Too HOT!

I’ve replaced psu & aio.

Like other posters have stated, you had no flow in your water loop. You should be able to tell this by touching the input and output hoses to your cpu waterblock. Both sides should be about equally hot because of no flow and the heat is just conducting through the water. If you had flow, the inlet hose would be cooler than the outlet hose. You might be able to sense flow by squeezing a hose to nearly shut, then slowly releasing to detect the change in flow. Do at your own risk though, I’m not sure how the AIO hoses are designed.

You said you replaced the AIO, what do you mean by that? Isn’t it pretty unlikely to get two bad AIO’s? I think you mean you flipped the radiator vertically. If you’re pump is running, then even though you did this, there is still air trapped in the pump. You want the air to go into the radiator, so you need to move the pump below the radiator, or lower in any case, so the air bubble will flow upwards and out of the pump. It would be useful to do this with a spare power supply providing power to the pump at the same time, to get the air bubbles all moved out of the lines.

My first guess would have been to check the power input to the AIO pump. It’s possible your bios settings are set to AIO disabled, it’s possible the AIO pump cable is not properly connected.

I’ve experienced something similar to this before, and it was due to my failure to reconnect the pump power cable. The core’s thermally throttled and my machine ground to a halt. Yes, I was still able to youtube but before I opened hwinfo my next clue was task manager’s cpu speed staying at the lowest idle frequency. My experience was on an i5-2500k which I ran at 4.7 Ghz for years and just recently retired to second string.

Your conjecture about bad PSU and bad MOBO are pretty unlikely. It’s possible the AIO pins are bad. In which case, I’d recommend to try connecting the AIO pump to a chassis fan header.

If you think you can manage using a multimeter, an inexpensive one and some wire cutters and spare wire could be used for some voltage verifications.

2 Likes

Did you try under clocking and lowering/raising the voltage, to see if the chip is stable/cool at lower speeds?

Looking at the long list of screenshots, it looked like the cpu was reading over 100c in BIOS, at 0.9v.
That seems real hot for that voltage, so I could well be wrong, and I Do Not Know what is an appropriate voltage for speed of your chip.

But a lower clock speed might help test?

0.9v seems low to me.

1 Like

My AIO is a CoolerMaster MasterLiquid LC240E RGB. For me, it was the first time I’d ever had liquid cooling, AIO or otherwise. I assembled this computer 2019/01 and within a few days, the AIO leaked green coolant on multiple parts shorting out and destroying the CPU, Motherboard, & GPU. All were replaced and seemingly worked successfully w/in a week.

2021/08/02 I replaced the psu, upping from 550 to 650, both corsair.

2021/08/16, @bigsqueaks : when the shutdowns got to the point that the machine wouldn’t stay on more than 1 second, I replaced the AIO with an identical new AIO. This is my 3rd identical new AIO.

2021/08/18 at @Kocytean 's recommendation, flipped the radiator so that the hoses were on the bottom. @Ghan : zero cooler mount wiggle. looked for but unable to find any BIOS reference to “Multi-Core Enhancement” to disable.

2021/08/19 at @wertigon recommendation, I tried running the machine with the side panel removed. Same day at @d0rk 's suggestion, I started putting my fingers near (within a cm) from various spots identified as likely VRM and felt negligible heat.

People didn’t like my use of OpenHardwareMonitor so I switched to HWInfo64 and provided images.

20219/08/20 I posted imgs from excel of the log. W/o anyone telling me definitively that the issue was the MB or the CPU, I ordered a replacement MB, the less expensive part, which won’t arrive until 2019/08/26.

2021/08/21 @Exard3k : I re-thermal pasted at nearly everyone’s suggestion.

@redocbew : the AIO has a radiator w/2 fans, tubes, and the block that attaches to the CPU. The fans are spinning. Currently, nothing, including the cpu block is hot.

@Gnuuser : the tubes between the radiator and cpu block are not disconnect-able so I don’t think I would be able to] cut into one and install another component and frankly, I don’t think I’d be competent if it was possible.

@MisteryAngel : the auxtin1 is always high, but it’s not constant. please refer to the graph of the log. as to using a different cooler, this one is brand new as of 2021/08/16. My decision to try an AIO was because I was afraid of implementing a liquid cooler but believed that this device would be more effective than just a fan. Also, I agree that the CPU temps are more worrying.

as @HaaStyleCat points out, when it gets hot, they all get hot. Yes, ALL parts were new.

bigsqueaks : Once when the new AIO was installed after attempting to play POE a shutdown occurred shortly into the game. When I touched the tubes, both they and the CPU block were very warm. The CPU block has been re-thermal pasted since that time. I’m assuming that the chassis fan header doesn’t respond to the CPU’s temp and so I’m wary about connecting the AIO radiator fans to them.

Trooper_ish : I have never played around w/any overclocking, cpu, memory, volting, or otherwise for fear of getting things too hot. No I haven’t tried lower clock speeds.

Regarding the possibility of my miswiring something: Two wires come from the CPU cooler block [B]. The one with 3 ports has tabs to “polarize” orientation and is connected to 3 pin header on the right in [A]. The [B] wire w/4 ports connects to a complicated wiring contraption [E] that receives independent power for the 3 button mini controller for the lights. [C] connects directly to the 2 radiator fans and has [D] a 4 port header-port with tabs that polarize orientation. It is connected to the header on the left in [A].

per the z390 manual[https://download.asrock.com/Manual/Z390%20Pro4.pdf], the 4 pin header is cpu_fan1. inexplicably, the 3 pin header adjacent to the 4 pin cpu_fan1 doesn’t appear in on the z390 mb layout.

wa





wf

Found the diagram of the prescribed wiring:
PrescribedSetup

New to thread and haven’t read the entire thing, but has the orientation of AIO fans been verified? Are they pushing air out the front of the case?

No, they’re air pushing into the case, pulling thru the radiator. there is a slight grill on one side which, per the diagram, seems to facing the radiator.


I can’t speak for certain as I can only feel the blow near the edges where it is definitely inward toward the mb.

The fan direction is the air moves towards the hub. So in the picture posted above (third post in thread) it seems like the brackets supporting the hub are against the radiator, which would mean the air is moving towards the radiator and out the other side. In other words it’s push air through the rad out the front of the case… at least that’s what I’m seeing. If that’s the orientation then ALL the fans are exhaust fans.

Can you post a new pic of the radiator fans with the system off.

Yup i saw that.
But i don´t think you really have to worry about auxtin1.
That has nothing to do with the vrm.
AUXTIN is just auxiliary temperature index.

But the cpu temps itself there is definitely something going wrong.
If you have an aircooler laying around for a quick test.
Then i would recommend doing that.

1 Like

Highly suggest this as well. Really seems like a mounting or pump issue. I usually have to use bios to set pump speed in the thousands range depending on manufacturer and pump design (anywhere from 3-4k). That’s all I can think of without my hands on it.

With all new parts I can’t see it being anything but a cooler issue especially with doing replacements of parts again… and with a new cooler of the same design has to be a install or operation error I’m sorry to say. Not to be rude or mean by any means.

Also i could highly recommend to try other software for temp,
read outs as well just to compare numbers.

1 Like

Also, looking at your pdf… image 1.3 Motherboard layout, I’d use connector #2 it is actually listed as a waterpump connector. It should have enough power to run the pump if amps is a issue as well. Then can connect the dual fan connector to the cpu fan connector if you like.

@anon35662433

If you click it big, you can see the grill of the fan behind the blades. I think that is the way the pamphlet that came with it showed to install it.

1 Like

@ MisteryAngel, which other software for temps would you recommend. I’ve currently got HWInfo64 and OpenHardwareMonitor.

1 Like

There are a lot of messages going around but Please start with the basics!
Is the pump actually running?

Following the manual, pump should be connected to a 4pin CHA_FAN1/WP, CHA_FAN2/WP and CHA_FAN3/WP.

The 3 pin header you are talking about next to the CPU fan, is not a fan header! it’s something like a test connection. so it won’t work if you plugged the pump in there!

Try to plug it into the port between the CPU socket and the M2 slot, that should work. and it should be quite noisy, if that is the case, run it for 15 minutes and look at if the temperatures are going down.

2 Likes

Aida64 or core temp are also two you could try.

1 Like

So I’ve got a question. What were the voltages like? Overlay them? In the spikes are spikes or droops occuring? Just curious at this point

1 Like

simple test.
when running, put the tip of a screw driver on the pump block, put the other end to your ear hole and listen…
whirring?.
silent?.
bubbling?.
what can you hear?.
psst! only the first one is good :wink:

1 Like

Yep that’s exhaust… the pamphlet is showing it as it exhaust because most people mount them at the top of their case where you want it as an exhaust. In your scenario it’s mounted in the front where you would want it as an intake… flip the fans and I’m pretty sure you’ll solve your problem.

BTW, there should be arrows on the side of the fan… the arrow is pointing in which direction the air moves.

1 Like

@nutral
I don’t yet have stress test confirmation, but you are officially my hero! I rewired per your suggestion and then performed @anon7678104’s screwdriver to ear test…whirring. I think this has fixed my issue.

I currently am running multiple things in the background including POE, POB, as well as youtube videos playing 2x and am seeing VERY low temps on cpu. I’m baffled how this functioned since 2019/01 w/o melting. I just hope that 2 year heat stress hasn’t perm-damanged my cpu. Thank you for your insight.

3 Likes

You were just running in natural circulation mode. Just think of the energy savings of having no pump and your cpu running at idle power!

2 Likes

Yes, there is that, but I think i’ll stick with the lower core temps after 6 hours of celebratory game play w/concurrent video watching.

1 Like