GTZ core i7 loop order

[Splave]

First off components

d5 Vario
GTZ w/1366 kit
Black Ice 3x120
Micro Res
3x Slipstreams
and 7/16 5/8 tubing
All on dd torture rack


Because GTZ uses injectors and is I guess you could say and accelerator, I would assume it would benefit having the pump right before it. I came to this conclusions by starting linX, then turning the speed of the pump down half way (temps reached as high as 86c and rising) then turned it up all the way (temps lowered and peaked at 68c after 3 minutes)

Does this order sound appropriate?
PUMP-Block-Rad-Res

Im currently running
PUMP-Rad-Block-Res

Which I figure now I must have been drunk when I configured it :rofl:
LMK what you guys think :up:

[WaterFlex]

3x Slipstreams
I would get better fans. Slipstreams suck a lot. Use Scythe S-FLEX 12cm 1900RPM (S-FLEX SFF21 (G)) for your Black Ice.

[NaeKuh]

pump b4 block on the GTZ.

[Splave]

so its worth it to change then? I mean Im redoing the loop tomorrow regardless

The fans are alright I have them shrouded but they are loud as hell but move quite a bit of air. I unplug two of them when Im not benching or stability testing and temps are still fine.
I was using medium speed yate loons one of them started buzzing and the other lost a blade.

Thanks guys

[eligray]

loop order is wayyyyyy overrated. No matter how things are ordered, you will have the same overall restriction and flow rate everywhere in the loop. I think I remember that even if you're only pumping at 1 gpm, it takes something like 400W to heat the water up 1 degree on its way by. IMO (which appears to be backed up by fact), order is virtually irrelevant. A more likely reason for temps is an improper mount on CPU. I own multiple WC loops, including most parts in your rig. Before rearranging, at least post why you think it will help.

cheers
Eli

[John Planet]

It may not make too much of a difference in overall scope, but my GTZ is plumbed directly after my 120.3. Modern rads are far less restrictive than any commercial CPU block and it makes sense to me to have recently dissipated coolant slamming into my GTZ with no heat source in between.

[naokaji]

A more likely reason for temps is an improper mount on CPU.

The wonderful thing about the GTZ is that you can't do an improper mount, the screws are done in a way that you end up with a even mounting pressure, no more multiple attempts needed like with *insert other block of choice*. So as long as you don't screw up on the tim it's a guaranteed good mount.

[Splave]

loop order is wayyyyyy overrated. No matter how things are ordered, you will have the same overall restriction and flow rate everywhere in the loop. I think I remember that even if you're only pumping at 1 gpm, it takes something like 400W to heat the water up 1 degree on its way by. IMO (which appears to be backed up by fact), order is virtually irrelevant. A more likely reason for temps is an improper mount on CPU. I own multiple WC loops, including most parts in your rig. Before rearranging, at least post why you think it will help.

cheers
Eli

did you read my post?

"I came to this conclusions by starting linX, then turning the speed of the pump down half way (temps reached as high as 86c and rising) then turned it up all the way (temps lowered and peaked at 68c after 3 minutes)"

so its obvious it likes extra pressure

[eligray]

did you read my post?

"I came to this conclusions by starting linX, then turning the speed of the pump down half way (temps reached as high as 86c and rising) then turned it up all the way (temps lowered and peaked at 68c after 3 minutes)"

so its obvious it likes extra pressure

It doesn't care about pressure, its all in the flow rate. Just trying to help...


The wonderful thing about the GTZ is that you can't do an improper mount, the screws are done in a way that you end up with a even mounting pressure, no more multiple attempts needed like with *insert other block of choice*. So as long as you don't screw up on the tim it's a guaranteed good mount.

I have the block, and mounted one on an i7 just yesterday. As you said, it is easy to use incorrect amounts of thermal paste, even if the screws are dummy-proof.


It may not make too much of a difference in overall scope, but my GTZ is plumbed directly after my 120.3. Modern rads are far less restrictive than any commercial CPU block and it makes sense to me to have recently dissipated coolant slamming into my GTZ with no heat source in between.

Yes, i agree. This way will get you better temperatures than putting in heat before the waterblock. The issue is the large amounts of energy required to heat water. So though you will have better coolant temperatures, it would be on the order of a fraction of a degree, and nothing near responsible for 8xc temperatures. I can turn my WC fans off completely, and never break 45c on a stress test. With those high temps, something is very wrong.

[warriorpoet]

It doesn't care about pressure, its all in the flow rate. Just trying to help...

Got that backward ;)

[Splave]

never break 45c on an i7 fanless during a stress test is probably a stretch but okay.

Temps are fine when I have the pump at full bore, bad when even %50 so why not put it pump before block?
Even as you say its all in the flow rate, the flow is the highest right after the pump is it not?

thanks for the help

[warriorpoet]

never break 45c on an i7 fanless during a stress test is probably a stretch but okay.

Temps are fine when I have the pump at full bore, bad when even %50 so why not put it pump before block?
Even as you say its all in the flow rate, the flow is the highest right after the pump is it not?

thanks for the helpPressure is highest; flow is constant :up:

[eligray]

never break 45c on an i7 fanless during a stress test is probably a stretch but okay.



Sorry about that, i didn't mean on the i7. Haven't temp tested that yet.


Pressure is highest; flow is constant :up:

Do you have any information on as to why this might be? I'd be really interested to hear it :)

[eligray]

Temps are fine when I have the pump at full bore, bad when even %50 so why not put it pump before block?
Even as you say its all in the flow rate, the flow is the highest right after the pump is it not?

thanks for the help

Go ahead and try it, it can't hurt things.

And no, this is exactly what I meant. The flow through one area cannot be any higher than it is through any other area. Otherwise your pump would be pushing away more liquid than its sucking in ;)

So go ahead and make the change, but only if you believe that the changed pressure will help you out, and not a flowrate change.

So if the pressure statement is correct, then i should really be running a very tall T-line. Is this true as far as you know?

[warriorpoet]

Do you have any information on as to why this might be? I'd be really interested to hear it :)Should be in the stickies. Keep in mind-- we're talking about closed loops here.

[eligray]

Should be in the stickies. Keep in mind-- we're talking about closed loops here.

as opposed to...

[eligray]

I hope that by pressure you don't mean the pump's max. head pressure. All that's good for is overcoming resistance in the loop, with the ultimate goal of attaining the highest flow rate possible. Like any phase system, the less flow you get, due to increasing resistance (or captube), the closer the pump will get to its maximum pressure differential. In a normal loop, the pressure differential is nowhere near the pump's max head pressure. Once again, it depends on resistance throughout the loop, which is directly effected by flowrate

[warriorpoet]

as opposed to...
An open loop where flow is variable.

No, I am not referring to maximum head pressure.

[Splave]

now we are talking space talk, eligray you sound like a pretty intellegent guy for sure :) Im going to try with the pump right before the block and will post my results. I bet it makes little difference either way.

Thanks for input guys will post when I have some results

[warriorpoet]

Bobby,

The biggest thing you can do for your temps is change your fans. While the Slipstreams may seem liek they push a lot of air, they lack badly in static pressure. They're really designed to be free-air case fans. This is especially important if you're using a Black Ice (e)Xtreme rad.

Also, while we can take educated stabs at what's causing your poor performance given a list of components (the BI rads are notorious flow killers, for instance), good pictures tend to be much more helpful.

What are your ambient temperatures? What are you using for coolant? Any sharp bends? Did you take your GTZ apart, and if so, are you sure you put it back together EXACTLY like it's supposed to be? How many mounts have you tried? All of these are things which could be affecting your temps.

[hellcamino]

I have to admit I am curious still on the flow vs pressure debate, using a pump with higher head pressure means nothing if you aren't adding enough restrictions to the loop to lower the flow rate and increase the actual fluid pressure.

If excess of 1 gpm yields no measureable gain and typical use runs at approx. 1.5gpm or higher it would seem to me to be merely a matter of choice or space considerations that dictated which pump you choose.

If someone can show me where it makes a difference in a given scenario where you are pushing 1.5gpm+ through a given block with different pumps that have differing capabilities in terms of max head pressure I'll be impressed and grateful. I am not asking for flow and pressure specs here but for real world actual temperature variation, IMO at this time as long as flow rates are equal pressure means nothing in the ranges we are dealing with.

[John Planet]

Do you have any information on as to why this might be? I'd be really interested to hear it :)

I'd like to see that tool. If flow is is based on total restriction, and constant, why wouldn't pressure be limited by the same factors and constant as well?

[eligray]

I'd like to see that tool. If flow is is based on total restriction, and constant, why wouldn't pressure be limited by the same factors and constant as well?

what tool?... First off, I still am not totally sure of what you mean by a closed vs open loop. The ability of coolant level in the T-line to fluctuate to keep pressure constant?

As for what you said, I'm not entirely sure what you are asking here. A pump will always be trying to reach a pressure differential between its two sides equal to its max head pressure. However, the fact that water is allowed to flow from one side to the other tries to reduce this deferential. If a loop has high restriction, then the pressure will be higher than the loop average at the loop average after the pump, and lower than the loop average before the pump. But of course, as I said, the T-line keeps whatever area it is attached to at the pressure level equal to the height of coolant in the T. I'm just guessing at what you were asking here, could clarify?

If average loop pressure really does matter, then a 10-foot tall T-line would help. Is this what you're saying?

Hope yours temps drop, and thanks for the compliment :)

Eli

[John Planet]

(typo) I meant to say "too." :eek:

[eligray]

As for the restriction and head pressure comment above, Here's a quick equation I'm making up off the top of my head:

(Flow)=(PeakGPM)*[(Maxhead)-(Restriction)]/(Maxhead)

Making my head hurt :P


(typo) I meant to say "too." :eek:

:ROTF:

PS: Have I mentioned I'm 16? :p: