Modding a Maze3?...

MadDogMe · 10-22-2002, 10:26 AM

Kudos to DangerDen on being up there with the top performing blocks,they certainly work well enough.
Having said that the M3 is an awful design now I look at it!, it's size is good for a TEC, but for a CPU core?...

I noticed after my first real good look that there are in affect two channels that run parallel to each other over the core area, then merge into the long one that runs around the edge/outside of the block, these two channel are in fact halving flow across the vital area!

, then when it merges it is restricting flow in an area where there's no need/use, at the furthest point from the core. What can I do to improve this that does'nt involve totaly demolishing the block?. something that can be done without milling...

Turn it into a high flow block?, move the inlet (use 5/8 in/outlets) to the far opposit side from the outlet and 'do away' with the 'vertical' parts of the maze, so in affect it is an 'open' block with three horizontal fins?. maybe lower the height as well?, halve it to improve flowrate/velocity through the block and dimple the BP with a drill (ala Swiftech) for turbulence and to thin the BP thickness down. How would I work out optimum area to leave?, does anyone know the BasePlate thickness?...

Lower the level of the two channels over the core?, use silicone sealant to lower the Lexan top over the channels to improve flow rate through them?. would this yield any difference do you think?...

The last idea I can't explain without a drawing, but it'd involve having the outlet on the opposit side, in the area to the left of the inlet, I'd open up another 'hole' symetrical to the one there already is, to make the flow symetrical, I'd put a sort of < where the outlet is now to help direct the flow, I'd need to remove a metal from below where the outlet stops as well...

Which of these ideas do you think would yeild the most?, I'm putting my money on the 'open' block, with either an uprated pump or lowered, preferably lowered...

Blackeagle · 10-26-2002, 09:18 PM

I must be going crazy! (I have been told at work that this is likely.) I've never built a rig, but have been casting a eye at that same block with regards to "improvements". My thought was to dimple the base of the channels as you said, but also to take a dremmel and cut shallow groves at a angle on all walls. Inner walls would have the grooves running from bottom and angled upward (in relation to flow) and outside walls the grooves would run from top down to base of the channels. The idea would be to try and instill a twist to the flow going through the channels as well as gain surface area by both the grooves and the dimples. I hadn't thought of lowering the height of the channels to increase the velocity of the flow however, good idea.

I'm afraid I got lost in your description of the other alterations involving the moveing of the inlet.

But it would be interesting to "improve" a maze 3 with the other 3 changes and then compare it to the performance of the maze prior to being improved.

I don't have a rig built yet so I can't do it. Would you please let me/us know how it turns out?

#Rotor · 10-26-2002, 09:52 PM

one thing I would have done, is to cut some slots into that fin in the middle , more surface area and more turbulence, I would also half the effective channel height... it's way too deep.... the top of the channel probably never see any heat at all, so in essence it's just an easy bypass for a lot of lazy water....

but I never even got close to one of those, so I'm just imagining things, really

Since87 · 10-26-2002, 10:12 PM

So MadDog, been reading oc-forums?

I posted the idea of filling in the top of the central area in the Maze3 two days before you posted this. It's in this thread.

LiquidRulez · 10-26-2002, 11:26 PM

Quote:

Originally posted by Since87
So MadDog, been reading oc-forums?

I posted the idea of filling in the top of the central area in the Maze3 two days before you posted this. It's in this thread.

Who cares who posted what first!

If you do all the cryin'

, what the h&ll are the babies going to do???:shrug:

JUst kidding.......dont go postal on me!

Just make em' and let us know the results.
After all, thats all we're interested in anyway.....RESULTS

MadDogMe · 10-28-2002, 03:59 AM

What's oc?, overclockers?.

No it was a spontanious thing, I'm not into plaugarism

. I have noticed before though that things seem to 'come in three's', bet someone else, somewhere else, has been pondering it too, if I'd seen your thread I'd have posted a link or joined in with you. I don't read the OC cooling forum much though cause it's hard to find the gems in all the dross about aircooling and such

...

It is a waste of space though is'nt it?, Since reading DodgeVipers thread about using springs as turbulators I've been thinking about putting something like that in there, what do you think?, we could get away with using half the volumetric area!!. it'd be easy just to lower the top by fixing something to the lexan, but we might as well go the whole hog eh?...

I thought about cutting grooves or just widening the outer channel, to try and balance it to the two inner ones, but I don't think there's enough room

...

I'd like to insert Cu pins into the baseplate in the 2 channeled centre as well!, or solder some extra fins as a last resort. can you buy Cu in rods?, can you get pure Cu brazing rods?...

***EDIT**** I don't believe it!, just went and had a look~see and there's a bit about puttoing in 'Cu coils' and a quote about DodgeViper too

, you are now entering the twighlight zone...
Wonder what else I'll find?...

MadDogMe · 10-28-2002, 04:22 AM

Well there you go!, I did'nt even know there was a H2o section...

Since87 · 10-28-2002, 07:14 PM

OK, NP.

When you make your living in engineering someone else taking credit for your idea is a serious offense. I've had it happen enough to make me pretty touchy. Sorry about that.

As I mentioned at oc-forums I'm planning on molding a piece of epoxy in the Maze3 itself. Got to figure out how to get it out though. One suggestion I've gotten, is to dam the areas I don't want filled, and coat the remaining space in vaseline. I'd be happy to hear other suggestions though.

MadDogMe · 10-29-2002, 03:47 AM

I've had simular happen to me lately, it's reasuring this happening in a way

...

I'd fix it to the inside of the lexan top myself, that way it's not obscuring any vital Cu surface area!, you could fashion a couple bits of plexiglass to shape then glue them to the lexan, use a glue you can get off till you're happy with the results. Or fill the channels with silicone sealant, let it cure then pull it out, then cut it to shape, stick it back to the lexan with more sealant. at least it's reversable and does'nt mark the lexan...

What do you think about some heavy duty turbulators?, copper coils like Dodge Viper?. we can afford to make them quite restrictive as we want, I'd like to widen the outer channel as well, make it a higher flowing block, restrictive is only good if used wisely...

10-22-2002, 10:26 AM	#1
MadDogMe Thermophile Join Date: Aug 2002 Location: Just shut up ;) ... Posts: 1,068	Modding a Maze3?... Kudos to DangerDen on being up there with the top performing blocks,they certainly work well enough. Having said that the M3 is an awful design now I look at it!, it's size is good for a TEC, but for a CPU core?... I noticed after my first real good look that there are in affect two channels that run parallel to each other over the core area, then merge into the long one that runs around the edge/outside of the block, these two channel are in fact halving flow across the vital area! , then when it merges it is restricting flow in an area where there's no need/use, at the furthest point from the core. What can I do to improve this that does'nt involve totaly demolishing the block?. something that can be done without milling... Turn it into a high flow block?, move the inlet (use 5/8 in/outlets) to the far opposit side from the outlet and 'do away' with the 'vertical' parts of the maze, so in affect it is an 'open' block with three horizontal fins?. maybe lower the height as well?, halve it to improve flowrate/velocity through the block and dimple the BP with a drill (ala Swiftech) for turbulence and to thin the BP thickness down. How would I work out optimum area to leave?, does anyone know the BasePlate thickness?... Lower the level of the two channels over the core?, use silicone sealant to lower the Lexan top over the channels to improve flow rate through them?. would this yield any difference do you think?... The last idea I can't explain without a drawing, but it'd involve having the outlet on the opposit side, in the area to the left of the inlet, I'd open up another 'hole' symetrical to the one there already is, to make the flow symetrical, I'd put a sort of < where the outlet is now to help direct the flow, I'd need to remove a metal from below where the outlet stops as well... Which of these ideas do you think would yeild the most?, I'm putting my money on the 'open' block, with either an uprated pump or lowered, preferably lowered...

10-26-2002, 09:52 PM	#3
#Rotor Cooling Savant Join Date: Feb 2002 Location: Dione, sector 4s1256 Posts: 852	one thing I would have done, is to cut some slots into that fin in the middle , more surface area and more turbulence, I would also half the effective channel height... it's way too deep.... the top of the channel probably never see any heat at all, so in essence it's just an easy bypass for a lot of lazy water.... but I never even got close to one of those, so I'm just imagining things, really __________________ There is no Spoon....

10-28-2002, 03:59 AM	#6
MadDogMe Thermophile Join Date: Aug 2002 Location: Just shut up ;) ... Posts: 1,068	What's oc?, overclockers?. No it was a spontanious thing, I'm not into plaugarism . I have noticed before though that things seem to 'come in three's', bet someone else, somewhere else, has been pondering it too, if I'd seen your thread I'd have posted a link or joined in with you. I don't read the OC cooling forum much though cause it's hard to find the gems in all the dross about aircooling and such ... It is a waste of space though is'nt it?, Since reading DodgeVipers thread about using springs as turbulators I've been thinking about putting something like that in there, what do you think?, we could get away with using half the volumetric area!!. it'd be easy just to lower the top by fixing something to the lexan, but we might as well go the whole hog eh?... I thought about cutting grooves or just widening the outer channel, to try and balance it to the two inner ones, but I don't think there's enough room ... I'd like to insert Cu pins into the baseplate in the 2 channeled centre as well!, or solder some extra fins as a last resort. can you buy Cu in rods?, can you get pure Cu brazing rods?... *EDIT** I don't believe it!, just went and had a look~see and there's a bit about puttoing in 'Cu coils' and a quote about DodgeViper too , you are now entering the twighlight zone... Wonder what else I'll find?... Last edited by MadDogMe; 10-28-2002 at 04:04 AM.

10-26-2002, 09:18 PM	#2
Blackeagle Thermophile Join Date: Oct 2002 Location: U.S.A = Michigan Posts: 1,243	I must be going crazy! (I have been told at work that this is likely.) I've never built a rig, but have been casting a eye at that same block with regards to "improvements". My thought was to dimple the base of the channels as you said, but also to take a dremmel and cut shallow groves at a angle on all walls. Inner walls would have the grooves running from bottom and angled upward (in relation to flow) and outside walls the grooves would run from top down to base of the channels. The idea would be to try and instill a twist to the flow going through the channels as well as gain surface area by both the grooves and the dimples. I hadn't thought of lowering the height of the channels to increase the velocity of the flow however, good idea. I'm afraid I got lost in your description of the other alterations involving the moveing of the inlet. But it would be interesting to "improve" a maze 3 with the other 3 changes and then compare it to the performance of the maze prior to being improved. I don't have a rig built yet so I can't do it. Would you please let me/us know how it turns out?

10-26-2002, 10:12 PM	#4
Since87 Pro/Guru - Uber Mod Join Date: Sep 2002 Location: Indiana Posts: 834	So MadDog, been reading oc-forums? I posted the idea of filling in the top of the central area in the Maze3 two days before you posted this. It's in this thread.

10-28-2002, 04:22 AM	#7
MadDogMe Thermophile Join Date: Aug 2002 Location: Just shut up ;) ... Posts: 1,068	Well there you go!, I did'nt even know there was a H2o section...

10-28-2002, 07:14 PM	#8
Since87 Pro/Guru - Uber Mod Join Date: Sep 2002 Location: Indiana Posts: 834	OK, NP. When you make your living in engineering someone else taking credit for your idea is a serious offense. I've had it happen enough to make me pretty touchy. Sorry about that. As I mentioned at oc-forums I'm planning on molding a piece of epoxy in the Maze3 itself. Got to figure out how to get it out though. One suggestion I've gotten, is to dam the areas I don't want filled, and coat the remaining space in vaseline. I'd be happy to hear other suggestions though.

10-29-2002, 03:47 AM	#9
MadDogMe Thermophile Join Date: Aug 2002 Location: Just shut up ;) ... Posts: 1,068	I've had simular happen to me lately, it's reasuring this happening in a way ... I'd fix it to the inside of the lexan top myself, that way it's not obscuring any vital Cu surface area!, you could fashion a couple bits of plexiglass to shape then glue them to the lexan, use a glue you can get off till you're happy with the results. Or fill the channels with silicone sealant, let it cure then pull it out, then cut it to shape, stick it back to the lexan with more sealant. at least it's reversable and does'nt mark the lexan... What do you think about some heavy duty turbulators?, copper coils like Dodge Viper?. we can afford to make them quite restrictive as we want, I'd like to widen the outer channel as well, make it a higher flowing block, restrictive is only good if used wisely...

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)