'm not mistaken on full load. Added with system, probably around 80W because CPU will have work to do.

It'll be running at around 360W on single GPU setup.

Add 2 more cards, it's another 560W on the system or so. You can round each of the card to about 300W, so it comes to that estimated figure that I mentioned.
pspslim007
post Mar 22 2017, 01:10 AM

On my way
****
Senior Member
699 posts

Joined: Apr 2010
From: Malaysia


bro, you need some cleaning to do, that damn dusty man, it hurts my soul. sad.gif
TSasyrafwersh
post Mar 22 2017, 01:22 AM

New Member
*
Junior Member
14 posts

Joined: Jan 2017


QUOTE(goldfries @ Mar 22 2017, 01:07 AM)
Yes you're wrong. Recommended wattage is 1024w.

Actual draw of 974w or whatever the number is happens ONLY if all cards are maxed out and CPU also maxed out.

R9 390 is a power hogger, around 275W if I'm not mistaken on full load. Added with system, probably around 80W because CPU will have work to do.

It'll be running at around 360W on single GPU setup.

Add 2 more cards, it's another 560W on the system or so. You can round each of the card to about 300W, so it comes to that estimated figure that I mentioned.
*
according to miners . rated using Kill A Watt device : Attached Image

1x R9 290 = <250 W
3x R9 290 = < 750 W
Total around 900W

i can apply low power consumption and underclock to bring down the wattage further by changing the algorithm mode :

this is snippet from the miner website :

-a Select algorithm:
"-a 0" (default) means autoselection.
"-a 1" fastest mining mode for 290/290x and 270/270x cards.
"-a 2" low power consumption mode for 290/290x cards. A bit less speed (1%) but less power consumption (10-15%).
"-a 3" fastest mining for old cards (6xxx and 5xxx).
"-a 4" fastest mining for recent cards (280x, 270/270x etc).
You can also specify mode for every card separately, for example, "-a 1234".
It is very important for good mining speed to select correct "-a" and "-h" values for your cards.
For old cards 5xxx and 6xxx use "-a 3" mode only, otherwise you can get "LowDifficultyShare" errors from pool.

2 Pages  1 2 >Top
Topic ClosedOptions
 

Change to:
| Lo-Fi Version
0.0168sec    0.35    6 queries    GZIP Disabled
Time is now: 29th March 2024 - 05:29 PM