Overclock.net banner

PCH temperature out of control on Z690 Carbon

42K views 23 replies 15 participants last post by  roumen7  
#1 ·
Hello Everyone,
Like many here, I was quite excited to be finally upgrading after 7 long years(almost) to a brand new platform, and this weekend, i finally had all the parts to assemble my new system.
Everything went fine, surprisingly the build booted in 1 go and I didn't have to bother troubleshooting much(I have bad luck with these things).

I soon started loading up all kind of benches and monitoring programs(HWMonitor) and that's when i saw something I wasn't expecting from a premium board. The PCH temps are just insane, the VRMs are a little toasty but i can live with them, but the PCH.... well let me put the numbers and screenshots here for you :

1) PCH Idle Temp : 78C
2) PCH Temp under Gaming load : 90C + (I say plus because i stopped the games once i saw these sky high temps)

P.S Corroborated these readings with what BIOS shows, they are identical, so either the sensor is faulty or both BIOS and HWMonitor are giving accurate reading.

To give you some more context, here is my rig info(No Overclock) :
CPU : Intel i7 12700KF
CPU Cooler : Lian Li Galahad 360 AIO
Motherboard : MSI Z690 Carbon Wi-Fi
RAM : GSkill Trident Z5 5600Mhz CL36 16X2 Kit
GPU : Nvidia RTX2080 FE
PSU : Corsair HX850i
Cabinet : Lian Li O11 XL with 10 Fans
Side 3 and Bottom 3 set to Intake | Top 3 set as AIO Exhaust and Rear also set as Exhaust

Storage (All driven by PCI Lanes linked to PCH and not CPU) :
1x 500GB Samsung 970 EVO
2x 1TB WD SN850 Black in RAID0
1x 2TB Kingston NV1

Things I have tried to alleviate the problem :
1) Tried putting my GPU in the second and first slots, first slot ended up providing about 3 degrees relief, taking the idle PCH temp to 75
2) Set the PCH voltage to 0.820 fixed, reduced another 3 degrees at idle

But even with these 2 fixes this leaves me at 72-73C idle, which is still a lot and the PCH still shoots to 81-82 during short gaming sessions.
Image

Is there anything else I can try, can I lower PCH voltage further without any detrimental effects ?

I wanna give MSI the benefit of the doubt and it could be that this is the norm for Z690 considering how much this Chipset is providing in terms of I/O.
Or it could be a BIOS issue
Or it could be something I don't want to think about, ie poorly designed PCH heatsink which has mounting pressure/contact issues.

I would love to know what others who may have gotten this board are seeing on their systems. Apologies for the long post, would really appreciate some help here fromfellow community members.

Thanks
 
#3 ·
Thank you for your reply, but as i said, BIOS also shows the same reading, so it cannot be that HWMonitor is faulty. The moment i saw 90c on PCH, i rebooted the PC jumped into the BIOS, where i was greeted with 87C PCH temp.

But i will definitely give HWinfo a shot.
 
#4 ·
Sounds about right for a case that is not known for great airflow. Even in high air flow/open test bench the newer chipsets/pch run hot and seeing temps in the 70-80+ is not unusual. Even at idle they will be in the 50-60c+ range in an open test bench.

There are voltages you can lower and things you can turn off in your bios to lower temps 10-15c but in reality it doesnt matter. Those components can take those temps.
 
#5 ·
That is news to me, i knew AMD's X570 had similar issues and it even came with a chipset fan. but even in its case i never saw 90C+ temps being reported.

I am curious to know what are these chips rated for ? How high is too high for me to lose sleep over ?

I will try to take off the side and front glass panels of my case and do a test, and see how it fares.

Apart from that i also have another theory, please correct me if i am wrong :

I am thinking this might have something to do with the fact that i have 4 NVME, of which 2 are GEN 4 in RAID0.

I will try to do the following this coming week just to see if it helps :
1) Move the Gen 3 OS drive the 970 EVO to the top M.2 which is linked directly with CPU
2) Move the Gen 4 RAID drives to farther slots which dont share the heatsink with the PCH OR are far enough to give some breathing room to the PCH
3) Undervolting more, i know i am not gonna damage it by undervolting, but i wanna know how low is too low to affect day to day system stability and PCI E/IO performance/throughput.

Appreciate your input.
 
#6 · (Edited)
I dont know about AMD, I was refering to Intel z490 and z590 as I have owned and built those. Currenly also have z690 Strix d4. I think pch temps are in the 55-60c range at lower use, dont really pay attention to it.

The more drives you use and things like thundebolt or whatever else runs through pch will raise temps. Just the added heat from 4th gen ssds adds heat inside the case. Some of the 980 Pro NVME drives I have had idle temps are in the upper 50s to 60c range in cases with pretty good airflow.

I remember seeing people over on the Asus support forum posting worried about pch temps from z490/z590. I guess i havent heard of any failures from too high of pch temps on z490/z590 and they have been around a little while. But yeah lower temps are better in most cases.

Edit plus of you have an aircooled GPU dumping heat onto everything it is not surprising to see higher temps on everything it comes into contact with.
 
#7 · (Edited)
According to Asus, at least on the z590 chipsets, are fine with those temps and are rated quite high ( up to 108 deg C).but I still don't like it. There are posts everywhere with owners experiencing the same problem. apologies - the posts I was thinking about regard z590 temps.

I guess I lucked out with my z690 Hero.. my pch temps are typically in the high 50s during use.
If it were me, I would look at swapping out thermal pads / paste under the heatsink. However unsure if that would void your warranty 🤔 You should not have to turn down performance or turn off features for better temps on these pricey z690 boards imho.
 
  • Rep+
Reactions: Barry and ghost_z
#8 ·
Mobo manufacturers and Intel's standpoint is the z690 chipsets are fine with those temps and are rated quite high..but I still don't like it. There are posts everywhere with owners experiencing the same problem.
I guess I lucked out with my z690 Hero.. my pch temps are typically in the high 50s during use.
If it were me, I would look at swapping out thermal pads under the heatsink. However unsure if that would void your warranty 🤔 You should not have to turn down performance or turn off features for better temps on these pricey z690 boards imho.
I know right, i guess i will try moving the NVME drives around and see how it goes, otherwise, will use this board for a couple months till the pricing is sorted out, and then jump to another more stable and cooler board.
 
#13 ·
So my hunch was indeed correct. And the fact that i had the audacity too run a raid 0 on these gen 4 drives, o boy, maybe this motherboard or rather the chipset itself isn't cut out for that.

But just as you said doing the following made sure that i never see 90c ever again.

Moved the OS drive(even though its a gen 3 Samsung 970 evo) to CPU's m.2
Moved 1 of the raid0 WD SN850 drives from the middle heatsink/slot to the bottom one(in hopes that it will reduce the thermal load on the middle heatsink which directly connects to the PCH heatsink)

These 2 changes helped bring load temps down to 80-82c. Which is still high, but not as crazy as 90c+. I guess i got to live with it.

Will try using Gelid thermal pads as a last resort but that's about it, I don't see the point of RMA'ing the board and being without a PC for god knows how long in this starved market.

Appreciate you corroborating what i was seeing.
 
#17 ·
Re PCH temp., I have an ASUS Z690 Hero board, 12900K CPU, ASUS 3090TI, 32GB RAM at 5200MHz, two M.2 drives & four SATA drives. The PCH temp was reaching around 69°C with case fans off (Corsair 7000X) and about 65°C with them running at minimal speed. My particular case has an optional metal shroud to tidy internal appearance that channels front fan air flow away from the motherboard, so thought I'd try a simple trick. I purchased a 25 x 80 x 80mm fan (Arctic F8-80, 2,000rpm spec.) from Amazon for just over £4 and simply placed it under the GPU, close to the motherboard and pointing towards the chipset area. No need to mount it, it is quite happy just sitting there. With the case fans off, the PCH temp is now around 55°C. The fan has a 3 pin connector, runs at approx. 1,500rpm and is barely audible so thought I would share this. Placing it in the case and plugging it in is only a 1 minute job. I have made no BIOS adjustments, everything is running at defaults and I use no power saving features in Windows. As a side note, when I play games I have seen no significant increase in PCH temp, even before I fitted the extra little fan. I've seen nothing even close to 90°C, so maybe I've been lucky and just happen to have a well mounted chipset heatsink. Anyway, hope this helps somebody to keep their pride and joy a little cooler.
 
#19 ·
I hate to revive a old thread. But I figured I would give the solution to the problem for someone who finds this by googling.

Go inside your MSI bios, and set PCH voltage to “Auto” it’ll drop PCH temps about 20C+
 
#21 · (Edited)
I solved this issue. Seat your contact frames again. With the thermalright you must not make the last 1/4 turn which over tightens it. I initially dropped my PCH min voltage that got me from 80 degrees to 70. Stayed on that for 3 months. Now I'm on 55 degrees. Let's hope no permanent damage was caused to the m.2 drives. M.2 drives dropped temps also.
 
  • Rep+
Reactions: Oglethorpe