Quote from: ctmk01 on April 27, 2017, 07:01:39 PM

Quote from: carloshd on April 27, 2017, 12:03:01 PM

Quote from: EvilC0P on April 21, 2017, 08:35:45 PM

Quote from: yoonline on April 21, 2017, 07:20:38 PM

Quote from: mafaz on April 20, 2017, 02:49:06 PM

Hi guys....

I have (had Sad

) 42 MSI rx 470
12 Sapphire Nitro+ 470
12 Sapphire reference 470 cards.

Latelly ... all off them started falling like leaves. I mean they either die or burn up.

6 msi - 3 sapphire nitro and 2 sapphire reference cards died all during past 3 weeks....

Anyone else have that many breakdowns lately?

I dont even know what to blame now. Bios mods... motherboards.... PSU or claymor v9

Sapphire reference cards all worked for 5 months with only 2 of them die before this nightmare....
Sapphire Nitro+ never broke for 5 months before this....
I recently purchased 30 MSI cards and 6 already dead in like 3 weeks.... they tend to bring out smoke while sapphire cards just die.

My rigs are 6xGPU all (used to be.... LOL) motherboards are BTC PRO and Z97 fatality.
PSU are Be quiet pro 2x 1200 and and Platimaxe's 1600.

I used to use -96 core woltage in afterburner.

I started thinking that might be the reason. But i changed it to 0 and 2 more cards died. SO im pretty much lost.

Any ideas are welcome.

Or maybe someone else has unusual high rate of failing cards recently?

thats is creepy... can someone confirm these die rates? maybe you have a bad setup of power distribution or yours psu are failing

i had one of my msi rx480 sparkle and burn in front of me when booting up my system after a power outage.
realized the pcie riser had some burn on it, it was a version 003. when running hwmonitor the voltage were spiking to like 500000 ! at first though it was a glitch but then i thrwe them all away and got new risers, v007 this time and hwmonitor reports normal voltage value, no more spiking to 500000+ .

So i would suggest double checking your pcie risers, they may be the faulty part causing that problem

You really were right, 2 "rise 003 version" burned, I have 2 "rise 006 version" is perfect.

4 x RX 480

https://ibb.co/hNTdJ5
https://ibb.co/ewRXd5
https://ibb.co/g64JJ5

fyi,
http://www.playtool.com/pages/psuconnectors/connectors.html#peripheral

the 4 pin peripheral power cable only good for 5 amps that's 60w for 12v

and SATA power cable side only rated for 4.5 amps

i am not surprised it would melt over 75w constant load overtime. (and you should know RX480 will draw more that that on pci-e power draw.)

you need to get pci-e 6 pin powered raisers (do not use the provided sata to pci-e 6pin power cable with raisers if any!). and do not use extension to keep in specification with standard 18awg wires.

only safe way is to get PSU with adequate PCI-E 6pin power outlet.

I am using VER 006C 1703 btw

often monitor temperature with infrared thermometer gun on the connectors to make sure that no loose connection that cause electrical sparking that will cause heat up and melt your parts.

The 75 watts is not supplied by the powered riser. The main power for the GPU comes from the six or eight pin 12 volt power connector specifically designed for high end graphics cards. The power to the riser replaces the supplementary power that would be provided by the PCI-E slot which is not much power at all. The use of powered risers is to relieve the stress from the motherboard of supporting PCI-E bus power to too many cards, but this is by no means the main source of the graphics card power source. I am using risers with a molex to SATA adapter with SATA power from the PSU. I check the temps and the wires stay below 30C, which is about my room temp where I am located. They should not melt or get hot unless very thin, cheap wires was used to make them.