Yes, to all of the above - every strip down and rebuild.
Dark Rock 4 owners - i need re-assurance.
- Thread starter Plec
- Start date
Yes, stuck stock on this afternoon in spare 10 minutes but yet to find time to test (too tired now - think i've written the equivalent of a novel and my ear aches from phone calls - i'm frazzled :/). Hopefully, early tomorrow morning - as would like replacement hardware by Friday if possible, Tuesday latest.
Pins are fine - and as confident as i can be that it's not CPU related, given how the problem presents. But never, say never with a problem as obscure as this.
And every chance it's a MB problem in some guise - but again nothing obvious that sticks out on close inspection.
Pins are fine - and as confident as i can be that it's not CPU related, given how the problem presents. But never, say never with a problem as obscure as this.
And every chance it's a MB problem in some guise - but again nothing obvious that sticks out on close inspection.
Going to test now..
I'm kinda hoping it fails.
Same result with stock cooler.
@EsaT, @orbitalwalsh, @Danny75, @doyll, @tamzzy, @Journey, @Quartz and anyone else i've missed:
Consensus would be really helpful here:
Crux, the NVMe drive fails to install windows once secured into case (zero case attacments/cables) - but works fine breadboarded. NOTE: it will install Windows fine via a standard SSD within the case - and runs fine. So, problem leads me to think NVMe slot/stress(?) related - due to case fixing - leaning towards motherboard. (NVMe tested and passed in another computer.)
Tried
@EsaT, @orbitalwalsh, @Danny75, @doyll, @tamzzy, @Journey, @Quartz and anyone else i've missed:
Consensus would be really helpful here:
Crux, the NVMe drive fails to install windows once secured into case (zero case attacments/cables) - but works fine breadboarded. NOTE: it will install Windows fine via a standard SSD within the case - and runs fine. So, problem leads me to think NVMe slot/stress(?) related - due to case fixing - leaning towards motherboard. (NVMe tested and passed in another computer.)
Tried
- Tested NVMe in a different build (Tomahawk based) - worked fine
- Tried standard SSD within build with DR4 and worked fine within case
- Tested different coolers
- Different USB/Downloads for installs
- Different memory
- Different GPU
- Checked CPU pins
- No warping
- Temps were fantastic with DR 4
- ...
Last edited:
I know - the one thing i haven't tested - different case :/
Son's case was Q110 (and complete PITA to work on) - he wanted an optical drive bay.
/really, really, don't want to have to strip down another system
Tested in horizontal and vertical - each rebuild.
Yes, but then the windows install, trying different memory - and chasing my tail with work.takes 5 mins
9 off mobo screws out then back in
I have another system i could pull apart (my son and daughter are home - i think they would freak if i ripped the guts out of their new systems.). But, i'm very much at the FFS stage, due to time constraints - i'm sure you get it...
Worth a shot - i'll give this a quick try as may save me attacking another system for parts...
When this is eventually, up and running, i'm naming it Jordan:
It's the build equivalent of Katy Price, high on Ketamine while doing the school run in her Pink Land rover. It's all over the place and every point of access leads to an eventual crippled Windows/windscreen...
You do work?..
And i'm working from home now (did a lot anyway), only go in if they need face to face - but my work load has doubled due to recent events. This, computer issue has come at a very bad time - i may have to throw money at it.
It's the build equivalent of Katy Price, high on Ketamine while doing the school run in her Pink Land rover. It's all over the place and every point of access leads to an eventual crippled Windows/windscreen...
That sentence has confused my already addled brain
You do work?..
And i'm working from home now (did a lot anyway), only go in if they need face to face - but my work load has doubled due to recent events. This, computer issue has come at a very bad time - i may have to throw money at it.
**** - I'm going to have to try a different case. 4 screws (top 3 and second one down nearest edge of backplate loosely fitted to limit sliding) and, still operational with driver installs and 6 restarts. I realise, very limited testing but it would have normally corrupted by now.
It's not really proved anything more that i'm already assuming but it's another piece of evidence that's forced my hand into testing in another case after i've finished a few calls :/ (Damn you @Danny75
- i knew i had to do it, but i needed a shove, your test gave me no way out, kudos, and #$*&ing bloodies!)Yes, i've been pretty thorough:
OK, no surprise, Windows becomes corrupted in a different case too - so very damning for the motherboard.
Put in an RMA (waiting for reply- but will have to buy replacement now and get them to reimburse when they've done their testing as I need a working system now.
YAY - of sorts - knackered. Stripped down the old system in approx 3 minutes flat, It's going to take a lot longer to assemble - cables and drives everywhere....
Put in an RMA (waiting for reply- but will have to buy replacement now and get them to reimburse when they've done their testing as I need a working system now.
YAY - of sorts - knackered. Stripped down the old system in approx 3 minutes flat, It's going to take a lot longer to assemble - cables and drives everywhere....
Cheers @Danny75
Ordered a replacement from OcUK - should arrive tomorrow.
Awaiting reply from Customer Services forum but they're usually pretty good (suspect they're busy with lots of panic buying/prices) and i've been as detailed as i can, outlining fault. Plus, the RMA (refund of the new board - as it's gone up) should be a formality, in theory. Especially, if i have an up and running system tomorrow...
But, appreciate that they're probably busy so refund delay is small price to pay for up and running workstation (i hope!).
Last edited:
OcUK are dealing with the RMA as I type (very quickly and courteously i might add) - so much easier using their CS forums.
Forgot about the heatsink (would have bought the other bequiet) - but the Dark Rock 4 should be fine on a fully functioning board. I've fitted bigger coolers on flimsier boards and never had a problem like this. Admittedly, no m.2 slot...
Ordered the same as i need to get a workstation up and operational as soon as possible now. I was familiar before with the board - now i know it too well, but will help with speedy setup.
Well, the guy in the returns forum has been excellent - testing needs to be completed, as i reported a fault, but I will receive a full refund regardless, when testing is done. The guy was, succinct and answered all my questions quickly and accurately.
I've not had to return anything to them, personally, in ages, as i make all the people i build for buy and return via their own accounts. But, credit where it's due @Jargan (hope he doesn't mind being doxxed), made the whole process very simple - kudos. A relief with what i'm having to juggle this week (kids home too) - i feel like i'm spinning plates while trying to model the solution to the Travelling salesman problem...
@Yewen, @fiveub (preferred 5UP
) and @StigBourne were excellent in the past too - presuming they still are but not crossed paths in a while (although i think yewen moved on - i can't keep track, i've been on these forums too long.)
Anyway, a few more hours, then beer o'clock - it's virtually Friday, @tamzzy
*Will keep you posted when receive new board tomorrow (hopefully).
I've not had to return anything to them, personally, in ages, as i make all the people i build for buy and return via their own accounts. But, credit where it's due @Jargan (hope he doesn't mind being doxxed), made the whole process very simple - kudos. A relief with what i'm having to juggle this week (kids home too) - i feel like i'm spinning plates while trying to model the solution to the Travelling salesman problem...
@Yewen, @fiveub (preferred 5UP
) and @StigBourne were excellent in the past too - presuming they still are but not crossed paths in a while (although i think yewen moved on - i can't keep track, i've been on these forums too long.)Anyway, a few more hours, then beer o'clock - it's virtually Friday, @tamzzy
*Will keep you posted when receive new board tomorrow (hopefully).
Thanks, @doyll, appreciated your input.
Unfortunately, i didn't be get my new motherboard today as it looks like it got held up in processing. I think they're experiencing much more demand than normal, and probably reduced staffing levels (guessing) - so i won't have any updates until Monday
Well, DPD have informed me i'm receiving my replacement motherboard this morning - so either OcUK did me a solid or DPD are expediting all deliveries regardless of weekday/weekend (original slot was due Monday).
I've received a replacement DR 4 too (different company as had credit to use), as i wanted that ruled out in diagnostic process. In hindsight i regret not buying a Noctua/Mugens but no stock - as I have nagging doubt about the DR 4's fitting, and the pressure it applies, via its method on this particular board when using the m.2 slot. I realise the stock cooler did the same - but that was after the placement of the DR4. This time i may install stock cooler first test and then the DR4 (if i have time) as that will be as conclusive as my testing could get if the problem was to present itself again.
Or, i install the DR 4 with the motherboard already installed into the case which should limit any flexing... (But would be a PITA to fit.).
*Realise - that the DR4 shouldn't be an issue - and have read nothing like this anywhere else - but as mentioned, no other examples of an NVMe and DR4 combined with this board so no data to cross reference (none that i've found with a brief search.). If it did fail, it would point to a board vulnerability - but for now i'm hoping it pops into life and allows me to get rid of some boxes, and concentrate on other things when not working...
Thoughts?
I've received a replacement DR 4 too (different company as had credit to use), as i wanted that ruled out in diagnostic process. In hindsight i regret not buying a Noctua/Mugens but no stock - as I have nagging doubt about the DR 4's fitting, and the pressure it applies, via its method on this particular board when using the m.2 slot. I realise the stock cooler did the same - but that was after the placement of the DR4. This time i may install stock cooler first test and then the DR4 (if i have time) as that will be as conclusive as my testing could get if the problem was to present itself again.
Or, i install the DR 4 with the motherboard already installed into the case which should limit any flexing... (But would be a PITA to fit.).
*Realise - that the DR4 shouldn't be an issue - and have read nothing like this anywhere else - but as mentioned, no other examples of an NVMe and DR4 combined with this board so no data to cross reference (none that i've found with a brief search.). If it did fail, it would point to a board vulnerability - but for now i'm hoping it pops into life and allows me to get rid of some boxes, and concentrate on other things when not working...
Thoughts?
Thinking of buying the Noctua NH-U12S - i can only get the Cromax for next day delivery so £10 more for a cooler you're never going to see :/
Looked again at the DR 4 (especially the replacement) - my gut keeps telling me that it may be contributing factor, and the Noctua fitting i know is not as pressured/brutal - and as @EsaT said:
This has been bugging me too. As mentioned, i've fitted a couple of these on other systems no issues - but these had no m.2 slots (close by). But, with the m.2 slot so close - and as i have fitted the PITA that is the DR3 in the past - I was just happy that the install process was easier. But, now i've had to fit one right next to an m.2 slot which has given out ambiguous results - and time to reflect - i'm doubting the motherboards capability to support this type of clamp fitting compared to the lesser sprung screw fitting (not worried about weight).
When i took out the new replacement to inspect - the retention bar obviously hasn't been bent/moulded yet by fitting - and the gap between bar and bracket is huge on both sides (bigger than in my initial picture with a pre-fitted so pre-bent bar - reminder: https://imgur.com/75LJZvl). The 'initial' force really does seem excessive to mould the bar to top of HS, when closely inspected. Don't get me wrong - it's a great, secure fit, quieter running than the Noctua - cheaper too - but I'm not prepared to go against my gut over £20 with return costs. Too little time and i would kick myself self if i had the same problem develop when i've considered that this could be a possible reason (I still blame the board design if this is the case - but i need to move on and i know a spring loaded cooler works fine.)
Looked again at the DR 4 (especially the replacement) - my gut keeps telling me that it may be contributing factor, and the Noctua fitting i know is not as pressured/brutal - and as @EsaT said:
This has been bugging me too. As mentioned, i've fitted a couple of these on other systems no issues - but these had no m.2 slots (close by). But, with the m.2 slot so close - and as i have fitted the PITA that is the DR3 in the past - I was just happy that the install process was easier. But, now i've had to fit one right next to an m.2 slot which has given out ambiguous results - and time to reflect - i'm doubting the motherboards capability to support this type of clamp fitting compared to the lesser sprung screw fitting (not worried about weight).
When i took out the new replacement to inspect - the retention bar obviously hasn't been bent/moulded yet by fitting - and the gap between bar and bracket is huge on both sides (bigger than in my initial picture with a pre-fitted so pre-bent bar - reminder: https://imgur.com/75LJZvl). The 'initial' force really does seem excessive to mould the bar to top of HS, when closely inspected. Don't get me wrong - it's a great, secure fit, quieter running than the Noctua - cheaper too - but I'm not prepared to go against my gut over £20 with return costs. Too little time and i would kick myself self if i had the same problem develop when i've considered that this could be a possible reason (I still blame the board design if this is the case - but i need to move on and i know a spring loaded cooler works fine.)
Last edited:
It does - but from my experience (I'm not as informed as you with coolers) - i've found that the bar method has to apply, initially, more force to level out the bar - this has never concerned me in the past.
The Noctua fitting does not exact the same amount of initial force for fitting (neither does the Mugen 5). And although i accept the DR 4 is within tolerance - from personal fitting it's got a greater pressure tolerance than some of its peers. That coupled with my very strange issue (very weird everything considered), which i agree may be unrelated, but for the sake of £20 i can remove another possible factor (i accept it may not be.)
Put, it this way - had i fitted the DR 4 and got the same results i would feel sick to my stomach - mainly because i would have kicked myself. Purely, as i anticipated that it might be a contributing factor and spare time is a novelty at the moment...
You're killing me @Danny75
The NH-U12s are babies in comparison - and i'll only be fitting finger tight (i normally do).
Hopefully, my bar/spring comparison makes sense - it's that initial first fit, when it bends the bar. My unproven theory is that it 'may' have been a factor in compromising the slot - but again i blame the motherboard, still have a lot of love for the DR 4 - prefer it's cooling/quiet ability.
Last edited:
Balls, i'm debating fitting the new DR 4, but with the pre-fitted bar from the old one. This would take out that initial extra exertion of bending/flattening the bar.
If it failed again - it has to be a design fault with the motherboard.
Thoughts, @doyll, @Danny75, @EsaT, @orbitalwalsh? Or, does the Noctua argument make sense?
*I'm really short of time (writing this in-between doing 2 other things) - but i'm wasting even more time due to a nagging suspicion that i'm experiencing a fault due to a rare combo of DR 4, specific motherboard and use of NVMe...
So much simpler had it been a suspicion of a dodgy cable, PSU, memory...
If it failed again - it has to be a design fault with the motherboard.
Thoughts, @doyll, @Danny75, @EsaT, @orbitalwalsh? Or, does the Noctua argument make sense?
*I'm really short of time (writing this in-between doing 2 other things) - but i'm wasting even more time due to a nagging suspicion that i'm experiencing a fault due to a rare combo of DR 4, specific motherboard and use of NVMe...
So much simpler had it been a suspicion of a dodgy cable, PSU, memory...
Last edited: