Jump to content

Aaron1001

Member
  • Posts

    30
  • Joined

  • Last visited

Awards

This user doesn't have any awards

3 Followers

Recent Profile Visitors

908 profile views

Aaron1001's Achievements

  1. Oh wow. I haven't been back on the forum in quite awhile. It's super cool to see people take my work and improving on it. I know my build log kind of just ended abruptly (college started), but I did end up finishing the PC, and I've attached a photo of the finished product. I ended up using it for a year but sold it once it became apparent that a gaming laptop is much more suited for the on the road work that I do in college. But for that one year, it was a fantastic beast of a machine to use. @DommPaone just a few mistakes I made and that I want to give you a heads up on. 1. acrylic sheet has terrible tolerances from sheet to sheet. make sure the slots are wide enough. 2. mineral oil dissolves oil-based thermal paste. you will need some liquid metal thermal paste for both your CPU and GPU. 3. when buying a GPU, have the first fan be as far away from the IO as possible. something like an EVGA card. my Asus card would make splashing sounds if it ramped up to anything more then minimum speed. 4. get a PSU with an always-on fan. helps with circulation. but make sure the fan blades are below the oil level. 5. I don't think I specified in the build log, but when the mineral oil goes from the radiator to back into the tank, there is a T coupling + hard tubing that routes the oil back to the front and shoots it through the CPU heatsink and in front of the GPU fans. It's in my CAD files, but the bends are way too tight. it's much better for it to move forward a few inches. here is a sheet with some thermal numbers. I personally like to run my PCs a little hotter and quieter, but with a less aggressive OC, the numbers fall drastically. If anyone has any questions about mineral oil PCs, I'm more then happy to answer them. I'm super excited to see where you take this project, and wait eagerly to see you finished build!
  2. the idea behind the sheet is that even though the CPU does not get hot enough to melt it, it does become malleable, filling in the microscopic gaps. However, if what you say is true, I will go for the full metal TIMS. I'm sorry if this sounds a bit rude, I have quite a bit of hardware at stake here, but if you have some sort of documentation, or perhaps your own experience that you can share with me, that you make me feel a lot more confident using them. i bit the bullet, and took the risk with the gallium compound. Works like a dream for the CPU. I will warn against using them on ASUS direct CU cards though, since their contact surface has large grooves in it where the heat pipes touch. But my computer is no longer overheating. Thanks!
  3. Even though the application process sounds terrifying, I might just use that for the CPU. That still leaves out the GPU though. hopefully i can find something similar to it. Thanks. I was thinking about that as well. But without weeks of testing, I wouldn't be confident that it wont fallout like regular thermal paste. And if it does fallout, I wont have a computer anymore.
  4. This post has a bit of back story to it, so here is a tl;dr. Has anyone used thin Indium sheets (similar to aluminum foil) instead of thermal paste? what thickness did you use, and what were your results. I have recently built a mineral oil PC (you can see a unfinished build log here), and one week in, I'm already facing problems. The mineral oil loves eating away at my thermal paste, and every day, my temps are climbing higher and higher. The CPU is still under control (but I can see thermal paste fallout on my ram sticks), but the GPU, under load, has gone from the mid 50's to the mid 80's. I know that it is highly unlikely, but it would be amazing if @Slick can chime in and tell me what paste he used for his build that lasted for so long, and if @LinusTech can explain why as seen in the HP Spectre 2016 Review, he has stopped using the mineral oil PC. My best guess would be that there is some rubber in the capacitors on the motherboard that swelled up and caused the capacitor to pop off, or IC diamond gets eaten away much slower them my paste, and only now has started to overheat. Whatever it is, some insight into long term submersion would be great. this site has great detail on mineral oil cooling large servers. A source I really wish i knew about sooner, and one I highly recommend anyone tacking something similar to read. They recommend Indium sheets as a replacement, but there is very little info about this online. But I assume that if it was better then thermal compound, then everyone would be using it. Anyone with experience using this material can give some insight to its performance?
  5. $300 wont get you very far buying new, so for low budget machines, I would highly recommend looking at the used market. Just last week I helped a friend build a used gaming PC for around $500, and he ended up with a computer that runs a i5 3570k and gtx 960, which can run any game with very nice settings at 1080p. a couple suggestions: look on craigslist for older systems. Sandy bridge and ivy bridge systems still have many years left in them, and sell for far less. Barter a lot. Have no shame. you can usually bring the price down quite a bit. ($20-$30). Ask for lower then what you think they will sell at. Sellers will usually meet you in the middle (price is $300, you ask $250, they say $275. done deal) On eBay, for CPUs and GPUs, hunt out parts that have been overclocked. Those usually sell for less. Be warned, items that start at $0.01 attact a lot of watchers, and will usually sell for market price regardless of their low starting price. The only thing I had my friend buy new was a hard drive, but he kinda goofed up and got a 1TB sshd for $100. So really, the system could have been $450. Yes, you are going to have to do a lot more work and research, but if you are willing to put the work into it, you can get a damn good PC for $300
  6. I have a 6800k running at 4.5ghz at 1.385v, and I've run a stress test on it for over 24 hours so I know that it is stable. However, recently, the CPU has started to become unstable. My computer crashed in the middle of gaming. Using AIDA64, and running a CPU test leads to no crashing at all. However, if I do the floating point test (FPU), my computer crashes within seconds. my thermals are within limits (maxing out around 80 degrees). I've already tried to raise the voltage up to 1.45 to no avail. I know i can just dial back the overclock, but does anyone have any ideas to why my OC is suddenly failing after passing a 24 hour stress test, and running stable for weeks. -thanks in advance
  7. thanks for the quick reply. Im a bit of a silence freak, so your method is definitely a no go for me. Currently, I have it running at 2101, and just letting the voltage go higher then it really needs to. I know 13 mhz makes absolutely no difference, but it is upsetting that I pay good money for a card, and it wont do what I tell it to do
  8. I have a ASUS strix 1070, and I've been attempting to overclock it, but I have hit a wall. I know my max overclock (2114), and I know what voltage to stay at (1.075v), but my GPU just doesn't want to cooperate with me. Sometimes, the voltage will jump up to 1.093, and the clock will move up with it to 2124, which will lead to instability. and sometimes, the voltage will drop to 1.050v, and the clock will drop with it. I've put in a old gtx 460, just to make sure I wasn't going crazy, and indeed, the core clock I put in, is the core clock it runs at. The voltage I put in is the voltage it runs at. anyone have any ideas for how i can make my card run at 2114 at 1.075v under load, no more, no less? thanks in advance
  9. It would make sense if i was drawing like 10 watts less then average. But drawing nearly 40 watts less then tdp when overclocked? and nearly 70 watts less then what other people are seeing? that isnt just luck.
  10. I never thought of that, but that does make quite a bit of sense. Never said i wasnt excited about amd being faster clock for clock, just that i wasnt excited by their, intel and AMDs, higher end offerings. I want to see AMD competitive in the mid $100 range as i see that being the best case scenario for everyone.
  11. I have a 6800k, overclocked to 4.5ghz at 1.35v. im only drawing around 105watts according to AIDA64. This seems way way too low. I've checked bios, and I know i am not limiting power draw, and I also know AIDA isnt too far off because my CPU temps are extremely low for a stress test. Anyone have any idea why, and how to solve it? My best guess is my old PSU is limiting me since it is some no name brand without even a 80+ rating. But if that is the case, shouldn't my computer shut down?
  12. because if they overclock their chip. It shows that not only is AMD faster clock for clock, they can also achieve similar clock rates.
  13. I honestly cant tell if you are trolling or if you are serious. But, if you really are serious, I propose a question. Why would they underclock the intel chip and not overclock theirs? If everyone saw a 3.6ghz amd zen beating out a 3.6ghz 6900k, EVERYONE would be losing their shit at how amazing this new chip is. but they didnt. why not?
  14. well, Intels non k cpu's cant go anything above specified... (yes I understand you can mess with the BLCK frequency) but, you are cherry picking my points. I never denied that it cant. just said that it might be able to, might not. i don't think anyone can fault me on that point. if you overclock the AMD chip, you HAVE to overclock the intel chip. and when you do that, i am pretty confident intel will still win out. And before you wrangle my neck, I am not hating on AMD. I will quote a part of my own post that pretty much sums up what I think about this new CPU.
×