http://www.customcooledpc.com/maze4/001.jpg
http://www.customcooledpc.com/maze4/002.jpg
http://www.customcooledpc.com/maze4/003.jpg
http://www.customcooledpc.com/maze4/004.jpg
http://www.customcooledpc.com/maze4/005.jpg
http://www.customcooledpc.com/maze4/006.jpg
Note the O-ring and the allen wrench were extra and I chose the metal mounting hardware option over the nylon. I don't like this set of mounting hardware much. I would rather have screws with a pan head instead of having to use a nut on the backside. Also the nylon washers and springs are to big and the springs are not stout enough. I had to completely compress the springs to get a acceptable (to me anyway) amount of pressure. Just as well not use them at all. I never liked using springs anyway.

More pics and results later, hopefully tomorrow afternoon....

Here is my first set of numbers:

XP1700+@1800mhz
1.75V
Abit KD7

Thermocoupler under CPU with socket mod: 43C
Onboard from Windbond software that came with board: 40C
Room temp: 26C
Water temp: 30C

This is under a Distributed Folding Load.

Going to load up CPU Burn now and see what it will do.

It would be a better design If It had a center input and two smaller outputs, To Me at least.:evilaugh: :evilaugh: Nice pics though.:D My cpu has a Temp of 105F (41C according to the LED readout) right now and the motherboard Temp is 93F. I know water cooling can with or without a Tec make the CPU cooler though.:rolleyes:

Your temps mean absolutly nothing. Sorry to break it to you. ;)

Well whoppee due. They do to Me and the Bios agrees with MBM, So there. At least I have heatsink mounting holes on ALL 3 major places on the motherboard (CPU, NB & SB) . But then Epox is known for It's overclocking ability and what Do You have? An Asus or an Abit? Abit isn't too bad, Asus I wouldn't touch with an Anti-Matter 10' pole. If I did neither the Asus nor the pole or Maybe Me would exist of course.:evilaugh: :evilaugh: :evilaugh: :evilaugh: :evilaugh: :p :p :p :p :p

What the hell are you talking about? Your temps are meaningless. The onboard probe is garbage and is also thrown off by all that CFM your pushing around the socket. No shit it is going to read the same in the bios, that means NOTHING. The onboard probe is simply not accurate especialy if there is ANY air flow around the thermisters. have you read anything on this site?

Do you even know what the thermocoupler I am using is? Has nothing to do with the onboard probes.

You can post your BS temps around but your not going to get away with it on THIS site.


http://forums.procooling.com/vbb/sho...&threadid=6943
http://forums.procooling.com/vbb/sho...&threadid=6949
http://www.procooling.com/articles/h...hermal_dio.php
http://www.procooling.com/articles/h...sting_cali.php
http://forums.procooling.com/vbb/sho...&threadid=5765
http://forums.procooling.com/vbb/sho...&threadid=6442

fyi- the Bios and MBM both use the save variable to bring you the exact number from the same equipment.

Well When I get an Enermax unit like this,
http://www.plycon.com/fans/enermax_35_blk_a.jpg
I'll hook up My custom made probe and
We shall see what the difference is then, Ok? We'll see what an Epox can do with the Gemin Cool.

LOL, thats not going to be much better than what your using now. Read the damn links I posted. You might learn something. ;) Use at very least a type K thermocoupler (the worst type, which I am using) and preferable a type J or what PH has done in the links posted.

The 8RDA DOES NOT produce accurate temp readings :rolleyes:

Forgot to add I removed the worthless spings and crappy mounting hardware from DD and used my own hardware and no springs.
http://www.customcooledpc.com/maze4/007.jpg
-----
Thermocoupler wires running under the CPU with the socket mod and the thermocoupler is attached to the bottom of the CPU.
http://www.customcooledpc.com/maze4/008.jpg
----
http://www.customcooledpc.com/maze4/009.jpg
The thermocoupler reader. It still isn't perfect but it is better then the onboard.

I ran CPU Burn for an hour and temps showed .5C warmer on the onboard and stayed the same on the TC. The TC only reads whole numbers so it probably did go up .5C on it to.

Also note these temps are with the HydroThruster 500GPH pump at full speed. Don't have a GPH measurment at this time....

Jaydee,

for those who aren't fortunate/or smart enough to have a thermocoupler, are there any other commercial WB's you can compare those maze4 results with or just with your homebrew's?

thnks

I am interested in this as well. I am waiting to get my Maze 4 before I even start any testing with the Maze 3, that way I can see a side by side comparison of both.

My best friend has the EXACT same components as myself, cpu, mobo, ram, watercooling accesories. Only difference will be the Maze 4 vs the Maze 3. We will set it up using the same water mixture, in the same room with same temp, etc. to see what differences, if any, there are.

Will let you know the results as soon as I get the Maze 4 block tomorrow.

No, I don't have any other commercial blocks and don't plan on getting any. I just bought this Maze 4 to compare my own blocks to. There should be some better "official" reviews of this stuff sooner or later I would think. Maybe pH will have some of these blocks when he starts testing.

P.S. the thermocoupler can be had for $17.00 at Wallmart.

You keep mentioning this thermocoupler and socket mod??? Sorry, new to all this, but what is this socket mod?

Are you referring to some way to get the thermocoupler inside to read accurate tempatures without damaging the socket?

Damn good read. Thanks!

Okay, the three PCs are as follows with various tests. The room was air conditioned to 75F (outdoor thermometer). The following are LOAD temps with all three PCs doing full load runs, using the onboard thermistors. Yeah, I know these suck ass, but for comparison's sake they will show a temperature difference, okay? Don't make me teach a course on relativity, guys! Anyway, as you all know, these are all on one loop pushing approximately 7-8GPM/1000CFM with a common coolant store. Here we go:

#1: Athlon 700@770, Black Edge (34C)
#2: Athlon XP 2100+@2550, 1.70V, Black Edge (48C)
#3: Athlon XP 2500+@2650, 1.80V, Maze3 (46C)

#1: Black Edge (33C)
#2: Black Edge (46C)
#3: Maze4 (52C)
Maze4 more restrictive?

#1: Black Edge (35C)
#2: Maze3 (44C)
#3: Maze4 (53C)
Maze3 less restrictive? #1 is odd here.

#1: Black Edge (34C)
#2: Maze3: (45C)
#3: Black Edge (50C)
The Black Edge does better?

Currently I have the setup at BE/M3/M4, and I'm not happy with the performance at all. I've had to back my speed on the 2500+ down to 2550 Mhz and down .25V because of the temps and sudden instability due to it. I would NOT recommend this block over the Maze3 unless they redesign it. It appears that they were trying to shave costs off of the manufacturing end to make more profit at retail. By trying to pass off an inferior product at the same price by telling you it'll increase performance ... well, that pisses me off. The Black Edge block does better in my tests and it is a 3/8" block that costs around $25 new. I wouldn't buy one of these Maze4s for any reason from what I learned.

I am not understanding here. Maybe I am not seeing the numbers right, or maybe my logic is flawed.....but wouldn't you want to run all 3 blocks on all 3 different comps to run a comparison test?

Maybe it is just me but, you say the black edge performs better, but you never tested the Maze blocks on comp 1 at all, and the Maze 3 outperforms the Black edge on comps 2 and 3. So how does the Black edge perform better?

I can see where the Maze 4 does in comparison on comp 3, but you never tried it on comps 1 or 2.

No offense, but this isn't exactly what I would call a conclusive test. 8P

Well you can extrapolate a rough guess from what he posted.

On system 2 the Maze3 performs 1-2 degrees better than the Black Edge. On system 3 the Maze4 performs 2-3 degrees worse than the Black Edge. Hence, the Maze 3 appears to be the best, the Black Edge comes in second, and the Maze 4 comes in last.

From a designer's perspective, the lack of a center inlet, has got to really hurt the performance. This Maze4 is actually a step back, as it's now down to a simple side to side flow configuration.

The side to side flow configuration only has one advantage: it's the least restrictive. The problem is that the added flow can't make up for the simplicity of the design.

It reminds me a lot of that French block, who posted that again?
http://forums.procooling.com/vbb/sho...&threadid=5731

That french design is actually superior though, because of the added turbulators, and all the tweaking added to it.

I have to agree: there are better blocks out there than this Maze4, for the price. Better luck next time, DangerDen!

I don't know I am not having near the dissapointment airspirit is. 13C over water temp under CPU Burn and DF load isn't to bad at all. :shrug:

I just re-read airspirits post and caught onto the fact he is using a 2500+@ 2650mhz at 1.80V. :eek: I am only running 1.75Vcore at 1800mhz on one of the cooler Tbred A 1700+'s. I have a theory that there isn't enough copper on the Maze 4 to handle the higher load he is using especially being there is no center inlet. The Maze 4 has much more copper and possible a thicker base as does the Black Edge. When I get home tonight I will cranck up the Vcore and OC as high as I can get and then re-post the new numbers. My board will go up to 2.35Vcore so I should have a little success.

Again, part of my problem could be due to the parallel configuration between the PCs and the load balancing that causes with the water flow. I would guess from my tests that the M4 is the most restrictive of the blocks (more so than the BE), and that causes it to perform bad. Even if this is the case, however, I'm pumping at least 2GPM through it since my #1/#2 loops are 3/8" and my #3 loop is 1/2" ID. This could be completely due to the system I'm using, but even if this is the case it would suggest performance risks for different users. Why take a risk if you don't have to?

Actually, 1/4" is appropriate, for this design, IMO.

I think that if anything, you could get some business out of modding these: add a center inlet, try to jet that inlet a bit on either side of the center fin, and that should yield a substantial improvement. You might even be able to shave a mm or two, off of the baseplate. Alternatively, you could drill a cup, ~1 or 2 mm deep, >5mm wide on either side of those fins, which would be a heck of a lot easier to do.