Repository navigation
Replies: 7 comments 15 replies
|
Did you also activate the ACPI fix in the bios? I tried that on one of my boards, and it would crash with errors in some games and completely crash the board in other games. |
|
Yeah I tried with 8 core unlock on and off and ACPI injection on and off but made no difference, if it were to affect my setup though would it not have been a problem with both watercool and simple air cool with original heatsink back on, once I put it back together to original nothing else was changed and issues went away. Thanks |
|
I have this issue with one of my boards, 8 CPU core unlock seems to be perfect, but running certain games will throw a greens screen of death, and no amount of messing with the CPU will fix it, undervolted under clock, nothing works, I shut those two cores off, and it OC's to 4175mhz and 40CU unclock at 2250mhz, sill well under 65C and no crashing. I have two other boards that have fussy CU's, I have shut a couple of them off, but the CPU cores are fine, so its a mixed bag. But one thing to note, your testing on the bench, so none of the VRM's are getting cooled? do you have some fans blowing on them? as outside of my case, there is zero airflow over them. |
|
Same for the VRAM, that's not being cooled either, both VRM and backplate are actively cooled inside my case. |
|
I did put the whole thing together inside the main 2 case shells and run it and it was exactly the same as on bench, I haven’t OCd CPU I do have all CUs unlocked and tested with 8 core unlock and ACPI off and on, I run the exact same setup with this shitty setup change nothing and can play cyber punk ratchet and clank, 007 all running well, is there a chance I could have shorted something with backplate and prob should have thermal padded the underside of APU, gonna complete a REDUX case with another board I have and then put this together again, is there a good compound I should use to glue the little heatsinks down |
|
I don't think there is anything you could have shorted with the back plate, if you have over tightened the AIO mount and its flexing, then maybe something could be touching, I've built a number of systems over and over, as I iterate designs I strip down and rebuild mine, over and over again, I have not had any issues, well the very first one I did, I used Thermal Grizzly Pro Putty, and I found out after slopping that crap everywhere, its electrically conductive, lucky the only thing it effected was the fan, it stayed running even after you shutdown the board, I spent many hours bathing the entire board in 100% IPA and scrubbed it with a tooth brush to clean it all off, board works great now, so they are pretty robust, but I have seen people knock off SME items, that have caused random issues. All I can say is pull it down and inspect everything, if you had the back plate off, check that no foreign objects got in there, same for the APU mount, also you don't need PTM on the AIO setup, It's not worth hassle, it makes bugger all difference, it does help with the stock cooler, but you have more than enough thermal headroom with the AIO to not even warrant PTM it anymore, the system overclocked runs cooler than Ryzen Workstation |
|
I never had the back plate off but it did warp somewhat when aio was fitted, thought maybe if I put the putty in behind it may help stop the flexing, I will also make sure I put on as little pressure as I can, I’m playing just cause 3 on it now in that simple one fan setup and it would have crashed 3 times by now before, works perfectly now it’s back with its original heatsink been playing on it for days, cloning M.2 drives with Clone Zilla also works, both boards I have run with that same setup pretty much the same, with the board I hadn’t opened yet running a bit cooler. |
|
Also with the VRM's on the plug side of the APU, make sure you not shorting anything, as space is very tight, and did you shave a few mm if the fins that are under the larger black ones where the temp sensor is located, as they will stop the bigger black ones from making proper contact. I use thermal plaster on mine now for a more robust connection, those sticky pads are somewhat crap at times, they do the job, but you can knock them off easily, the thermal plaster is a more permanent solution, you can also use some Automotive copper based silicon, like the RTV stuff from Permatex or Loctite, I have used this as well, and it works really well, and you can pull the heatsink back off and the silicone comes off easy, below is one of my first boards I did, you can see the orange goo, that's Copper based high temp RTV silicone, that works exceptionally well, I got a bit keen with that one, but the other side I was a little more strategic :) heatsinks were just random junk I found in one of my many "bits" boxes: 
|
|
This is the thermal plaster I now use: https://link.amazon/B04FgwUjS And if your interested in the silicone adhesive, I think JBweld do it as well: https://link.amazon/B0bFpDNRQ |
|
This is the area you need to check, as those little heatsinks are too high, and you need to shave off at least 1mm off the last fin thats under the black ones: 
|
|
These are the hottest and most important VRM's, and is why I have positioned an active fan above them, they will get stinking hot real fast when pushing the APU, the other ones, not so much. |
|
I thought I might need to do that because I had seen another thread where that was mentioned but the little silver ones were able to slip under the black ones without fowling I don’t think they were stopping the black ones from making contact but will have a look at doing that, should I use a compound to stick these heatsinks down |
|
Yes, I did say it somewhere, I have so much info scattered all over the dam place, kind of why I started this discussions section, I need to try and keep it in one place, I would just start with shaving down that fin and stick them back on and test again, just pop a 120mm fan over the top if your on the bench, just to keep some air flow over them, you don't need anything powerful, once you happy, then I would get some Thermal Plaster like what I linked to, at least for those VRM's, as they do get very hot, specially when you OC the GPU, I have been pulling over 500w from the wall at times with this setup, and no issues apart from melting the PCI-E plugs :) so you will want to get some Micro-Fit connectors with 18 AWG. The heatsinks are big enough with an active fan over them, but if you where to rely on just fresh air intake in my case like I do for all the others, then you would need larger heatsinks, which as you can see, can be quite hard, I would love for someone to create a nice machine aluminium slab that would cover all of them that we could just glue down, like a 10mm thick heat spreader, if I had time I would model somthing up in plastic, get all the heights right, then CNC it from some billet, but I just don't have time for it, and my mini CNC is still in the box, where its been for like 2yrs as I have not had time to even set it up :). I did just found a Voron Zero 120mm2 build plate I had packed away, I'm going to drill that out using my back brace as a jig, and replace the back brace with that, and see how much better the board performs with 8mm thick slab of flat Aluminium instead of the plastic brace, might be interesting, as these beds are less than $50 delivered now, and they are super flat, high grade Aluminium plate, and they are the perfect size, 120x120mm, the bolt holes are 96x96mm. |
|
Hi, would a ARCTIC Liquid Freezer III 240 All-in-One work with the Pro V2 case? Thank you! |
|
Sorry, not a chance, I have those coolers, they are way to big, the pump is massive and stands up on its side, the mounting system is completely wrong for my mounting arrangement as well. |
|
If you want to build the Pro-V2 and want to use a Flex PSU, then you have to use the recommended Thermalright AIO, if you're going to use an Meanwell LOP PSU, then you can use other types of coolers, but they will still have to bolt up like the Thermalright one, there is many, usually the cheaper ones are like this, plenty on Aliexpress like the Teucer AIO, but they will all only work if your using the LOP PSU, and I don't recommend people go down that route, not unless you have good electrical experience, as its an exposed power supply, and can be potential fatal if you touch the wrong thing. |
|
Okay thank you for your quick response! |
|
I have rebuilt this a number of times now and I am encountering the same issue each time. I have tried MX 4, MX 7 and TPM, I have tried using thermal plaster to stick the Igeek heat sinks and then I cut the original heatsink and used that. The longest I had it running was a couple of hours but always encounter crashes that I never see even using the stupid 1 fan blowing down the heatsink in the pic above. I’m going to convert this board to an air cooled redux case using a heatsink from another board then try again with that board on water cooled setup. It’s a headscratcher as all temps are low and not much of an over clock. The last time I put the heatsink back on and tested with the one fan all the issues cleared up. |
|
Same board with another boards heatsink fitted and two shit fans set on top. Not tight on APU so temps high but played for hours with no issue Fitted redux back brace and temps brought down near 30 degrees now better contact with APU, tested for hours no issue |
|
Peeled back heat sink and completed into another air cooled redux broke the little screen as I had printed the wrong faceplate and snapped it on the way out of it |
|
Ouch, yeh you need to pry it out from the inside, I print with a slight under flow ratio so parts fit together more easily, if your flow ratio is a little on the high side then the screens can be tight, but prying from the inside is safer than pushing it out. Really weird issue with the liquid cooler, I'm not seeing this, is the cooler working correctly? could the pump be playing up? |
|
Yeah I think I tried a little push from the front and didn’t realise how delicate the screen was, cheers I will take a look at my flow ratio, its default at the minute. |
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
This is to see if anyone else has hit the same issue and knows what causes it and to help others that have this problem and how to get back out of and what it isn’t.
I had a strange issue when I installed the liquid cooler on a fully working board I had tested with a simple air cooled setup.
Latest Bios installed with 8 CPU cores unlocked and settings set
Blue Pill taken
All CUs unlocked on GPU and running at 1850 default toml file
Original heat sink removed and water cooler fitted with PTM9750
Small Igeek heat sinks fitted using sticky back that comes with them
Corsair commander duo using specified firmware setup on laptop then the script run for Corsair C Duo, config altered and able to access its web gui
Tests run and games start crashing sometimes during loading sometimes in game after a short period
Sometimes whole system would hang, reboot or go to white screen
Temps were brilliant for CPU and GPU 36 or 37 degrees
Temps were ok on Corsair hub probes 55 probe 1 43 probe 2
Thought something had happened to my data on m.2 so took out another BC250 that had just arrived and tested that with my one fan setup, everything worked perfectly
Thought I may have a short from one of the small heatsinks so removed them in batches retesting as I went but nothing helped.
Fitted into case to rule out airflow problem but that didn’t help.
Removed liquid cooler, reinstalled old heatsink then retested and everything worked fine again
The only things I can think of are one or more components now with small heatsinks over heating or maybe a short between backplate and board
Small heatsinks I only used sticky pad that comes on them but noticed when heated it doesn’t take much to dislodge them, is there a non conductive compound I should use to stick them better and if I do should I remove the sticky pad from each of them first
All reactions