Jump to content

jughead0

Member
  • Posts

    5
  • Joined

  • Last visited

Awards

This user doesn't have any awards

jughead0's Achievements

  1. See, I did all that already. The system is unstable at default BIOS settings. BIOS is the latest.
  2. Yup, only use those. Like I said, BIOS is up to date. I guess I'll have to ask a friend to make me a memtest thumbdrive. What's weird is, as far as I understand, Prime95 doesn't stress RAM, but my system stopped working after that test even now after I cleared CMOS.
  3. Hello all. TL;DR I've played around with some RAM overclocking and power management and not my system reboots after 2-5 minutes of ANY use or even just idling. So I've recently built a new AMD system and I'm having some problems. I will try to write it up in as much detail as I can remember so there's full picture. After I've assembled the PC and installed Windows all was fine. I've disabled some annoying lights on the MB and started updating all the firmware and software. So AMD chipset driver, BIOS and GPU driver are all the latest. Then I've installed all the monitoring software necessary because my idle CPU temps seemed too high (49-62C randomly). I have since read that this is basically fine for 3600, this isn't the issue now. After I tweaked some stuff in BIOS like disabled PBO, set up a custom CPU fan curve and then adjusted the power profile in Windows (AMD High Performance, min power to 1%, max to 100%) the temps became better and the fan stopped ramping up and down. Then I decided to overclock RAM and IF. Found some OC manual OC guide (since Nanya chips in my RAM isn't supported by the DRAM Calculator), set the RAM to 3200CL16 and the IF to 1600. The system booted and seemed stable in light stuff like web browsing. Also at that point the PC was on for like 4 hours while I was watching some videos and reading stuff about overclocking and power management, everything was just fine. Didn't get to play any games or load the system in some other way though. Then I decided to try and bump the RAM to 3600. Set the voltage to 1.4V, the IF to 1800, didn't touch the CPU voltage at that point. The system got stuck in a bootloop. So fair enough, I discharched the power button, cleared CMOS, booted into BIOS. Then dialed in the last successful RAM configuration (3200CL16, voltage on Auto), set up the fan curve again, didn't touch anything else, booted into Windows no problem. I still was getting this temp pattern in HWinfo, so after some further reading I've decided to do a voltage offset of 0.1V on the CPU and try and run Prime95 to determine whether my load temps were okay. System seemed stable with the offset, so I ran the test, the temps were sitting at around 67C for the first ~5 minutes, the suddenly shot up to 85C and stayed there for the next ~5 minutes more until I shut it down. That's where it all broke down. After I left the PC idling to cool down, the temps dropped to around 65C and refused to go lower, the CPU fan was ramping up and down (I think 65C was around where my fan curve was supposed to rise). That continued for ~3 minutes, then the system rebooted. Upon loading the temps were about where and how the were before Prime95, but then after ~3 minutes the system rebooted again while I was watching some Youtube. I turned it off, let cool for around an hour just to be sure, cleared CMOS, left BIOS at default settings, loaded into Windows and right away changed the power profile to how it was at default. The system still keep inexplicably rebooting after 2-5 minutes of any light use. I checked that all the connectors are seated properly, the cooler is screwed down tightly, I reseeded the RAM into each others slots (dunno why but whatever). Still the same behaviour. I'm out of options and losing hope. This is my system below. Everything is brand new except for the GPU (around a year old) and the PSU (around 3 months old), both were working in my i5 3470 system with no issues. The case is cooled sufficiently, my MB and drive temps are around 40C, the GPU is in low 30C at idle. CPU - AMD Ryzen 5 3600 (OEM) Motherboard - Asus ROG Strix B450-I Gaming ITX RAM - HyperX Fury 2x8Gb 2666MHz HX426C16FB3K2/16 SSD - Samsung 970EVO 250Gb MZ-V7E250BW GPU - Palit GTX 1050Ti PSU - Seasonic FOCUS Plus Platinum 650W Cooler - Noctua NH-U9S Case - Cougar QBX My only theory at this point is that there might be something wrong with the BIOS seeing that the latest version is about 3 weeks old. But didn't manage to find any complains about it so far from other people. pls help
  4. So are most sleeper PCs, that's the point I think. Retro aesthetics and modern insides.
  5. So I just rewatched the series on sleeper PCs and it gave me an idea: why not build a modern gaming PC not inside an old-looking case, but inside an old-looking display? As far as I know, CRT displays went up to 21 inches in size and most of them had a pretty bulky behind. That would allow to put, say, a mATX motherboard (since it's kinda square) parallel to the screen, just like modern towers with glass side panels. That would leave enough clearance between the glass and MB for the GPU(s) and whatnot. Also, as you can see in the example above, they usually had quite a lot of perforation at the back for cooling which can be utilised for the same purpose. Now, because there usually wasn't much perforation on the sides close to the screen (where almost all of the components would be), this build would require a custom water-cooling solution. Keeping that in mind, I made this EXTREMELY rough schematic of what the insides would look like: The dual radiators would be positioned, as seen above, on both sides of the case, probably with dual 120mm fans each. Maybe one in an intake config which in theory would benefit the airflow, although my "expertise" on that front is really questionable. The tubing would have to be routed at the top of the case over the motherboard. Also, as you can probably tell from the image, the IO would be facing downwards and also (probably) hidden inside the body. That means that all the cables (which would probably be a single video output cable for the, you know, ACTUAL monitor) would need to run through the back and bundled with the power cable. If needed, front IO could be modded into the front-bottom of the case. Obviously since the "case" would literally just be and empty plastic shell you would need to make quite a number of metal frames to mount all the components on. Or maybe just cut them out from a regular case, but custom-making would probably be the way to go. So far this whole thing seems kinda silly and weird. but not exactly outstanding. However, the icing on the cake would be the front glass, which would be electrochromic and change its transparency when PC turned on. This is what the thing itself looks like, if you aren't familiar: This is how it's usually utilised. Also in house windows, showers, etc. And, from what I gather, there are companies out there that custom-make glass windows like this. Also, as you can see, when such glass is in its non-transparent position, it very much resembles the kind of greyish tint that old CRTs had when they were turned off. So the finished case would have the glass wired to the power button so it would turn transparent simultaneously with the PC booting up and reveal all the components inside. This is what a poorly photoshopped version of what I'm talking about looks like: That's pretty much all I have to say about this. The reason I wrote this all up is that I don't really have neither skills nor the resources to execute something like this so I'm just hoping someone gets inspired by this idea and runs with this. And then posts the result here, hopefully. Thanks for your time!
×