That was weird I just RMA'd my card last week and my replacement was is still on Hold. Im still waiting for their explanations on why am I experiencing multiple failures and why alot of people are experiencing this as well. I dont know if they would give me a valid reason on this. This is very annoying now.
My leadtek (winfast a400 GT TDH) geforce 6800gt is doing crazy things
- Thread starter dr_jared
- Start date
- Status
Well as I've said I have no problems with this card as yet, but I suppose it's possible I could at any point.
If this card ever did start showing lots of artifacts even on the bios screen, I think I would first look at how it was seated in the agp slot, how securely it was screwed in and so on. Maybe how well the whole motherboard was mounted in the case could be a factor for that matter.
I'd also want to check how well the heatsink was mounted on the card. I'd probably take it off, make sure there was good contact and lots of grease or pad in the right places, and make sure it was put back together so it would stay in place.
Maybe you wouldn't want to do that if you were looking at a new card you could just RMA or replace under warranty, but anyway I'm hoping this card here will last long enough so that won't be an issue and I like to avoid customer service and mailing stuff around generally as rule in any case.
Well it's tough wasting 300 bucks on stuff like this. Hope nobody has that happen.
If this card ever did start showing lots of artifacts even on the bios screen, I think I would first look at how it was seated in the agp slot, how securely it was screwed in and so on. Maybe how well the whole motherboard was mounted in the case could be a factor for that matter.
I'd also want to check how well the heatsink was mounted on the card. I'd probably take it off, make sure there was good contact and lots of grease or pad in the right places, and make sure it was put back together so it would stay in place.
Maybe you wouldn't want to do that if you were looking at a new card you could just RMA or replace under warranty, but anyway I'm hoping this card here will last long enough so that won't be an issue and I like to avoid customer service and mailing stuff around generally as rule in any case.
Well it's tough wasting 300 bucks on stuff like this. Hope nobody has that happen.
Hehehe Well I did that to the last 2 cards they replaced for me. The first card got artifacts and was sent back for RMA so I got the replacement. The replacement after few months smoked my motherboard!! I read some reviews on other cases in Newegg the card did that to them too. Then it broked again so I send in for RMA again and give me another new one. With a new board diff model and new chasis on my computer the card run for few months again then it poofed!! Not working again. I was working on an AC room with also a very high wattage reputable PSU. My other cards wont have this problem. Bios of mboard and vcard had been updated as well and all the possible things. Now I have sent this back to leadtek to check on it again and was found dead.. They will want to replace by Im asking them that they owe me some explanations because alot of people are hving this problem! And replacing the card is not simply resolving my problem. I wonder what would they do about it. I hope they would send me something would not repeat this nightmare due to this is used for work not gaming pleasures.
I really dont know. Maybe because Im in USA? and my card didnt even make it to a year. Its always under one yr when it broke down. I guess that is why they still entertain my card. I really dont know their policy on International orders though. Did you check the rules of their internationals? Maybe you should them on email about it.
atomicplayer
Posts: 30 +0
Their warranty policy makes me worry because it says that its one year since the first card... the one year warranty doesn't restart with a new card... so if my second card (that I'll probably receive mid-january) is still faulty, Ill have to send the card back to them quickly and hope for the fourth one to work because my warranty ends on february 25th
sry if my english s*cks.. im from quebec, canada
sry if my english s*cks.. im from quebec, canada
atomicplayer
Posts: 30 +0
according to the manual, its a one year warranty. so leadtek should take their card back within the first year... the bad thing is that the warranty doesn't restart if we get a replacement card. so if for example their card breaks each month (which is a fact...), after one year and a month, we won't be able to use it because it will be impossible to send the card back to them again.
My card is like this. When I recieve the new RMA card that its a nother 1 yr warranty on it. So everytime there is a new replacement the warranty period resets as well. I read a couple of issues on Leadtek I think the whole 6800xx line is a design flaw Im really not sure about it. But according to my research it is somehow like that. So replacing with another leadtek of any 6800xx would be the same problem.
atomicplayer
Posts: 30 +0
ok so its a good news... according to the manuel it doesn't work like that but if you say that it resets for your card, it should for our card as well.
garlicbread
Posts: 9 +0
well I've had mine back recently and everything seems to now be okay
I think most retail places will only offer a 1 year guarantee at the most
For the Ultra cards I believe the guarantee with leadtek is 3years
For non-ultra cards I'm not sure
(I have noticed for the AGP models I think they only sell the Ultra ones now)
have you tried mailing support@leadtek.nl which is the european one
they'l probably ask for a copy of the reciept via mail / fax or email (I just scanned mine in and attached it has a jpg)
I think most retail places will only offer a 1 year guarantee at the most
For the Ultra cards I believe the guarantee with leadtek is 3years
For non-ultra cards I'm not sure
(I have noticed for the AGP models I think they only sell the Ultra ones now)
have you tried mailing support@leadtek.nl which is the european one
they'l probably ask for a copy of the reciept via mail / fax or email (I just scanned mine in and attached it has a jpg)
atomicplayer
Posts: 30 +0
Ive never had any problems with my ati radeon 9200 and ive been using it since spring 2004. and it works 24/7...
my 6800 gt lasted 5 months.... :approve:
my 6800 gt lasted 5 months.... :approve:
Yeh PC Nerd that is really constructive criticism...
Anyway, it sounds like Leadtek just messed up with this range of cards, so I imagine nVidia aren't too happy with them.
I for one will get a different make next time...or even ATI. I have a temporary 9700 Pro in my PC, and while it may not be quick it is quiet and actually reliable, neither of which my 6800 is!
Anyway, it sounds like Leadtek just messed up with this range of cards, so I imagine nVidia aren't too happy with them.
I for one will get a different make next time...or even ATI. I have a temporary 9700 Pro in my PC, and while it may not be quick it is quiet and actually reliable, neither of which my 6800 is!
PC nerd is off topic....
Anyway I think only LEADTEK 6800xxx series has problems. Other manufacturers dont encounter this problem. I think this is a whole design flaw on the card. I hope leadtek would do something about this. But I dont think they would invest to redo the design of the 6800xxx series since 7800 is already taking over and was being shipped out already. I hope they give me another series of card instead of this nightmare Im encountering. It is my 3rd card already in 2 yrs i think.
Anyway I think only LEADTEK 6800xxx series has problems. Other manufacturers dont encounter this problem. I think this is a whole design flaw on the card. I hope leadtek would do something about this. But I dont think they would invest to redo the design of the 6800xxx series since 7800 is already taking over and was being shipped out already. I hope they give me another series of card instead of this nightmare Im encountering. It is my 3rd card already in 2 yrs i think.
garlicbread
Posts: 9 +0
well mines just crapped up yet again same sort of interferance as before
although not quite as bad, both outputs are still working at the moment but anything at a high refersh rate and the card stops responding
WARNING WARNING
I don't know if this is the cause but it's the last thing I did before mine crapped up completly (again)
I'd been messing around with the Bios settings as I was getting hardlocks every 30 minuites or so within windows, I tied the problem down to a USB Bluetooth dongle that was causing the problem
anyway one of the settings I'd altered was the Hypertransport CPU to AGP clock rate
in the past I'd always had this set to 1000Mhz the top setting
although recently I'd toned it down to 600Mhz to try and stop the hardlocks (which were actually being caused by the BT dongle)
After figuring out that it wasn't the card I'd upped the Hypertransport setting to 800Mhz
Boot into Windows
screen goes black (no signal)
Reset
ARRRRSE!!!
I think the combination of by A8V board with my new dual core CPU somehow affected the Hypertransport speed, which in turn affected the card
From looking at my own card I can tell they've blobbed some more of that heat sink stuff onto the GPU, but I know they couldn't have replaced the GPU
I suspect that it's one of the surface mounted cap's that's overloaded due to the HyperTransport speed and caused a timing issue on the card
the interferance is on both the digital and analogue outputs which suggests that it's probably something ether in the power supply or digital timing circutry
So in otherwords for those users that get they're cards back don't set the Hypertransport speed to anything more than 600Mhz as this could be the cause ( I can't be sure)
Back to my 4600TI
time to start looking for a new MB with PCI-E and perhaps a 7800
(Linux support for ATI is okay but I don't think the best in the world)
although not quite as bad, both outputs are still working at the moment but anything at a high refersh rate and the card stops responding
WARNING WARNING
I don't know if this is the cause but it's the last thing I did before mine crapped up completly (again)
I'd been messing around with the Bios settings as I was getting hardlocks every 30 minuites or so within windows, I tied the problem down to a USB Bluetooth dongle that was causing the problem
anyway one of the settings I'd altered was the Hypertransport CPU to AGP clock rate
in the past I'd always had this set to 1000Mhz the top setting
although recently I'd toned it down to 600Mhz to try and stop the hardlocks (which were actually being caused by the BT dongle)
After figuring out that it wasn't the card I'd upped the Hypertransport setting to 800Mhz
Boot into Windows
screen goes black (no signal)
Reset
ARRRRSE!!!
I think the combination of by A8V board with my new dual core CPU somehow affected the Hypertransport speed, which in turn affected the card
From looking at my own card I can tell they've blobbed some more of that heat sink stuff onto the GPU, but I know they couldn't have replaced the GPU
I suspect that it's one of the surface mounted cap's that's overloaded due to the HyperTransport speed and caused a timing issue on the card
the interferance is on both the digital and analogue outputs which suggests that it's probably something ether in the power supply or digital timing circutry
So in otherwords for those users that get they're cards back don't set the Hypertransport speed to anything more than 600Mhz as this could be the cause ( I can't be sure)
Back to my 4600TI
time to start looking for a new MB with PCI-E and perhaps a 7800
(Linux support for ATI is okay but I don't think the best in the world)
atomicplayer
Posts: 30 +0
when you talk about hypertransport, you mean the front side bus or what ? my fsb has always been set to 800 mhz before and after the leadtek's card crash
garlicbread
Posts: 9 +0
I don't think it's the FSB, it's not the speed of the CPU
I think its the speed with which the CPU talks to the AGP port
on the A8V it's under
CPU Configuration -> HyperTransport Configuration
there are 3 options
HT Frequency 200Mhz / 400Mhz / 600Mhz / 800Mhz / 1000Mhz
HT Data Width (UpStream) ether 8bit or 16bit
HT Data Width (DownStream) ether 8bit or 16bit
I think on some other motherboards it show up as some from of multiplier setting instead of a frequency setting
EDIT
I think this is probably AMD only
it might be different for the P4
I think its the speed with which the CPU talks to the AGP port
on the A8V it's under
CPU Configuration -> HyperTransport Configuration
there are 3 options
HT Frequency 200Mhz / 400Mhz / 600Mhz / 800Mhz / 1000Mhz
HT Data Width (UpStream) ether 8bit or 16bit
HT Data Width (DownStream) ether 8bit or 16bit
I think on some other motherboards it show up as some from of multiplier setting instead of a frequency setting
EDIT
I think this is probably AMD only
it might be different for the P4
garlicbread
Posts: 9 +0
Just found this on Toms Hardware Guide
http://www.tomshardware.com/2005/12/15/dual_graphics_platforms/page2.html
"How To Overclock The AMD Platform
Unlike Intel-based computers, an AMD system has no front side bus. Instead, HyperTransport is used to link the processor to the chipset. It has a system base clock of 200 MHz; a multiplier of 5 is applied in order to hit the 1,000 megatransfers per second for the HyperTransport protocol (which corresponds to 1,000 MHz here).
However, since the memory controller is part of the processor, there are fewer bandwidth requirements between the CPU and the chipset. A similar Intel Pentium 4 or Pentium D based system has to use the front side bus for every memory access, which adds latency to each operation. This all means that the resulting system or HyperTransport speed is far less important than one would believe with AMD based systems.
Despite this, a flexible overclocking range is still very important, in order to overclock either the CPU or the main memory, because the goal is to find the best performing combination of system speed, processor multiplier and memory clock speed. These tuning measures result in noticeable performance boosts."
and this
"Increasing The System Clock Speed
Every Athlon 64 motherboard other than the lowest cost models will offer you a BIOS option to alter the system clock speed. It defaults to 200 MHz with all Socket 939 processors, but clock speeds of almost 400 MHz can be reached today. However, this has more of an effect on other components than it has on performance, so be careful here.
Raising the system clock will automatically increase the HyperTransport speed as well, since as mentioned before, this is a multiple of the system clock speed. Look for an option called 'HTT Link' or similar. It should either be set to '1000 MHz' or to 5X, both of which mean that HyperTransport runs at five times the system clock speed. Possible options usually are 800 MHz (4x) and 600 MHz (3x). Whenever you overclock the system clock, make sure you reduce the HTT multiplier accordingly, because for the sake of system stability, you should not exceed the 1,000 MHz figure by much."
I had the BIOS setting in the A8V for overclocking set to AI originaly (which I think means just a little bit)
and then to None or Standard (for Standard speed) after the first card failure
I've not overclocked my CPU at all during the second fail which suggests that a design flaw in the card leaves it unable to handle the higher transfer rates that really it should be able to handle (assuming this is what caused my second failure)
I'm now planning to no-longer use the card and to get a A8N-SLI Premium instead with a 7800 PCI-E, and to re-use the old board with an older card as a server
I definitaly won't be getting a Leadtek card again, I think the Gigabyte ones may be the best at the moment though
http://www.tomshardware.com/2005/12/15/dual_graphics_platforms/page2.html
"How To Overclock The AMD Platform
Unlike Intel-based computers, an AMD system has no front side bus. Instead, HyperTransport is used to link the processor to the chipset. It has a system base clock of 200 MHz; a multiplier of 5 is applied in order to hit the 1,000 megatransfers per second for the HyperTransport protocol (which corresponds to 1,000 MHz here).
However, since the memory controller is part of the processor, there are fewer bandwidth requirements between the CPU and the chipset. A similar Intel Pentium 4 or Pentium D based system has to use the front side bus for every memory access, which adds latency to each operation. This all means that the resulting system or HyperTransport speed is far less important than one would believe with AMD based systems.
Despite this, a flexible overclocking range is still very important, in order to overclock either the CPU or the main memory, because the goal is to find the best performing combination of system speed, processor multiplier and memory clock speed. These tuning measures result in noticeable performance boosts."
and this
"Increasing The System Clock Speed
Every Athlon 64 motherboard other than the lowest cost models will offer you a BIOS option to alter the system clock speed. It defaults to 200 MHz with all Socket 939 processors, but clock speeds of almost 400 MHz can be reached today. However, this has more of an effect on other components than it has on performance, so be careful here.
Raising the system clock will automatically increase the HyperTransport speed as well, since as mentioned before, this is a multiple of the system clock speed. Look for an option called 'HTT Link' or similar. It should either be set to '1000 MHz' or to 5X, both of which mean that HyperTransport runs at five times the system clock speed. Possible options usually are 800 MHz (4x) and 600 MHz (3x). Whenever you overclock the system clock, make sure you reduce the HTT multiplier accordingly, because for the sake of system stability, you should not exceed the 1,000 MHz figure by much."
I had the BIOS setting in the A8V for overclocking set to AI originaly (which I think means just a little bit)
and then to None or Standard (for Standard speed) after the first card failure
I've not overclocked my CPU at all during the second fail which suggests that a design flaw in the card leaves it unable to handle the higher transfer rates that really it should be able to handle (assuming this is what caused my second failure)
I'm now planning to no-longer use the card and to get a A8N-SLI Premium instead with a 7800 PCI-E, and to re-use the old board with an older card as a server
I definitaly won't be getting a Leadtek card again, I think the Gigabyte ones may be the best at the moment though
Hello all. I joined to boards to post my experience with this card. I, like many of you, order this card from NewEgg. It was a refurb, and I got it on the 16th. I plugged it in, powered on, and low and behold had the exact same problems described here. I tried many, many different things to get it working (many of which were posted here by other members). I was just about to submit the RMA to NewEgg, but recalled that one person said the weight of the card may be causing problems with the contacts.
I didn't think that was the problem because I had even tried laying my PC sideways before powering on, so the weight of the card shouldn't have been a factor. I figured it couldn't hurt anything to give it one last try before RMAing it though, so I powered it on while holding the plastic part of the card up with my hand. Voilà, it worked perfectly. I went through my misc. electronics box to find something the right height, and it turned out a miniDV tape fit perfectly under the plastic edge of the card without blocking the fan to hold the card at the right angle for the contacts to be secure. The tape holds the card about 2cm higher than it would sag without the tape. If I take the tape out, all the problems come back. I put the tape back in and they go away.
For the past two days I have been running the card hard, and it has performed beautifully. It is just the gigantic size that caused the problem in my case. It is worth seeing if it will fix your problems as well.
Check out the attached photos for my solution. It actually is very secure in the case, but I put a couple pieces of packing tape on it to ensure it would stay put.
I didn't think that was the problem because I had even tried laying my PC sideways before powering on, so the weight of the card shouldn't have been a factor. I figured it couldn't hurt anything to give it one last try before RMAing it though, so I powered it on while holding the plastic part of the card up with my hand. Voilà, it worked perfectly. I went through my misc. electronics box to find something the right height, and it turned out a miniDV tape fit perfectly under the plastic edge of the card without blocking the fan to hold the card at the right angle for the contacts to be secure. The tape holds the card about 2cm higher than it would sag without the tape. If I take the tape out, all the problems come back. I put the tape back in and they go away.
For the past two days I have been running the card hard, and it has performed beautifully. It is just the gigantic size that caused the problem in my case. It is worth seeing if it will fix your problems as well.
Check out the attached photos for my solution. It actually is very secure in the case, but I put a couple pieces of packing tape on it to ensure it would stay put.
Attachments
atomicplayer
Posts: 30 +0
hmm it seems a valid reason why we have having trouble with this card. and its maybe why I started having problems when I tried the dvi plug (the ones who have read my first post will understand). I probably moved the card a bit just enough to impede the contact between the card and the agp slot.
I'll try smeg's solution when I receive my card in january..
I'll try smeg's solution when I receive my card in january..
Duff card
Hi there folks new to the board after searching for the very information this thread has provided me with - many thanks. Bought the said mentioned card of a mate who had no probs for about a year when I bought it from him. I hadn't used it until a month ago (waiting to build rest of rig). So I'm well pissed off with it considering I've only used it for about a total of 6hours in 1 month - hardly excessive and my mate who I bought it off didn't abuse it. Thought there could be a glimmer of hope with the minidv prop solution but I've got mine in a desktop case so it sits in square with no weight pulling it out. Oh well live and learn, now waiting for the XFX 6800GS 256MB extreme cards to ge to the UK so I can finally start using my machine.
Hi there folks new to the board after searching for the very information this thread has provided me with - many thanks. Bought the said mentioned card of a mate who had no probs for about a year when I bought it from him. I hadn't used it until a month ago (waiting to build rest of rig). So I'm well pissed off with it considering I've only used it for about a total of 6hours in 1 month - hardly excessive and my mate who I bought it off didn't abuse it. Thought there could be a glimmer of hope with the minidv prop solution but I've got mine in a desktop case so it sits in square with no weight pulling it out. Oh well live and learn, now waiting for the XFX 6800GS 256MB extreme cards to ge to the UK so I can finally start using my machine.
- Status
Similar threads
