SOLVED 3770K Large idle to load temp increase

I recently upgraded to Ivy and started overclocking the 3770K and noticed higher idle temps than most people write about in the threads here with similar voltages/clock speeds. I let it go because it was not that bad within a few degrees C. I OC'd to 4.5GHZ with CPU volts @ 1.34 and only using the multi. 45 x 100. @ 4.5 my idle temp is ridiculous. 52 52 56 52. load, forget it. upwards of 90c. Why such a big jump? Why so hot @ idle?
I also read about de-lidding the CPU and reseating the top but I don't know if I am ready for that or not yet.
I resat cpu (using pea size TIM method in case it matters)
checked for air in lines and radiator
made sure the pump was moving the water like it should

Water loop is swiffy mcp 655, 3/8 lines, corsair RAM block, Raystorm CPU block, stock Maximus V block, 240 Rad, Reservoir

I appreciate any help.
Thanks

Well rainman maybe this video could answer some of your question here. It seems you have a very hot CPU that really needs to be De-lidding! So watch and see if this is about right for you. AJ.

1,
As this video says you can have bad Temps even @ idle before you OC!

IB always has a big jump from idle to load.
You should be able to run 4.5GHz with a lot less vcore than that, try 1.3, if that works then try 1.25.

Ugh, stop with the delidding advice people, LOL! Its not for the feint of heart yet thrown out to do like its a bios setting or something!!!

A couple of things I notice... first, that loop is pretty full in that you have a couple of restrictive and frankly, unnecessary blocks in there (ram and chipset). That said, you have enough pump, not so sure about raddage cooling everything on the board well though. You also didnt mention your ambient temperatures either. Those play a huge role in idle temps.

I also think that voltage is high for the clockspeed. Is that the lowest it will go?

As far as the deviation, that is about normal.. I idle (different system, dont compare), at 27C or so, and load around 60-65C depending on ambient.

Bobnova said:

IB always has a big jump from idle to load.
You should be able to run 4.5GHz with a lot less vcore than that, try 1.3, if that works then try 1.25.

Click to expand...

Will do thanks. I started messing with this using Asus' pre set OC profile. Probably why my vcore is so high.

Earthdog- My ambient here the past 3 days is like 75F, 96 F outside.....
Crazy hot. I know the RAM block is a waste for moderate OC I just like Watercooling! lol I did notice the stock waterblock on the board is restrictive so maybe it's time to rethink the loop.

i started to answer to Ajay57 before you guys responded and de lidding is a big deal and I don't know if I am ready for that but from the video is doesn't look to bad just hate to mess up and ruin it!
Thanks guys i am going to try said advice and see what happens.

What you might consider is a pair of Ts, run the RAM/MOSFET blocks in parallel to the CPU.
That way they will still get some cooling (not a lot, admitedly, but they need roughly squat anyway) and the CPU will get a lot more flow than it does right now.
That's what I'd do, anyway.

Sorry about the De-lidding advice, but for some people its the way to go, if all else fails to work!! Maybe next time fully noted ED!!

It really isn't, though.
It provides a drop in temps that might allow a slightly higher OC, an OC that won't be noticeable anywhere except benchmarking.
At the guarantee of killing the warranty (while enabling more voltage to be used, which will kill the chip faster) and potentially killing the CPU on the spot.

Honestly, it sounds like a bad mount...maybe too much or too little TIM. Temps can be pretty sensitive to how much TIM you use. What I usually do is apply the TIM, squeeze my block down on it, then remove it and take a look. If you've got too much (like it's spilling out over the sides of the IHS), wipe off the excess with a q-tip. If there's too little (so it's not reaching all the corners of the IHS), add a tiny bit more. When you're set, squeeze it down and slide it around a few mm in each direction.

Bobnova-I do bench for fun, but I also do not want to kill the chip for a couple extra MHz when I am not going to place very high on the BOT anyway since my cooling/knowledge is limited to water not extreme cooling and everyone benched this chip. It's not like it is a P4 925 or something where a couple extra MHz MAY get me a higher ranking. So what you say makes perfect sense. Thanks for eliminating that option completely. Although I considered it, I really didn't want to have to do it.

LoneWolf121188 - I did apply a tad too much TIM. It was oozing off of the side of the IHS just a tad so when I reassembled it I made a smaller pea size,applied the block, removed it and it was good so I just cleaned and repeated. Thanks

As far as the problem I found that as said the loop was restriced a bit. How ever my main problem was my coolant. I used Mayhems Aurora Tharsis Red but must not have mixed enough demineralized water with it. When disassembling my loop I noticed a white residue on the tubing and in the res. I also noticed it in my rad ports so I decided to check the cpu block. When I took the copper plate off, the metal restricting plate or water channeling plate (sorry do not know the correct name for this part) was completely clogged with this white residue which I believe is the "Aurora" additive being too thick. I reassembled the block and the loop and just used demineralized water with no dyes and it is like night and day. Has anyone else had this or heard of this problem with the Mayhems Aurora? If I add more water will it kill the Aurora effect?

Thanks for the help guys.

Distilled + PT Nuke/silver kill coil only. You now know the reasons we incessantly suggest that..

Did you add a biocide with the water? You really should if you didn't....that would be PT Nuke or a silver kill coil as ED mentioned.

Earthdog - Now I see why first hand! It was a mess inside that block.

No biocide but I do have the silver kill coil. I will be ordering the biocide today. I just thought the coolant was pretty cool looking so I figured I would give it a try. I was aware of residues from dyes thanks to info on here from you guys way before I bought this stuff but I didn't think it would be THAT bad! Not worth it! I flushed everything out of the rad and res with demineralized/distilled water to make sure no residual was left. Anyway thanks again for all your help and suggestions.