Maze3 vs. Maze2 Guess who lost: ME!

[Jim]

Well I put the Maze3 in the box today. After reading the results on the other thread I knew I wasn't going to gain much.

All I changed was the water block. The hoses are in a slightly different location, but nothing significant.

This morning I ran Sandra Burn in for one hour with the existing Maze2 installation.

Results:
Before starting idle temps were 34-35C.

Ambient: 72f
CPU: 37C (highest temp reached)
Case: 27C

This evening after installing Maze3 and running pump only for 3 hours bleeding, refilling etc.

Before starting idle temp were 35-36C. Ut OH!

Ambient: 72F
CPU: 37C (highest temp reached)
Case: 27C

So that’s it. I am not a bit surprised. And I had to spend an additional $5.00 or so to ship back the first defective Maze3 I received with cracks in the plastic top.

However, I have known temperatures to drop in a day or two or three. The AS3 has to settle in and there most likely is still air in the heater core.

But even if I gain one or two degrees C, who gives a crap? So I spent about $60.00 to gain nothing, and have an inferior product installed in my computer that I feel is not a durable as the Maze2.

I did forget to point out that the bolt on the upper left is easier to torque down than with the Maze2 since I use 1/2" tubing. In addition I can see the water passing through the block. Oh boy!

Well that's it, I was going to sell the Maze3 and not waste my time installing it, but I have been know to make a bad call now and then. Oh, if you get one of these lap it, mine was low in the center. Lapped it with 250 up to 400 only, same as Maze2.

I will update if temps drops, or goes up after the AS3 settles in and the air is all gone.
Later,
Jim

Stuff>>>
1.33 Tbird @ 1468
Eheim 1250
1/2" stinky silicon tubing distilled water + water wetter
Maze3 Humm... :shrug:
Ford van heater core with 2 120 68 cfm Panalfo's @ 7 volts

[Fixittt]

Jim before you totally give up the ship, I would recomend putting a valve in ne of the tubings. So that you can adjust the flow rate. And try tweaking the flow. Maybe the maze 3 might shine with a lowered flow, to equal the maze 2 flow restricted design.

Just a thought, I knew we would not be seeing massive gains in temp reduction with any of the new waterblocks. We have pretty muchreached out limit with copper.

[Jim]

Fixitt-

Well, I didn't give up totally but you are correct about one thing in that the limits are here. I still have to wait a few days to see how things settle in. I have seen time improve temps with air and water cooling as well.

Valve, a plastic valve with a 1/2" barb screwed in each end? Home Depot, Menards?

Jim

[Cyco-Dude]

i dont think Home Depot or Lowes has fittings that small (and if they do, they'ld probably be very bulky), but you can look. you might also try browsing thru a local aquarium or pond shop...

if you find anything, let us (me) know

edit*
damn, im freon-cooled now? i havent even got my watercooling stuff set up yet. guess i gotta send it all back and make a phase-change system now

[ECUPirate]

Mine was low in the center too, which made it a real bitch to lap. I'd rather it be high in the center, and then flatten it out until the flat area was sufficient to cover the core. Even lapping it all the way, it would still take less time... but that's a mute point.

[Sherlock]

Originally posted by Jim
This evening after installing Maze3 and running pump only for 3 hours bleeding, refilling etc.

This morning I ran Sandra Burn in for one hour with the existing Maze2 installation.

Results:
Before starting idle temps were 34-35C.

Ambient: 72f
CPU: 37C (highest temp reached)
Case: 27C

This evening after installing Maze3 and running pump only for 3 hours bleeding, refilling etc.

Before starting idle temp were 35-36C. Ut OH!

Ambient: 72F
CPU: 37C (highest temp reached)
Case: 27C
stuff>>>
1.33 Tbird @ 1468
Eheim 1250
1/2" stinky silicon tubing distilled water + water wetter
Maze3 Humm... :shrug:
Ford van heater core with 2 120 68 cfm Panalfo's @ 7 volts [/b]

So you're saying that during 1 hour of sandra burn (100% cpu usage) the maze3 went from aound 36C Idle to 37c full load????

[ablaze]

cheer up Jim..as with alot of the stuff we do..the fun is in the TRYING..not the end result right?

[Brad]

sorry to hear that jim, maybe you should try remounting the block

[Jim]

Originally posted by Brad
sorry to hear that jim, maybe you should try remounting the block

Brad I think I agree.
last night I shut off the computer and let the pump only run, to get more air out and check for leaks.

This morning I powered up, and the case temp read 20c (using MBM) but IMMEDIATELY the CPU was at 33C! That ain't right. The CPU at start up should have read around 28-29 and then inched up as the case temps rose. I think I may not have enough AS3, I went very very thin this time.

My mount is tight, can't swivel the HS at all. SO after work today I think I'll pull it check the print on the block and remount. Clean the core off and apply the AS3 fresh and clean the block smear on some AS3 and wipe.

What do you guys think?
Jim

[UpAllNight]

I believe thats what i would try first Jim...by your temp readings it sure did jump up there real quick....which points to a bad seat on your WB

[Cyco-Dude]

its a start. as for the thermal paste, ive heard that a very thin coat is not as good as a thicker coat (you shouldnt be able to see the core after you've applied the thermal paste, just dont have a super think coat as thats not good either lol). i just put a small dab of paste on there, wrap my finger in plastic wrap, and spread it around like that.

[Jim]

Going to reseat and recoat with AS3 now.
Back later, if she doesn't blow up!
Jim

[ondaedg]

Jim, through hundreds of reapplications of thermal paste, I have found that using a good amount always equals or betters using a super thin layer. Reason being is that your block may be lapped, but these AMD/Intel chips sometimes have a curved surface. The last Athlon I had, you HAD to apply a good amount of thermal paste to fill in the valley in the center due to its concave (or is convex the word I am looking for?) surface. The pressure by the mounting system should squeeze out the extra paste. Trust me on this issue.

Also, if your temps don't improve that much, don't fret. The Maze 3s look sweet as hell in comparison to the Maze2s.

[MeltMan]

Its concave (dished). Convex is raised.

I had the same thing with my athlon. The duron I have now is ok tho.

[Brad]

all of my 5 DDen blocks are concave, I just think they all are hehe, must be how they do their milling or something. And every one else seems to find the same thing

[GeminiCool]

Just a quick note... the concave nature of the copper blocks is a function of the manufacturing / operator. Infact, you can reduce this effect by slowing down your feeds and speeds and NOT clamping the block so hard. Cooper is soft and you WILL bend it if you crank down on the vice holding the block.

See I guess there is a choice made here. Slow down the feed and speed and use less clamping pressure (increase your cost). Or crank it up clamp it down and let the end user rub each block. Think about it this way... if DD had to rub each one of the blocks for 15min to get them flat, your just increased your price $10.

Just an observation...

[jaydee]

Originally posted by GeminiCool
Just a quick note... the concave nature of the copper blocks is a function of the manufacturing / operator. Infact, you can reduce this effect by slowing down your feeds and speeds and NOT clamping the block so hard. Cooper is soft and you WILL bend it if you crank down on the vice holding the block.

See I guess there is a choice made here. Slow down the feed and speed and use less clamping pressure (increase your cost). Or crank it up clamp it down and let the end user rub each block. Think about it this way... if DD had to rub each one of the blocks for 15min to get them flat, your just increased your price $10.

Just an observation...

Good points.

Also heat will warp it. If they run their endmills hot the block will suck up that heat and tweak a bit especially at the thin points where the channels are. Also the last bit of copper stock I used was naturally lower in the middle from the factory. Usually has to do with the rollers they use to flaten it out into a bar. They are never perfectly flat at the factory.

[UpAllNight]

having been in the tool and die trade for 25 years i would like to add some things to this flatness of the maze 3.is it not true that the flatness only matter where it touches the core...the core is roughly .500 x .750 in size...and i have checked my maze3 with a .0001 indcator and i get it out no more than .0003 in this area.i need to check my core but i bet its out more than that...so all im trying to say is there is no need to get the whole back of the block flat...of which u will never do anyway with sand paper on a flat table....if u wanted to really get it flat u will need a laping plate...it would be a prefectly falt with groves in it in a hatch pattern...and u only use compounds to different grits just like sand paper...this takes a very good touch and alot of time....and only moving the block in a figure 8 pattern...changing grits to get the desired finish...all of this i believe is nosence with u are dealing with the small area as u are with the core size...ill check a core and see what it is actual out and reply back to this post


please if some one thinks iam totaly off base just let me know


byt the way i hope u can understand any off this i am in no means good with words...for that im sorry

[futRtrubL]

I agrre that it's only necessary to have the portion touching the core flat. However I don't see how you can flatten the center area of a concave surface without also flattening the edges.

Edward

[Cyco-Dude]

that would make sense.
nevertheless, i still lap the bottoms of all my heatsinks. ive also lapped all sides of my waterblocks down to 2000 grit (gotta have that mirror finish )

[UpAllNight]

u are right if its concave it would require lapping the whole bottom...but if u do this with sand paper on a flat surface u will in most cases put pressure on the fron edge when pushing and on the back edge when pulling making it appear to be concave...i just bet if u check one after it has been done with sand paper on a flat surface that the edges will be rooled to a certain extent...i mean checking it with a .0001 indcator...well im leaving mine alone and see what happens...if it was out bad in the part that comntacts the core i would surface grind it instead of laping it...but that the advantage of having a surface grinder...all in alll im happy with my maze 3 for what it cost...if i made one it would be a better machining job but it would cost 5 times as much to sell it for a profit

My maze3 came in pretty good shape for a production part for 42 bucks...the top was not cracked but it didnot seat completely against the copper bottom either but that is because of to much compression on the oring...i think thats the problem is the depth and the width of the oring...but i can see DD way of thinking better to have to much compression than not enough...plus just how much experence do u think the person has puting those maze 3 together and inspecting them and shiping them...u can only do so much for 42 bucks

[Jim]

Well yesterday didn't turn out good for me. :-(

I remounted my Maze3 and the system wouldn't boot. So after getting a new chip today and having the original one tested (it is dead) I am back up and running.

I picked up another 1.33 Tbird since I didn't want to spend much, and locally it costs more than Newegg, unless of course I was to get overnight shipping, then price is identical.

I was reckless when I tightened the nuts on the old chip and must have cracked it, however I can see no damage on the core or chip either!

Anyway I took my time put the Maze 3 on the new chip and the temps are about the same as before with both the Maze2 and Maze3. There is still air in the system so I want to give it a chance but the Maze 3 so far is no better than the Maze 2.
Guys have used "shims" when installing big HS's, anybody here use a shim?

So I am now a member of the chip destroyer club, not pleasant but I guess if you fiddle enough, it is going to happen. Puzzles me that I cannot see any damage or discoloring. I even put alcohol on it to expose any cracks up but nothing?

I will update with temps. Right now same as Maze2.
Jim

[UpAllNight]

oh man that scares me Jim...im about to put my system back together and im scared of those 4 bolts and not knowing how tight is tight enough....i would hate to kill my xp1800....well i will know after this weekend...im so excited i have my first water cooling rig together and in my case no leaks that i can see now for 24 hours...i have a chevette rad and a ehiem 1250 pump and a maze 3.....ill let ya'll know my temps(from 2 compu nurse)...at lest i hope i can let ya'll know...lol

I dont use a shim but maybe i should...dint want to becuse i have my xp unlocked with thermal grease and i have a probe touching my core to...i guess ill just press my luck...i have mounted many heat sinks with no problem...but i believe this w/b is a horse of a different color

[Jim]

Well-
I was hasty. I have put on water blocks 4 times before this failure. I ignored that the nuts should be put on hand tight, measured from the top of the block to a bench mark like the top of the washer on each bolt, and THEN put two turns on each nut rotating from corner to corner measuring for parallel after each sequence.

So, if you go slow and measure carefully you probably won't repeat my mistake.

I just wish the Maze3 would cool as GOOD AS my Maze2 has!
Right now one degree C higher with Maze3, of course I have a different chip so that is one more variable in the equation.
Jim

[UpAllNight]

well my temps wont do any good for anybody either beings that this is my first water cooling setup.

Thanks for the tips though but i guess im stuck with the maze 3 now to