Innovatek Block
 

I have the Innovatek block setup and running. I replaced the fittings with 1/4x1/2 nylon hose barbs. Under load, Ambient 16C, Water 16.5C, CompuNurse reads 25.5C, and KR7A on-board probe reads 31.5C. CPU is AXP 2000+ with Vcore=2.08V @1921MHZ.

Overall it's a really cool block, and I love the mounting mechanism. It's really easy to use, and distributes the pressure nicely over the core.


InnovaTek After 12 Hours
Water: 20C
Ambient: 18C
Compunurse: 29.6C load/26.1C idle
On board: 35/29C idle
A delta of about 9.6C at prime95 load

MAZE2, After 24 hours
Water: 21.3C
Ambient: 19C
CompuNurse: 30.7C load/27.6C idle
On Board: 33C load/29C idle
delta of 9.4C at prime95 load

Well, I've updated these numbers a few times. I don't know how much weight to give them. First, I reseated the MAZE2 a couple times before I got those temps. I may have reduced the performance of the Innovatek a little bit because I screwed the barbs down about 1cm farther than the original fittings. Also, I'm not sure how accurate my CompuNurse readings are. I'm afraid that the smaller footprint of the IT block produced lower temps because more air is getting to the probe. The flow rate seemed lower with the IT block, even though the MAZE2 had 3/8" fittings and the IT had 1/2" fittings. I noticed that the IT block has a tremendous heat capacity. The temps really rose slowly when I turned on the computer.

I guess my conclusions are that the IT block and MAZE2 block perform about the same. It's much easier to seat the IT block correctly and doesn't require use of mounting screws. The results are so close that I'm having trouble deciding which block to run, long term. I'm also convinced that the Compunurse probe isn't accurate enough for the tests. I think I have to wire up the in-core temp diode for a real test.

So who is selling the Innovatek block or kit?

I bought mine from www.highspeedpc.com

What kind of pump did you use with it?

My Setup:

Danner Magdrive-2 (rated 350GPH)
3/8" silicone tubing
Leviathon radiator with 1/2" fittings, 2x120mm Panaflo L1A's
Delux rubbermaid reservoir 1gal, 1/2" fittings.
Kr7A mobo
AXP2000+ @174x11 (1.921GHZ) Vcore =2.08V
KingMax DDR333 Vmem=3.04V

Which one of the Innovatek blocks did you use then?

I used the InnovaCool block

Exactly where and how do you have the CompuNurse connected Schoolie?

From looking at the pics of the block at review sights, the fittings being inserted farther really shouldn't make much difference the way the water jacket is done around the copper slug.

I stripped off the protective plastic at the end of the thermistor and the heat shrink on the leads. Then I taped the probe so that the tip was touching the corner of the CPU core. I put a dap of ASIII on the end of the Compunurse probe too.


Did I do it correctly?

so the temps are about the same then.....

I haven't seen a pic of the inside yet, it has been described to me a couple of times...

Yes, I would say the temps are about the same. I'm going to measure the MAZE2 again tomorrow.


This is a little peek of the inside:

http://www.pcabusers.com/reviews/wat.../image006a.jpg

so it is how I presumed it was, cool. Now for some heatpipe action

There's one thing I don't get. You're running the IT block with 1/2" barbs yet your tubing stays 3/8"? Is there a mistake here or did I miss something?

Other than that, the IT block seems really cool. I had thought of going with Maze-II, but now I don't know. Damn you Schoolie! :)

Coffee and posts in the morning
 

It'll work. Here's my blurb about attaching thermocouples:

Make sure that there is no contact between the two leads where you removed the heat shrink (bare or nicked insulation).

I position the probe so it is pushed against the side of the die (like running a cable at the base of a wall) with the probe tip at the center of the die's side for the most accurate, repeatable readings.

If you use ASIII to eliminate air pockets, be sure to inspect the die in the contact area to insure there is no exposed copper from cracking, etc. Just as with block mounting, the thinner the compund between the probe and die the better (excess on the outside of the probe is fine.) Be sure to press it tight against the die with the tape. Another way to do it that works well (although cleanup a little harder) is to use the silver conductive liquids used for unlocking. Just dip the tip of the probe in it and then apply it to the side of the die, holding it in place for a few minutes until it is mostly dry (use a plastic bag over your finger). Then cover it with tape.

If you have acess to it, either metal tape or kapton tape (yellowish in color) works best for attachment. The kapton is much easier to work with. If you use metal tape, be careful as the adhesives used tend to be very strong. Once it touches a surface, it is pretty much there, unlike kapton tape which can be easily pulled up and moved (use alcohol or finger nail polish remover and soak the metal tape before removing it).

A thin bead of silicone on the outside of the tape will provide a good barrier against air flow altering the readings as well as help hold the probe firmly against the side.

Whatever way you do it, just be sure to do it the same way for the two blocks.

Another pic of the Innovatek
 

Oh, and here is another pic of the Innovatek construction (just ignore where it's from ;)

Innovatek taken apart

Combine with the other pic Schoolie posted, you can see as long as he didn't put the barb so far in that it gets too close to the bottom of the water jacket cavity, it shouldn't affect performance that much, at least on the inlet side. (but the lower flow rate Schoolie noticed might be because he got it too far in, shrinking the distance between the barb outlet and the floor of the cavity too much)

Is this block suited for peliters? I´m thinking of getting blocks for my Cellery comp but I want the ability to peltier cool them later. Should I go with the Maze1C instead then (it´s a Abit BP6 housing dual Celleries)?

anyone archive the image ??

Innovatek taken apart

THG took it down

hozie: 1/2" fittings will improve the flow, even with 3/8" tubing because the fittings are the most restrictive elements in the loop. Also provides a nice, tight seal.

hedgren: Definitely can't use Peltier device. That's one strike against the IT block.

EMC2: Thanks! Good intructions on mounting the probe.

I was hoping to do more testing this morning, but work intruded again. Once caveat, the mounting device for the IT block doesn't have much clearance. I had to sand down the collar on the 1/2" hose barbs before the mounting hardware would fit.

To bad about the peltiers. It would be so good with that smart mounting system and all. Oh well, I wont be watercooling the Celleries soon so I´m not in a hurry exactly. Maybe Fixxit comes with some new smart block (massproduced ;) ) that don´t require the 4 mounting holes.

How typical that it as just that picture that was removed then :mad:

The attached image was produced by THG, all credits, yadda, yadda, yadda.

many thanks, EMC2

any idea what is ment by:
"The copper core is inseparably bonded with the aluminum housing using a special thermal procedure and absolute water tightness is guaranteed."

shrink fit ?

I want to cross drill it (#53 drill, 2mm OC from face - AMD spec) for a TC hole
looks like there should be plenty of room, leaving an undisturbed seal area above the hole

Nice picture. Thanks. It looks like they could have made the notches in the core deeper and doubled the surface area.

unregistered - it would be just a guess, but from looking at it disassembled and what they say, my hunch would be they freeze the copper core to get it to shrink and then press fit it into the body. I don't think you could safely drill into that base section as a result, since the seal in that case would actually be made partially by the sides of the "foot" (bottom lip of the copper core).

Schoolie- there are several negatives from a thermal transfer standpoint with the way they did those ridges.