- 14 months ago
It would seem I surpassed the time in which I could continue to add onto my previous troubleshooting thread. I'm aware that links to such are prohibited so I'll do my best to summarize and keep this tight.. I will however have to copy/paste my specs as most are old enough I can't make a part list, there is a few minor updates for clarifications and one major update the GPU and a minor update/testing tool a spare PSU
Dell/AlienWare Aurora R4 ALX (2012) yes I know "omg AW." It was a 50/50 with my parents, back when I was still able to work for a birthday after my laptop literally had some of it's capacitors pop..
X79 Platform Dell Custom Board Alienware 07JNH0 if I did the link thing right that has a breakdown of specs but I'll list what CPU-Z says: chipset as "intel Sandy bridge-E Rev-07" & Southbridge as "Intel x79 Rev 06" Bios is the latest A11 as of 2014, there have been no updates to it since.
- Has a Daughter Board that handles all IO/Lighting/Temp-Sensors/Fans/Power-button, Runs off of a Molex, couldn't find a board number but from my loose understanding it's required to make her run at all.
At one point the onboard Realtek networking started being really screwy so I replaced it with a Intel Gigabit CT pci-e x4 adaptor card.
CPU: i7-3930k 6 core 12 thread. Reports @3.2GHz but typically functions between 3.2 to 3.7GHz the 3.7 spikes seem to be single/dual core only, will maintain a 3.499GHz under load, synthetic or game across all 6 cores + threads. I have not overclocked this is it's own boost stuff.
Cooler: Corsair H80i AIO Cooler with Push/Pull Noctua Fans added in after 1 year of ownership replacing the stock AIO. (This AIO is now 4 or 5 years old, I've replaced the paste on it once.. maybe I don't really remember, either way temps are still well under 70c under heavy use and below 40c at idle typically.)
Memory: 16GB (4X4GB sticks) Quad Channel 1600Mhz DDR3 RAM, CPU-Z Reports them as Hyundai Electronics PC3-12800 (800Mhz) with CL11,11,28,39. 1.50V says they can go as low as CL8,8,8,22,30 but only at 609Mhz so I'm assuming based on em reporting as 1600Mhz that they are set to the CL11 setting listed as JEDEC #6
PSU: Dell Precision T5500 875Watt Power Supply 0J556T Hipro, has 80+Silver stamp. These are "server" esk PSUs in that they pull air through them from front and out the back instead of top/bottom then out the back, it has a massive single cable harness modular plug.
OS: currently Windows 10 Pro 1803, was a "upgrade" from Win 7 Pro back when they were offering that for free. (I waited till like the last moment of the offer and the transition was smooth, no errors or bugs besides Intel Rapid Storage Enterprise deciding to just stop but I don't have raid set up anyway.. however my BIOS lists my drives set for RAID rather than AHCI not sure how to change that or if I need to.)
BIOS: Though previously mentioned, it is ver A11 from 2014, no updates from Dell have been provided, the BIOS is UEFI though visually still looks like a Legacy BIOS as it was just before graphical UEFI ones were a thing I believe. It is also very stripped, there are lots of things "hidden" or just do not show up vs images of other BIOSs of X79 boards I've seen.
Storage: NO RAID. 1x 500GB SSHD OS/Boot. 2x 1TB 7200RPM Seagate data drives. 1x 3TB WD Green 5200RMP? Backup-Drive. Only Boot and 1 Data drive run at 6GB/s Sata III. The 2nd Data drive and the Backup are on 3GB/s Sata II due to what the board has.
Monitors: 3 Identical Samsung 21inch SyncMaster 215tw 1680 by 1050. 60Hz (though they tend to show as 59) I only game on the center monitor typically in borderless windowed mode, or a 1600 by 900 window to allow multi tasking.
Monitor connections from GPU to Monitors: Left: Direct DVI-D Dual Link cable from GPU. Center: From GPU to Display - Displayport 1.2 adapter to HDMI → HDMI cable into HDMI switch input → HDMI to DVI-D Dual Link cable into Monitor from Switch out. (This gives full 60Hz operation & is so I can run my PS4 off center monitor as well). Right: HDMI to DVI-D Dual Link cable directly from HDMI port on GPU.
USB Perpetuals: AW Supplied Keyboard, Logitech G600 MMO Mouse, Logitech G13 Keypad, Playstation Silver USB headset adaptor (used with Polk Audio Striker-Pro P1 headset). Actisys IR Wireless Interface Diabetic pump data uploader.
Major-updated component (This was part of a "care package" a friend from discord sent me along with a Corsair TX850w Non Modular)
GPU: EVGA GTX 1070 SC Gaming, ACX 3.0 8GB DDR5(Micron), Dual Fan open heat sink. GPU-Z reports: GP104 Rev A1. Bios 86.04.50.00.72 (this is latest update from EVGA's site in re cooling concerns on this batch of cards. Their free "thermal pad kit" is being sent to me next week, I'd love suggestions on a thermal paste please one with a proper applicator/syringe.) Clocks: Core Base 1595Mhz, Boost: 1785MHz, Memory in GPU-Z is 2002MHz. Single 8 pin PCI-E power connector. It is also running at full PCIe x16 3.0.
Item of note: Though my friend sent me a Corsair TX850w, I can not at this time fit it into my current case, such would require a full lower case tear down (and dell being dell thats a nightmare) & cable modding of the power cables for the drives. Further I have ruled out the Dell PSU's voltages being the issue so I'm not concerned by continuing to use what is already in there. Will explain further in the actual report.
Summary of issue I was having a odd glitch where my system would lock up, all input would stop, screens would go black yet their lighting arrays would stay "on" (it was creepy), monitors would fully go off, system would come back up like nothing happened, no errors or anything. This glitch has not presented itself again at all since mid/late December till now, before changing out the GPU for the 1070.
Two friends suspect it was my GTX 680, another my Motherboard and another the PSU. During talking within a discord for a logistics game I'm part of, they had me run a bunch of tests, one of which was checking the voltages.
I already had OCCT, HWMonitor and HWinfo64 installed so I used all three yet all three reported the same.
+3V was within margen 3.264V min to 3.312V high typically.
+5V raised flags showing lows of 3.742V to high as 3.826V (Google research suggests people freakout over PSU +5V going under 4.5V or so.)
+12V also raised flags: it pumps out anywhere between min of 14.950V to as high as 16.192V
The update As I mentioned in the part list, I was sent a "care package" by a kind soul from that discord group consisting of the EVGA 1070 SC ACX 3.0 & a Non Modular Corsair TX850w PSU. both are in excellent condition though I will note:
The GPU is part of the group that had "concerning temperatures" on the power and memory, I've already done the BIOS update as suggested by EVGA's page on the mater and out of an abundance of caution have requested their free Thermal Pad Mod Kit.
- I would appreciate suggestions on a good thermal paste as the kit, from video's I've seen, comes with stuff in basically a ketchup packet and I'd prefer something with a proper applicator/syringe due to having tremors that make my hands a bit shaky at times.
Looking up the PSU the newer ones of the similar branding have 80+ ratings & stamps, this one is old enough it doesn't.
- I had it tested at my local shop who used Thermaltakes PSU tester on it and it checked out fully within the zone across the board.
Testing the PSU & Mobo Because of the Dell-ness of the case I could not just plug everything into the Corsair PSU so instead I opted a rather Frankenstein dual PSU test allowing me to Isolate the Motherboard with the Corsair PSU while powering the rest of the system with the Dell PSU (Including the Daughter Board). Tests were carried out with both my GTX 680 & the new-to-me GTX 1070 one at a time.
Voltages did not improve at all regardless of configuration.. (Corsair powering just the Mobo, or Mobo & GPU.)
- +5V Rail still sagged as low as 3.6 and a high of 3.9
- +12V Rail still was drawing at minimum 14.8V to highs of 16.2V
Conclusion & Sought advice This sadly leaves me to believe based on googling the troubleshooting logic for Voltages is that if it's not the PSUs then it is the Motherboard.. would I be correct in this flow of logic?
If this is the case & with how expensive x79 LGA2011 Sandy/Ivy-Bridge-E boards still are (and they are all "used".) Would it be best to look at building a new system? If so, should I look into replacing the AIO I currently have with it's age? it's about 4 or 5 years old now but seems to be operating within spec.
I was also considering NZXT's GPU hybrid cooler bracket, considering the temperature concerns of the card.. but as it's not a iCX card it also does not show the VRM/Memory Temperatures.. so I only have GPU chip temp showing up in EVGA Precision.. not upgrading the card gives me a lot more leeway in budget but I wonder if doing such a hybrid cooler would be unnecessary even if it gave peace of mind..
- Note: I am from BC Canada so prices are what they are here.. which can be ouch.. for example the card I was sent would have cost me 600 CAD or so brand new after tax and shipping likely at the current lowest listed price from the part picker link I provided at the time of posting: Memory Express 589.99.
Again I kindly also request pointers to some good thermal paste for the GPU..
I appreciate your time, even if all you did was read this. Warm regards. ~Istolla