Removing Accelerator Plate from EK - Supreme?

[gazmtk]

Hey everyone :)
I know ive seen it somewhere on here in the past, but I cant seem to find it again :shrug:
What happens to the performance of the EK Supreme when the accelerator plate is removed??
and Eddy where is the high flow plate that Ive read about while looking for this?
I would gladly purchase one if it was made available
Regards Gaz

[twwen2]

I've also thought about this. Even dremelling one or two of the slits to make them wider.
I wonder what performance would be like without the plate at all? No doubt the block would be more free-flowing.

[BlazinTiger]

i dunno where to get them, but heres a review for alternative accelerator plates
http://www.xtremesystems.org/forums/showthread.php?t=187753
(looking at the results, plate #5 [see link] looks best results wise and w/o plate performs pretty badly)
maybe u could dremel to these designs?

[gazmtk]

hey BlazinTiger,
Thanx heaps, this wasnt what I saw before, but its just as good,
from what I can gather from this review you have linked, the EK Supreme is on par with a stock D-Tek?
so out comes the plate, and here comes the test of performance, im hoping the flow will be crazy now :)
Regards Gaz

[gazmtk]

Ok ive just played with my Supreme, took the plate out, and im glad i pulled it apart because it needed some serious TLC inside, the plate was covered with slug from when it was laser cut, so i gave it a sand, and cleaned out the micro chanels that had slug that had detached from the plate.
Now the fun bit, running a MCW30, EK supreme and a Micro Res off a DB1 was the test, no RAD here yet, was too lazy to grab it and throw it in, sorry
ok with the plate in.... it trickled out, like i knew it would be bad.... but that bad!!! and once i took the plate out, i actually got a decent flow! not surprising tho... its missing the restrictor!
now heres the next step, and dont worry i will be taking photos from here on in! Testing the thermals with and with out the plate, honestly if its running even close to AVERAGE temps with out the plate then its gonna stay out! if not off to go get a copper plate cut with wider slits!
any thoughts on this?
Eddy if you read this could you please PM me if you have any information on this??

[iboomalot]

if I were worried about flow due to it being so restrictive I would look at the XSPC Edge block instead

[gazmtk]

Why look at XSPC products when I have some great EK blocks already?? it looks great, but unless I want to go another D5 then i need to free her up some more...

[Martinm210]

The mod that I would try is chamfering the leading edge of the accelerator plate slots. There is alot of entrance friction losses occurring because the plates have a sharp entrance. A little dremel action would take care of that and may make a difference, but I wouldn't run it without the plate.

[systemviper]

Wow, some interesting idea's/. I am looking for clear acrylic top blocks, since i love UV liquid.
I love the MC-TDX and the MPC-Universal combo with their clear acrylic top and small rod
shaped cooling, it just has a great look to it.

But i also want to use the best and i loved the design of the XSPC Edge Acrylic and
almost purchased the EK acrylic supreme but then saw the XSPC Edge Acrylic and
was sold on them

But now i am thinking about the issues with the XSPC resi, that the CPU blocks
need to be checked very carefully.... with that said i might flop back to the
EK supreme acrylic just because i know it's top notch. So with lowering
restriction is a exciting topic..... thanks and keep us posted!

[iboomalot]

why Edge???

1. similar look
2. use martin's flow ecell spread sheet 2.72 gal/min vs 1.28 gal/min I can put one EDGE + two EK full GPU blocks + two MCW-30s + two Asus Mosfet blocks and still show 1.43gal/min
3. Resale value on your Supreme is good and you would loose very little money since people do want that block
4. frees up flow thus you don't need another D5
5. cost to replace the modded plate
6. Edge seems to work as good or better at keeping temps down
7. This block is restrictive and your increasing flow might up temps

IMO way to restrictive for the temps it gets thats one reason Iam replacing my Storm

its your money and time and I wish you good luck and what does having an EDGE CPU block have to do with your other EK blocks???

I plan on my next project using EK Mosfet blocks but not the EK Supreme for the above reasons

[systemviper]

Yea XSPC Edge :up:

[gazmtk]

Iboomalot,
Sorry I was tired when I posted last, wasn't meaning to be rude,
The rest of the blocks im probably going to use are EK blocks, plus I already have my EK supreme.... That said I am going to try what Martin has said (thanx Martin!) and if not I will wait until the XSPC Edge hits Australia...
Also I sent Eddy a Message about maybe sending me a couple more plates (i'll pay for them of course) so i can mod them, maybe build a test bench to see changes in temps and see how the flow changes with it, and if I was to reach a good result send it to martin for proper testing.
anyway thanx everyone for your input, and Iboomalot sorry for rudeness, i didnt mean it i can assure you
Regards Gaz

[iboomalot]

didn't take it as rude and you had a liget question :up:

let us know how that goes EK makes nice prouct but that block IMO needs a revision.

off to watch a good movie "NEXT"

[systemviper]

"NEXT"

next what?

[aspire.comptech]

I think thats the name of the movie. Hence the quotes around it...

[gazmtk]

I just followed Martins Idea of smoothing the edges of the accelerator slits, i quickly tested the water flow by my crude setup of testing...
it made a difference, but not the difference I needed sadly! :P
but thanx alot Martin for the idea, im thinking make 2 slits into one, and going from there..... any thoughts on that?
I will not stop until I find an acceptable middle ground for flow VS performance ;)
im still going to try it with no plate to see if there is a CHANCE of running it without the plate safely, but after Martins comment, i doubt it will be, but doesnt hurt to try right?!
Regards Gaz

[Pedalmonkey]

if you have a high pressure loop, which it seems you dont, removing the plate would be detrimental b/c the turbulence in the water over the pins aids in temp transfer. however maybe you can make the EK into a block that can compete with the Edge in flow :). however for a low flow block the XSPC is much better b/c it was designed to work without those accelerator plates

[Martinm210]

I just followed Martins Idea of smoothing the edges of the accelerator slits, i quickly tested the water flow by my crude setup of testing...
it made a difference, but not the difference I needed sadly! :P
but thanx alot Martin for the idea, im thinking make 2 slits into one, and going from there..... any thoughts on that?
I will not stop until I find an acceptable middle ground for flow VS performance ;)
im still going to try it with no plate to see if there is a CHANCE of running it without the plate safely, but after Martins comment, i doubt it will be, but doesnt hurt to try right?!
Regards Gaz


Yeah, never hurts to try, and the large slot idea might help too. Hard to say.... the one test we've seen with the different accelerator plates didn't seem to make much difference which plate was used. That' similar to my lack of finding any real nozzle performance from the V2.

The differences might be there, but with a 2 degree mount variability, it's just hard to measure without doing a dozen mounts...

[iboomalot]

NEXT is a pretty good movie with a nice story line and isn't too long or short only 1.5 hrs.

[gazmtk]

alright guys, today I am building a test rig! im gonna try to test thermal loads with the rounded plate that ive already modified, and then remove the plate all together and see.... im also going to swap the DB-1 with the D5 and see what improvements come to light ;)
ok thats all
Regards Gaz

[gazmtk]

Alright here are the rough numbers running off a DB-1 pump!!
the loop is as follows || DB1 => EK Supreme => MCR240 => DD round Res => DB1 ||
the CPU is a AMD X2 4200 (939) and its clocked to 2.2 ghz, 71.9 watts, I would overclock but im working with Dodgy ram!
I know the the X2's dont put out much heat! but its a good starting block, i know that with a Apogee GT it used to run at 33 degrees celcius with an ambient of 20 degrees (as it was doing this morning!)
The operating system is Win XP Pro SP2 btw
now using Prime 95 and core temp i have reached the numbers I am about to post, this is by no means accurate, but this is just my findings.
on Idle with No plate its 28°c for core 1 and 25°c for core 2, and under full load its 38°c on core 1 and 37°c on core 2
now with the plate back in its a different story, the flow really kills the temps!
Idle is 32°c on core 1 and 35°c core 2 (i think the mount maybe be off, as core 2 is normally cooler! and at full load its 55°c on core 1 and 56°c on core 2
thats a HUGE difference, the ambient temps stayed between 20°c and 22°c, I will be trying with the D5 tomorrow to see if i can reverse the results
but in argument, the DB-1 is pretty weak.
Martin can you please test the pressure of the EK with no plate, it seems to cool better than expected!
Well thats my findings atm, I will be retesting properly soon so I will take screen shots then once the PC is back in the house and not in the Garage!
Im definately gonna have to reseat the block over and over to get a good average, the scores on the Supreme with the plate in dont seem right to me, but it could also be the pump!
anyway thats all for now,
Screenshots will be coming up in the next couple of days
Regards Gaz

[gazmtk]

Ok didnt get to throw the D5 on today, but I did manage to overclock the Processor to 2.75ghz with crappy ram,
scary thing is there is only a 3°c increase of tempratures when running with out the plate!!! thats AWSOME!
having WC this Mobo and processor in the past I know what temps it normally produces! and its running awsome with the EK supreme with out the plate
I want to get some screenies soon for everyone to see, but I first need to get the Network drivers installed, no network atm since ive brought it inside
will keep you all posted
Regards Gaz

[Martinm210]

Interesting, good work!

Sure, I can try it without the plate...

[iboomalot]

3c is a pretty big jump when looking at test comparisons between blocks but overall 50c vs 53c might not matter

keep motoring :up:

[gazmtk]

Oops Sorry, was tired when posted AGAIN! but I meant that it only increased by 3c when i overclocked it while running without the plate, hope this clears things up!
Regards Gaz

P.S. will fix the network today so I can screen dump for you guys and girls! :)

[gazmtk]

Martin that would be awsome if you could test the pressure drop with out the plate ;)
thanx mate