Nvidia 30X0 series waterblocks

Removing those pads means you won't have direct contact cooling to the VRM. Maybe try to find padding that absorbs vibrations better?

my understanding is this is typically what causes coil whine, vibrations not being completely absorbed by the thermal pad and/or backplate

interestingly enough EK's new active water block backplates promises to completely eliminate coil whine

They do not promise to completely eliminate coil whine rather possibly lessen whine: EK EXPO | CES 2021: Unlocking the Technology - ekwb.com

Another thing is the benefit of thermal pads between the PCB and the backplate since it reduces coil whine. Thermal pads act as absorbers – they are soft and could lessen the coil whine caused by the vibration of the inductor coils

So could and lessen, no promise of completely eliminating it. As soon completely and eliminate, will open themselves up for people saying misleading advertising etc. But your not wrong on the whole, if you can absorb the vibrations or remove them, all the better to try and reduce whine which in theory the backplate will try to do.
 
3090, Heatkiller block and backplate with an MP5Works backplate cooler.

My Memory Junction Temperature sits at around 56-58. It skips 57 at the moment!
Tried to buy one of these backeplate blocks but no stock. Got a ram block sat here but cant use it cos of the vertical mount...
 
Tried to buy one of these backeplate blocks but no stock. Got a ram block sat here but cant use it cos of the vertical mount...

I was unsure of how secure their mounting is as mine is vertical as well, but it's actually fine though I think I used the bigger clips. It's a fairly thick backplate. Maybe with more pressure the temperatures would be lower, but as two sets of thermal pads are already involved I can't see it making much of a difference!
 
I was unsure of how secure their mounting is as mine is vertical as well, but it's actually fine though I think I used the bigger clips. It's a fairly thick backplate. Maybe with more pressure the temperatures would be lower, but as two sets of thermal pads are already involved I can't see it making much of a difference!
Tbh getting it down the temps that you have now is more than sufficient for the longevity of the card... be interesting to see how long the cards last without the proper cooling of the ram chips on the backside. I can't see 90c+ over a few years be very healthy.

Did you plumb it in as instructed via parallel, think im gonna tap it in through the outlet on the res via a t piece then back out on the intake on the rad via another t piece. Just gonna look a little messy i think.
 
Tbh getting it down the temps that you have now is more than sufficient for the longevity of the card... be interesting to see how long the cards last without the proper cooling of the ram chips on the backside. I can't see 90c+ over a few years be very healthy.

Did you plumb it in as instructed via parallel, think im gonna tap it in through the outlet on the res via a t piece then back out on the intake on the rad via another t piece. Just gonna look a little messy i think.

I've been messing with fans again this wknd.

I am using a reference PCB as you know and regular EKWB block and backplate.

I've managed to get average GPU temps down to the low 50's across multiple game tests.

I took your advice on flipping back mounted fans and swapped top intakes back to exhausts.

Some higher mobo related temps, but better overall for GPU by a few degrees.

Just to let you know my VRM temps (with the HW Info update) get up to a maximum of 74c just gaming loads.

I heard they go a lot higher under mining or other higher use situations, but I am mainly gaming.

Pretty happy there I think. Will see what summer temps bring!

IMHO we do a lot to get the best temps possible,

I have had pretty good luck with components but I have also had cards randomly die for no good reason.

You could have a card running perfect temps and die and one at the limit that lasts till the apocalypse.

But TBH I think that is scheduled for this year anyway right? :D
 
Tbh getting it down the temps that you have now is more than sufficient for the longevity of the card... be interesting to see how long the cards last without the proper cooling of the ram chips on the backside. I can't see 90c+ over a few years be very healthy.

Did you plumb it in as instructed via parallel, think im gonna tap it in through the outlet on the res via a t piece then back out on the intake on the rad via another t piece. Just gonna look a little messy i think.
That's pretty much what I'll be doing apart from in will be from rad outlet and out back to res.
 
I've been messing with fans again this wknd.

I am using a reference PCB as you know and regular EKWB block and backplate.

I've managed to get average GPU temps down to the low 50's across multiple game tests.

I took your advice on flipping back mounted fans and swapped top intakes back to exhausts.

Some higher mobo related temps, but better overall for GPU by a few degrees.

Just to let you know my VRM temps (with the HW Info update) get up to a maximum of 74c just gaming loads.

I heard they go a lot higher under mining or other higher use situations, but I am mainly gaming.

Pretty happy there I think. Will see what summer temps bring!

IMHO we do a lot to get the best temps possible,

I have had pretty good luck with components but I have also had cards randomly die for no good reason.

You could have a card running perfect temps and die and one at the limit that lasts till the apocalypse.

But TBH I think that is scheduled for this year anyway right? :D

Not too shabby, glad the advice has paid off for your gpu temps at least. Was about what i was managing with the alphacool block around 52c max on that block on the current setup with bottom intake, back intake and 1 side intake with the top as exhaust (Dual loop though). I'm really tempted to use LM on the gpu as i saw a huge drop in temps with LM on the cpu block. To give a comparison (though its not direct since i'm using a different block now):

Bitspower summit EF - Using Kyronaut 58-65c depending on load
XSPC Raystorm EDGE - Liquid metal 48-52c depending on load

I reckon i could shave another 4-5c off the gpu with LM but its such a farce to apply, spent a good 20 mins slowly applying the paste to make sure its even. Literally 1 small dot was enough to do the cpu die.


You're vrm temps are still very good tbh, compared to mine when running something like prime95 i can easily hit 110c where it will throttle. It's a lot better now with a 40mm fan blowing over the heatsink but still gets hot.

Haha same, im hoping this card will last for a VERY long time yet... Though i have to admit everytime i disassemble and reassemble the card my heart skips a beat when i press the power button lol.
 
That's pretty much what I'll be doing apart from in will be from rad outlet and out back to res.
Works well either way, as the temps should normalise. I'd also be interested to see how that bumps up the water temps too. So far while gaming my gpu water temps stay around 28c and cpu around 26c. Never seen either go above 30 yet.
 
Yeah, don't think we'll see much of a rise, maybe a coupe of deg over a long session I expect. Keen to see the difference in avg clock!
The drop/increase in clocks for me at the next stepping point is 48c, so getting it to sit below 48c is my target though i dont think its achievable without mounting the gpu back horizontally and using LM. Could test it but a lot of faff haha.
 
Tbh getting it down the temps that you have now is more than sufficient for the longevity of the card... be interesting to see how long the cards last without the proper cooling of the ram chips on the backside. I can't see 90c+ over a few years be very healthy.

Did you plumb it in as instructed via parallel, think im gonna tap it in through the outlet on the res via a t piece then back out on the intake on the rad via another t piece. Just gonna look a little messy i think.

Yeah it's in parallel. The fittings they provide and the small diameter tubing mean you don't loose too much flow / pressure through the backplate waterblock as it's such narrow tubing. I doubt it really needs that much anyway.

I do have two pumps in series though, primarily for redundancy in case of single pump failure. The better head pressure is nice though, means I can run them slower and that I've been able to be quite cavalier with my use of right angle fittings to either keep things tidy or to squeeze everything into my case :D. (And I have a lot of places that need those right angles!)

I put two of these on the back of the card :

https://www.overclockers.co.uk/ek-w...t-adapter-fitting-black-nickel-wc-9ve-ek.html

Then ran the In with the In on the GPU and the Out with the Out. So yeah, parallel!

I didn't bother shortening the tubing, I just looped it as neatly as I could on the back of the card underneath the elastic loops used to attach the waterblock itself. Since your card is vertically mounted you could do the same and not have it look ugly.
 
Last edited:
Yeah it's in parallel. The fittings they provide and the small diameter tubing mean you don't loose too much flow / pressure through the backplate waterblock as it's such narrow tubing. I doubt it really needs that much anyway.

I do have two pumps in series though, primarily for redundancy in case of single pump failure. The better head pressure is nice though, means I can run them slower and that I've been able to be quite cavalier with my use of right angle fittings to either keep things tidy or to squeeze everything into my case :D. (And I have a lot of places that need those right angles!)

I put two of these on the back of the card :

https://www.overclockers.co.uk/ek-w...t-adapter-fitting-black-nickel-wc-9ve-ek.html

Then ran the In with the In on the GPU and the Out with the Out. So yeah, parallel!

I didn't bother shortening the tubing, I just looped it as neatly as I could on the back of the card underneath the elastic loops used to attach the waterblock itself. Since your card is vertically mounted you could do the same and not have it look ugly.
I've seen on facebook they're advertising a Serial version now with normal G1/4 ports at the top of the block... was wondering if thats a better choice to begin with but now im not so sure as i don't have enough room to deploy 2 90 degree fittings onto there without fouling onto the ram modules. Might just have to stick with the Parallel version when its back in stock.

Looked at some temps tonight and my my GPU memory junction temp is around 76c which is perfectly fine i think, don't think i would require any further cooling on that part tbh. For some reason i was expecting to see high 90's lol.
 
Not too shabby, glad the advice has paid off for your gpu temps at least. Was about what i was managing with the alphacool block around 52c max on that block on the current setup with bottom intake, back intake and 1 side intake with the top as exhaust (Dual loop though). I'm really tempted to use LM on the gpu as i saw a huge drop in temps with LM on the cpu block. To give a comparison (though its not direct since i'm using a different block now):

Bitspower summit EF - Using Kyronaut 58-65c depending on load
XSPC Raystorm EDGE - Liquid metal 48-52c depending on load

I reckon i could shave another 4-5c off the gpu with LM but its such a farce to apply, spent a good 20 mins slowly applying the paste to make sure its even. Literally 1 small dot was enough to do the cpu die.


You're vrm temps are still very good tbh, compared to mine when running something like prime95 i can easily hit 110c where it will throttle. It's a lot better now with a 40mm fan blowing over the heatsink but still gets hot.

Haha same, im hoping this card will last for a VERY long time yet... Though i have to admit everytime i disassemble and reassemble the card my heart skips a beat when i press the power button lol.

Thanks for the info.

I am afraid of LM due to a leak from a delid chip I once had!

It was one from OCUK where they do the job themselves.

Unfortunately, the seal had broken between the IHS and PCB.

I noticed I randomly started getting high temps and took the cooler off and the LM was so close to the mobo socket where it had seeped out.

Since then, I have been a little worried about the risk to reward ratio!

I guess with watercooling and LM, that is sometimes the chance you take.

You just have to be very careful, which I am sure you are.

Get out the old nail varnish :cool:

I know exactly what you mean. When I completed my first ever GPU waterblock and the rest of the system, I prayed to the PC gods to let it work!
 
So, on the EKWB 3080 FE Blocks, is there a layer of plastic seal on the acrylic window? It seems a bit duller than my FLT reservoirs but I don't want to start scratching around the edge in case there isn't anything there.
 
Thanks for the info.

I am afraid of LM due to a leak from a delid chip I once had!

It was one from OCUK where they do the job themselves.

Unfortunately, the seal had broken between the IHS and PCB.

I noticed I randomly started getting high temps and took the cooler off and the LM was so close to the mobo socket where it had seeped out.

Since then, I have been a little worried about the risk to reward ratio!

I guess with watercooling and LM, that is sometimes the chance you take.

You just have to be very careful, which I am sure you are.

Get out the old nail varnish :cool:

I know exactly what you mean. When I completed my first ever GPU waterblock and the rest of the system, I prayed to the PC gods to let it work!
It's not that bad, if you apply it correctly theres no issue honestly. Don't be afraid to use it, just make sure you don't use aluminum with it and you'll be fine.

I'm about to do some further testing with an external radiator using QD's. Just waiting on delivery but gonna start small with a slim 360 i have laying around and see what difference it makes before i make the big jump to MoRa3 big rad.
 
Back
Top Bottom