• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

5950x and PBO overheated something or busted custom loop?

Toldja it was a memory problem.. a few pages back..

Sammy B-Die is the way to fly for an absolutely problem free experience, imo.
 
Toldja it was a memory problem.. a few pages back..
Not so sure about that. I finally got around to testing the Corsair DDR4-3600 in a different system and it was fine just as I suspected.

I think I've ruled out the problem being the CPU and RAM as I've been running the old toasted 5950x and Nemix ram in my ML09B with the new B550 Phantom Gaming-ITX/ax for awhile now absolutely no problems since it was built. I took a month or so to enjoy the 4060 low profile and it was just excellent.

The NR200P with the old B550 Phantom Gaming-ITX/ax and newer 5800x and Crucial RAM seemed to be working fine for awhile but recently started to have problems again now that I've had a chance to put some time back on it trying to compare experiences between the 4060LP and RX6700 XT with Horizon Forbidden West. :(

Graphically messed up BSOD's returned also some SECURITY TPM error and the return of Ryzen Master not displaying telemetry data again. Looking in event log NO WHEA errors of any kind but a lot of app crashes related to an AMD process that must be supplying data to Ryzen Master, also Samsung Magician, and some Windows services crashing too. The system now refuses to apply the latest windows updates for some reason too.

At this point I'm going to try securing erasing the NVMe and do a complete fresh install.
 
So an update on the old B550 Phantom Gaming-ITX/ax (2 or 3 yrs old now) in my NR200P. The fresh OS install helped smooth out a lot of problems as I think it had crashed so many times the installation got messed up. The only issue remining was applications kept crashing randomly mostly when the computer was near idle workloads. I managed to finally get it stable again by adjusting the LLC for vcore and SOC to mid levels and changing some voltages. Looking back I almost can't believe I had been dealing with this since around Nov 2023.
SOC=1.2v, CLDO VDDP=auto at 1.0979v, CLDO VDDG CCD=auto at 0.9976v, CLDO VDDG IOD=auto at 1.0979v

Interestingly enough the new B550 Phantom Gaming-ITX/ax in my ML09B which has never had a problem also had similar voltages just on full auto and mid settings for LLC.
SOC=auto at 1.0v, CLDO VDDP=auto at 1.0979v, CLDO VDDG CCD=auto at 0.9976v, CLDO VDDG IOD=auto at 1.0979v

At this point all my hardware is working including the old ram and cpu.

The B550 Phantom Gaming-ITX/ax just has some kind of quirky VRM I guess.

Case closed.

It's been a long thread and nearly reached 40K views. Hope you have enjoyed the journey.

As for the other project taking a second look at thermals using the replacement Master SLI/ac I got on ebay - at this point I just don't have the time and I'm going to abandon that project.
 
Last edited:
Continuing with day 2... New tubing runs

I realised I didn't finish posting so I will start with a meme and continue

View attachment 303141

Let's break out the new tubing. I was pretty sure I didn't have enough scraps from the last change over from EK Duraclear to Primochill so I got another box and good thing too as it took a couple of attempts to cut the right length for the more direct runs for the new rad. I was getting pretty tired by this point in the day and 1/4 of an inch off ment not having a satisfactory run or an uncomfortable bend.

Below you can see my new drain layout with a rotating 90 on one side so I can tuck it away nice an neatly. For comparison I put the old setup side by side below. I also got rid of the ball valve on the other drain and replace it with the newer EK drain valve so I could reuse the ball valve elsewhere.

View attachment 303142View attachment 303143View attachment 303144

Let's take a look at the new rad


View attachment 303145

So I opted to route the tubes from the pump output directly to the 2nd rad and return back to the distroplate. The tubing run was fairly simple although I wasn't fond of putting the 90 back in there but I didn't have another 45 on hand so there it is.

View attachment 303146View attachment 303147View attachment 303148

Also since I had ordered some additional offsets I wanted to align the tube runs from the top rad a little better (and farther away from the motherboard I/O panel. At the time of ording I couldn't decided on black or nickel so I ordered both. Good thing as I was able to use them all in a useful way.

It took me some trial and error as I was test fitting rad and the offsets.

View attachment 303151View attachment 303152

The inner most offset was good and moved the tube away from the I/O panel and squared it up to the distroplate nicely.

View attachment 303153View attachment 303154

The outer most tube however was coming in causing the tubes to line up into a V pattern instead of parallel so I reoriented the offset so they are both pointing outward.

View attachment 303158View attachment 303159

Next that pesky RGB Cable. I managed not to squish it to death when putting the distroplate back in. It helped by moving it outward to the right then folding it toward the rear/side compartment. There was just enough space between the top rad and the distroplate to use a credit card to push the flat RGB cable between them up past the rad cover screws and tuck the cable at the top very nicely.

View attachment 303162View attachment 303163

Now lets run these rad tubes. I decided to keep the hot side temp sensor where it was before the loop enters the first rad on the top.

View attachment 303164View attachment 303165

Now let's use this beautiful new drain plug arrangement and pressure test the system starting with the first side of the distroblock.
Swinging the bad boy out there and hooking up the pressure tester.

View attachment 303166View attachment 303167

!!! Oh WTF I've got no pressure !!!! Pressure was quickly draining !!! All I could think of was at this point the new rad's fittings must have come loose when I tried to put the 90's back on. (because they were a real pita trying to tighten and I wiggled them a lot)

View attachment 303168

Ugh. Taking out both rad to retest.

View attachment 303170View attachment 303171

Both rads were OK. The problem is elsewhere but where could it be?

View attachment 303172View attachment 303173

Stupid me I forgot to put back in a plug! Major Homer Simpson moment that cost me a lot of time but at least it wasn't a catastrophic error.

View attachment 303174

Hooking it back up and testing both sides to be sure. Now working as expected.



View attachment 303176View attachment 303177

Top rad back in and pressure test it again.

View attachment 303178

Ok prepping the motherboard

View attachment 303179

(reached image posting quota)
... to be continued ...

Reinstalling Motherboard and GPU in and I skipped taking a lot of pictures at this point but I finished attaching all the tubing but not testing the whole system yet.

View attachment 303180

View attachment 303181

And this finishes day 2 in the late evening.

Next up Day 3....

continuing with Day 3....

Pressure testing. It looks good. Taking out the new cryofuel batch I mixed a week or so before and ready for some fun.

View attachment 303183View attachment 303184

Keen to avoid the mishaps I had with the funnel and too lazy to go get the painfully slow EK fill bottle I decided to hookup my EK res and use that for the fill.
Turns out this was a really great idea and filled my system very quickly. You can see the new EK fluid has a slightly yellow tint to it, not sure why.

View attachment 303186View attachment 303187

Opened the valve, gulg, gulg, gulg, in goes the fluid.

View attachment 303188View attachment 303189



View attachment 303190View attachment 303191

Using my handy PSU switch to turn on the pump. (was careful to not connect the pump's PWM during this time)

View attachment 303192View attachment 303193

It's filled and I'm letting the pump run for a bit checking for sound and looking at the flow indicator. What I see happening is the flow indicator is running slower than normal. To which is not entirely unexpected considering I did add a new rad to the system and some 90's. I remembered I forgot to flip the orientation of the QDC's so time to drain the system and flip them around.

View attachment 303194

Opening the valve to help drain the system enough the change the QDC's. Rebottling the new fluid for now.

View attachment 303195View attachment 303196

it was fast and easy.

View attachment 303197View attachment 303198View attachment 303199

Since I had the QDC's ready I took a bit of a detour and decided to try out this new filter I ordered. I thought this filter would be handy for the next time I have to setup some new rads.

View attachment 303200View attachment 303201

It's all working and reversing the QDC's did improve the flow indicator speed. Packing some things up.

View attachment 303202View attachment 303203View attachment 303204

Still have to put back in the bottom fans and the Corsair hardware. That new radiator fit just perfectly.

View attachment 303205View attachment 303206

First the fans. The dust filter that comes with the 011D is terrible because once you put the screws in for the bottom fans the magnetic sides of the filter are pulled a way a bit and doesn't stick to the case as well. I got some new dust filter and magnets to remedy this problem by making a custom dust filter.

View attachment 303207View attachment 303208View attachment 303209

eSata plugged back in and cable management on the bottom is looking good.

View attachment 303210

Next up Corsair cable management and having fun with a lot of zip ties.

View attachment 303211View attachment 303212

... to be continued ...
You're so cool, what are you doing with that red head
 
Back
Top