Planning to delid my 7700k - Some clarification needed

[Thaedalin]

Afternoon folks,

My i7-7700k runs extremely hot. I found that this was mostly due to the stock VCore sitting between 1.35-1.40. I would often peak 95 degree c playing games. I eventually managed to lower the VCore to about 1.25 stable at 4.5, but the temps still creep up into the 80s with my 212 evo. I understand that these temperatures are fine, but I want them lower in order to raise my clocks from 4.5 to 4.8+

I am planning to delid my 7700k to replace the TIM with Thermal Grizzly Conductonaut. I should have everything delivered by next week to first practise on my old 6600k and then move on to my 7700k assuming that everything goes alright.

My first question is regarding the optional step that everybody seems to have some form of different opinion on: Nail polish for nearby components. I am aware that none of the conductonaut should spill over if you use the right amount and spread it out correctly. However, for such a cheap price why not have the added security?

Will this nail polish be fit for purpose? https://www.amazon.co.uk/gp/product/B00UYJGYCK/ref=ppx_yo_dt_b_asin_image_o00_s00?ie=UTF8&psc=1

Ingredients can be found here: https://www.rimmellondon.com/en-gb/nails/nail-polishes/60-seconds-super-shine-nail-polish

My second question is regarding the relid stage. I have again seen multiple opinions on this. People say that you need to relid and match the sealant perfectly to how Intel had it, others say you can do small dots in each corner, and others say that you don't have to do any of that at all.

If I go down the route of just leaving the IHS atop the processor (as I quite like the idea of having no glue to ensure a close to perfect contact with the die and IHS), will the pressure from the retention bracket ensure that the IHS does not move and that the liquid metal covering the die also does not move?

This leads me onto my third question. Again, differing opinions everywhere I look. Should I apply the liquid metal TIM replacement to the die or the die and IHS? What are the differences?

 

[Justintime]

Never used the thermal grizzly but have used nail polish on chips with coolaboraty liquid metal, works fine. However in the early years i never used any insulator and never had an issue. As for relidding i never reglue and the pressure is more than enough to prevent movement. I apply to both die and IHS, perhaps not quite as much on the IHS. Have seen as much as 20C difference with some chips.

 

[pandem0nium]

I've delided a few in my time.

I've never done anything to the surrounding components and haven't had any problems.
A £5 chinese ebay special vice tool will do the job just fine.
No need to glue the lid back on, it can be held in place just fine with the motherboard socket retention bracket, and the temps will probably be better this way anyway. Position the IHS a few mm higher in the socket before clamping down the bracket as it'll drag it down a bit.
Apply to the die and bottom of the IHS. First time I tried just the die. Contact and temps weren't very good and I had to re-do it.

 

[Thaedalin]

Never used the thermal grizzly but have used nail polish on chips with coolaboraty liquid metal, works fine. However in the early years i never used any insulator and never had an issue. As for relidding i never reglue and the pressure is more than enough to prevent movement. I apply to both die and IHS, perhaps not quite as much on the IHS. Have seen as much as 20C difference with some chips.

I've delided a few in my time.

I've never done anything to the surrounding components and haven't had any problems.
A £5 chinese ebay special vice tool will do the job just fine.
No need to glue the lid back on, it can be held in place just fine with the motherboard socket retention bracket, and the temps will probably be better this way anyway. Position the IHS a few mm higher in the socket before clamping down the bracket as it'll drag it down a bit.
Apply to the die and bottom of the IHS. First time I tried just the die. Contact and temps weren't very good and I had to re-do it.

For repositioning without any glue, am I save to mark a line with a pen from the top of the IHS onto the substrate to check the line up with the retention bracket? I'm paranoid that without gluing it back down that I'll place it incorrectly. I've bought black sealant in the event that I need to RMA in the future. I'll remember to take pictures of the die and surrounding areas after I delid in that event.

Edit: I did look at coolaboraty liquid metal, but I seen that grizzly had a slight advantage in temps (sometimes ranging from -1 to -4 degree c)

 

[Justintime]

When you remove the IHS and clean the epoxy off it'll still leave an outline on the PCB so don't worry about alignment you'll see if its wrong. I only use the coolabratory because its all i have around and have used it for maybe a decade? To be honest most times i use MX4 if its not a massive overclocking deal or its just to bring temps down on a stock non K chip for less noise/silence, difference is not a great deal at all and its less annoying than liquid metal stuff plus easier to clean. I know i'm brutal but i've used a mallet, vice and block of wood from day one and never lost a chip

Also, something to note - If your case cooling or heatsink is subpar then it may not help at all. The only difference is it might take longer to reach thermal throttling than before and cool down quicker when theres no load.

 

[Thaedalin]

When you remove the IHS and clean the epoxy off it'll still leave an outline on the PCB so don't worry about alignment you'll see if its wrong. I only use the coolabratory because its all i have around and have used it for maybe a decade? To be honest most times i use MX4 if its not a massive overclocking deal or its just to bring temps down on a stock non K chip for less noise/silence, difference is not a great deal at all and its less annoying than liquid metal stuff plus easier to clean. I know i'm brutal but i've used a mallet, vice and block of wood from day one and never lost a chip

Also, something to note - If your case cooling or heatsink is subpar then it may not help at all. The only difference is it might take longer to reach thermal throttling than before and cool down quicker when theres no load.

My case cooling is as good as I can get it. I've got 2 front fans doing in take, and 3 back fans doing exhaust. I've got a 212 evo with a double fan setup, one blowing toward rear fans and the other exhausting toward rear fans. The only component that gets too hot for my liking is CPU, and that's when the side panel is on. Side panel off, it can still hit 80-85 but I'm convinced its because of how hot the 7700k runs due to the rubber scraps they use as TIM

I've got my old 6600k to practise on first before I move onto the 7700k, but I'm fairly confident I can get it first time. Seems like more of a common sense job than anything too difficult to do

 

[pandem0nium]

For repositioning without any glue, am I save to mark a line with a pen from the top of the IHS onto the substrate to check the line up with the retention bracket? I'm paranoid that without gluing it back down that I'll place it incorrectly. I've bought black sealant in the event that I need to RMA in the future. I'll remember to take pictures of the die and surrounding areas after I delid in that event.

Edit: I did look at coolaboraty liquid metal, but I seen that grizzly had a slight advantage in temps (sometimes ranging from -1 to -4 degree c)

It's not really an exact science, and you can't go too far wrong. It'll void the warranty, so you won't be able to RMA anyway.

 

[Justintime]

Just don't do like me, i once delidded/dedied a 2500k by taking the wrong chip from the tray

 

[Thaedalin]

It's not really an exact science, and you can't go too far wrong. It'll void the warranty, so you won't be able to RMA anyway.

Just don't do like me, i once delidded/dedied a 2500k by taking the wrong chip from the tray

Final question for you both: After delidding the cpu and separating the heat spreader from the PCB, can I use 99.9% isopropyl to clean the PCB without any worry that it may degrade the product in any way? I'll be using scotch tape to cover up the PCB before applying the liquid metal, and I'm slightly conscious of any residue the scotch tape may leave behind on the PCB. I know it shouldn't harm anything if it does, but I'd rather give it a once over with isopropyl after stripping the tape.

 

[MrRockliffe]

Before delidding, surely upgrade the cooler?

 

[Akagi]

Just don't do like me, i once delidded/dedied a 2500k by taking the wrong chip from the tray

Huh?

Before delidding, surely upgrade the cooler?

You would think so, but people are strange.

 

[Gray2233]

Huh?

Sandy was soldered, it wasn't until Ivy Bridge that Intel started using the thermal gunk.

 

[Akagi]

Sandy was soldered, it wasn't until Ivy Bridge that Intel started using the thermal gunk.

Oh I see he's saying he pulled the die off because he picked up a 2500K by mistake.

 

[Gray2233]

Oh I see he's saying he pulled the die off because he picked up a 2500K by mistake.

Yup.

Not the first time I've heard of that happening to some poor sod.

 

[random_matt]

You can always spread a very thin layer of sealant on the IHS edges.

 

[Thaedalin]

Before delidding, surely upgrade the cooler?

Nothing wrong with the 212 evo as a cooler? Also don't exactly have the space for a Noctua if that's what you're suggesting.

It's also just common knowledge that 7700k's run hot and always have. Even Noctua/water coolers have the same results.

 

[MrRockliffe]

Nothing wrong with the 212 evo as a cooler? Also don't exactly have the space for a Noctua if that's what you're suggesting.

It's also just common knowledge that 7700k's run hot and always have. Even Noctua/water coolers have the same results.

Well there is something wrong with it if you’re trying to run 1.35V and hitting 95 degrees in the process...

And no they don’t run hot, no yes a high end air cooler/AIO will make a difference to the thermal performance because of their larger quantity of fins and heat pipes, and having more fans.

 

[Gray2233]

Well there is something wrong with it if you’re trying to run 1.35V and hitting 95 degrees in the process...

Not necessarily.

The thermal gunk under the IHS could have degraded and be causing issues, in fact I find that more likely than an Evo 212 spontaneously going bad.

That said, other factors such as the current weather leading to high ambient temps and potential issues with case airflow could be at fault.

I'd also like to add that given how much people are paying for second hand 6600K and 7700K's at the moment and the supporting motherboards, I'd be tempted to put the two CPU's and motherboard up for sale and switch to a newer platform over the hassle and risk of delidding.

 

[MrRockliffe]

Not necessarily.

The thermal gunk under the IHS could have degraded and be causing issues, in fact I find that more likely than an Evo 212 spontaneously going bad.

That said, other factors such as the current weather leading to high ambient temps and potential issues with case airflow could be at fault.

I'd also like to add that given how much people are paying for second hand 6600K and 7700K's at the moment and the supporting motherboards, I'd be tempted to put the two CPU's and motherboard up for sale and switch to a newer platform over the hassle and risk of delidding.

Perhaps not in other situations, but having owned a 6700K, and now a 9700K, I’ve seen a notable thermal performance uplift every time I’ve upgraded the cooler.

If he wants to run a 7700K at 1.35V, it’s considerably less risky, quicker and easier to get a better cooler.

 

[Justintime]

Final question for you both: After delidding the cpu and separating the heat spreader from the PCB, can I use 99.9% isopropyl to clean the PCB without any worry that it may degrade the product in any way? I'll be using scotch tape to cover up the PCB before applying the liquid metal, and I'm slightly conscious of any residue the scotch tape may leave behind on the PCB. I know it shouldn't harm anything if it does, but I'd rather give it a once over with isopropyl after stripping the tape.

Residue won't harm, its non conductive and yes the alcohol is fine it won't degrade anything.