Direct Die Cooling Project.

[jaydee]

As a couple of you may know I got banned because of a discussion that got out of hand about direct die cooling. One guy built a cooler and showed ridiculous numbers and then come to find out he had his rad in the window sucking in cold air and was testing with the onboard probes. And then tried to pass of his overclock as a means of how good the block is. As anyone here with any experience knows that is not the way to accurately test. I wouldn't have had a problem with it but he actually wrote an article about it and by some ridiculous means it got posted. Just shows the lack of quality a lot of sites have these days. When questioned they simply said decent testing is a waste of time and I am full of it blahh blahh blahh.

Well as the childish hot tempered person I am I got carried away and got myself banned for 5 days. During that 5 days my first thoughts were to hang it up. Being I seen these statements about me in a certain thread. i will not post a link as it is irrelevant where this was said. This is just one of the reasons that motivated me to do a direct die block and to return here:

What an arrogant, condescending prick!

Well the guy could just be a little miffed that some noobie to HIS forum comes along and totally blows away all his self made water blocks and anihilates his best overclock. After all he worked very hard copying all those designs. And spending copious amounts of cash in the process. And you come along with a block made out of inexpensive hose barbs and glue and make him look bad. Shame on you!!! Don't you know who he is Zenn? He is the cooling god. Beware his wrath.

Or he could just be an asshat.

heh sounds like jealousy. i ageree he just cant stand the fact that you proved him wrong.

Agreed...I do not like the fact he is intent on making this a ProForums/Pimprig fight when that is clearly not the case. It's a calm cool collected vs. asshat case. "I use type-K thermal couples because I clearly have WAY too much time on my hands and can't be trusted with a budget." I find it hilarious that he is a "forman" but can't spell it.

Yeah, they are all of $20 at WalMart.

BA, it's useless to argue with such a prick. I know the kind. No matter what you say, no matter the proof, he is going to be a **** about it. If you leave the thread, he'll think he's won, but staying will prove only futile. It's a lose/lose situation. The biggest point he is missing: Regardless of temps, Zens cooler works. It doesn't leak. It allows good overclocks, so it HAS to be cooling efficiently. Such an awesome guy as JD should know that the higher the O/C, the more heat. This guy just makes me wann buttstroke his ass.

EDIT: I would like to see somebody build this diode reader, just so we can shut his ass up.

LOL

...being the sometimes hot head that I am, I have decided to keep my posting here at home and it will be their misfortune if they don't want to see what really does work.. I have unsubscribed to that forum and I will not contribute to their post count anymore...Good advise BA.
I still think do0d is a technojaronist and an asshat with BB's for balls...And he still hasn't taken the zennzzo challenge cause the hundie is still in my hot hands

well after reading that im all up for seeing how it plays out ,still to you zenn hats off you did something that has been running for 6 months now.but this cat keeps ranting and raving about how he can build it in 30 minutes ,but then again have not seen him do it,second hell some car manufactures dont even test cars for 6 months so your record is good here.im not even going to go through the time to post on that site of all the flaws that he has in his argument all i have to say to him is that with a attutidue like that he will allways be a follower not a leader and that is where his pride got hurt .he wasnt first

First? Leufkin was SELLIING direct die blocks 2 years ago!

"I would like to see somebody build this diode reader, just so we can shut his ass up. " slider

It's really not that difficult if you want to spend the time and money to do it. I personally have a life. And would rather spend my money on other things. When all is said and done Slider. Any test of cpu to cpu temp will always be apples to oranges. IMHO its a complete waste of time and money trying for better accuracy than you can take with the average thermal probe. Accurate testing requires expensive equipment and a controlled environment. CPU die simulators such as Innovatek's or the one Bill Adams built are the way to go. Anyone want to spend more on testing equipment than their rig is worth?

While I find it amazing people actually come up with this stuff, I can't help but question why? We had somebody asking about water cooling that had just come from Pro Cooling Forums. He asked about water cooling and got handed a physics lesson. In frustration he came here. When someone asks what time it is, you don't tell him how to build a clock. You simply say, "5:07". And you certainly don't berate him. Kudos to the pioneers of this extreme thinking. But I could do without the hangers on who like to lord their knowledge of arcane techniques over others. It's just another way for the insecure to feel better about themselves by belittling others.

In this case JD was told the sky was blue.
"Is not."
"Is too, I can see it."
"Your testing procedures are inaccurate. From your longitude and latitude and the position of the sun and the moon, given the thermal refractance varient times pie are squared. The sky can't possibly be blue."

"But it is."
"LIAR, cheater!!! Breeder of goats!! The sky can't possibly be blue."

Blah blah blah. I was rather pissed off at this show of ignorance and sheer stupidity. But after a few days it dawned on me most of these guys are still air cooling and couldn't sharpen a pencil let alone design and build a block. Notice I made a statement saying I could build a Direct Die block in 30minutes. Well I lied. It took me 34 minutes. I figured I could give up and let the idiots win, or I could do it and put an end to this Direct Die mumbo jumbo. So I bring you:

[jaydee]

Well being my host seems to be screwed I will attach the pic.

[bigben2k]

Welcome back JD.

IMO (personally) you were right: testing blocks with air/water temps outside of a normal range is a whole different kind of testing, and if compared with a normal block test, can seriously skew the results. Let's see the "Cascade" tested with sub zero coolant temps, and get this out of the way, once and for all (Oh, if I had my tools with me!!!)


So is this direct die cooling block using the same top you've had for your last trials?

[jaydee]

I used one of my dead Epox 8K7A's to get the mounting and what not lined out.



---
The Duron 1gig was my choice as my other 4 computers are being used for DF and I had this chip sitting on a shelf for backup. Will be mounting my Maze4 and My Copper lemon Block to this CPU first on my Abit KD7. Once well tested with those blocks I will setup the Direct die block and compare against the other two. Note I will have the type k thermocoupler mounted under this CPU for better accuracy and repeatablility. Alo have another type K thermocoupler and a thermister all calibrated together for the best accuaracy I can get with what I have.


---
Jets inside containment area.

--
The middle peice with jets lasered in and then milled around to fit inside the containment peice.

--
Containment peice with sloped side to direct the water out of the containment area and to the outlet.

[Phant0m51]

Come on already, I wanna see temps!!! Here's a little secret I may have come upon...Go get some tar, like the kind you seal your roof with, and put a very thing layer of it in the small crack between the Die and the board it's attached to. It'll make it so water can't get in, and it'll be small enough that it won't affect temps much. That way, you can have a permanent Direct Die WB, w/o worrying about the glue coming off of the chip!

[TheDanMan]

Very impressive JD, I too agree with you on that however I kept my hotheaded self from responding to the nonsense. I can't wait to see temps. As for the tar, it is your proc, I don't know if I'd do it until I hear someone has good results with the stuff.

[jaydee]

Originally posted by bigben2k
Welcome back JD.

IMO (personally) you were right: testing blocks with air/water temps outside of a normal range is a whole different kind of testing, and if compared with a normal block test, can seriously skew the results. Let's see the "Cascade" tested with sub zero coolant temps, and get this out of the way, once and for all (Oh, if I had my tools with me!!!)


So is this direct die cooling block using the same top you've had for your last trials?

Well my worst block performs great at sub 15C ambient temps. This guy thinks his overclock rules, but people are getting that overclock on that same CPU with AIR cooling. Without quality testing it proves NOTHING as to how well the block cools. I can't figure out how these guys think they are so far ahead of proocooling but yet cannot understand these simple concepts. Yet they think they are full of "Facts" and the leaders in direct die when leufkins was selling direct die blocks a few years ago before he went under. I can't figure out why they somehow came up with the idea we said it was impossible? That was never an issue. The issue is how effective it is against a "quality" made block and then the reliability of it in general. Sure one guy might run a year with it, but WHY? If a good water block is just as good then why risk it. Well as most of you should know by now I have little issue risking things to prove if something is good or bad right or wrong.
This is not a NEW idea and it has been tried over and over again. Some by our own members here and they have not had better results than a standard block from what they say. Well I will add jet impingement to make it as efficient as I can and get down to reality with it.

Note my new tittle I asked Joe to put in for me.

As for the top it is shown above.

[jaydee]

As for sealing, I am going to use RTV silicone. I do want to take this apart someday. Silicone can withstand the heat and it is flexible enough to help keep from ripping apart when the die heats up and cools down. And I want it to be as clear as possible.

[Phant0m51]

BOOOO!!!!! NOOOOO!!!! My tar looks cooler than your silly RTV Silicone. The silicon might work better than mine, but so what? Mine's BLACK and 0VVn3Z y3R AzZzZ!!!!!

Hehe, actually the RTV will probably be alot better in the long run, good call.

The only problem I see with using jet impingement is that the DIE itself may wear away in time. But then again, it might not.

[jaydee]

Originally posted by Phant0m51
BOOOO!!!!! NOOOOO!!!! My tar looks cooler than your silly RTV Silicone. The silicon might work better than mine, but so what? Mine's BLACK and 0VVn3Z y3R AzZzZ!!!!!

Hehe, actually the RTV will probably be alot better in the long run, good call.

The only problem I see with using jet impingement is that the DIE itself may wear away in time. But then again, it might not.

LOL.

The Die itself is actually made from silicone from what I understand. It will errode, especially with the way I have it set up here. 6 jets powered by a HydroThruster 500GPH pump. I am sure it would take a while though. Once it is up and running and all looks ok I will leave it run for a long while before taking it apart again or untill it fails which ever comes first. Biggest concern I have now is the jets plugging up.

[Phant0m51]

Why don't they make the DIE out of something that can transfer heat better? I don't know what, because everything that transfers heat well is also electrically conductive (except for good ol' pure H2O). I would really like to see how long it'll take your jets to kill the processor, if they ever do. Maybe take it apart in 9 months and see if there are any 'dimples' eroding the DIE? Just a thought.

[jaydee]

Originally posted by Phant0m51
Why don't they make the DIE out of something that can transfer heat better? I don't know what, because everything that transfers heat well is also electrically conductive (except for good ol' pure H2O). I would really like to see how long it'll take your jets to kill the processor, if they ever do. Maybe take it apart in 9 months and see if there are any 'dimples' eroding the DIE? Just a thought.

The way this is setup is the jets will hit the die and then the water washes aside, backup and then out the outlet. Not the best way to do it, but the best I have seen yet. What would be ideal isto take a peice of copper and drill holes all the way through it and place it ontop of the die and aim jets into the holes to where you have a Cascade style setup. Problem is the die is just to damn small. Even the die on this Duron Morgan is bigger then the die on my XP1700+ T-Bred A but not by much.

[ZaVkE]

JD, what is the cost of such a block?
i might be interested if you can make one with metal fittings instead of plastic, cause i will be using it with a eheim 1060 (2250l/h) so that animal produces to much pressure for most plastic fittings.

[jaydee]

Originally posted by ZaVkE
JD, what is the cost of such a block?
i might be interested if you can make one with metal fittings instead of plastic, cause i will be using it with a eheim 1060 (2250l/h) so that animal produces to much pressure for most plastic fittings.

My HydroThruster is rated at 2,200L/HR and has a higher PSI rating than the 1060. Not even close to enough pressure to do any damage to the nylon barbs.

This is what I have into the block:

$.50 x 2 Hose barbs. $1.00
$.20 x 4 Hex screws. $1.00
$.25 x 3 for 2" x 3" acrylic peices. $.75
$5.00 in engraving/cutting costs. $5.00

$7.75 is what I have into it right now if I paid for the acrylic and engraving which I didn't being I am the manager here at work.

Not interested in making these for anyone though for many reasons. Anyone can come up with a do it your self version though. It isn't nearly as hard as making a good regular water block. Just have to aim water at the core and make an exit. No big deal...

[ZaVkE]

It is easy indeed if you have the material, but here in belgium you can't go to any other shop and get those acrylic pieces, let alone have them cut

Thats why i asked ...

Just have to keep on looking then

[jaydee]

Originally posted by ZaVkE
It is easy indeed if you have the material, but here in belgium you can't go to any other shop and get those acrylic pieces, let alone have them cut

Thats why i asked ...

Just have to keep on looking then

I hear that. Sometimes forget how easy it is to get things here. Every hardware store on every other block carries everything you need to make something like this. I am a little more fortuanate to be a manager a place that deals with acrylics, metals, and what not daily.

But still I would not consider making anything like this for anyone else. I would need your CPU anyway (as the designs are spacific CPU critical, this will work on my Duron 1 gig but not any other CPU that has a different layout) even if I was up to risking it.

[redleader]

Why don't they make the DIE out of something that can transfer heat better? I don't know what, because everything that transfers heat well is also electrically conductive (except for good ol' pure H2O). I would really like to see how long it'll take your jets to kill the processor, if they ever do. Maybe take it apart in 9 months and see if there are any 'dimples' eroding the DIE? Just a thought.

Because silicon is a semiconductor. Kind of hard to make transitors out of copper

Actually someday we may see diamond transistors, but don't hold your breath, its very hard to dope. 5-6 times as conductive as copper tho.

[mhayenga]

jaydee, as a reader and sometimes poster. You kick ass, reading about your projects adds *signifigantly* to my enjoyment of these forums... Don't go away or bother with those ignorant ass punks.

btw, where the hell is fixittt?

oh yeah, and yet another annoyance for the temps

[jaydee]

Originally posted by mhayenga
jaydee, as a reader and sometimes poster. You kick ass, reading about your projects adds *signifigantly* to my enjoyment of these forums... Don't go away or bother with those ignorant ass punks.

btw, where the hell is fixittt?

oh yeah, and yet another annoyance for the temps

Thanks! Not sure what fixittt has been up to. Havn't heard from him in a long while.

If I have this weekend free I might have this block hooked up Sunday. Tomorrow and Saturday will be devoted to testing the Lemon Block and the Maze 4 first on the Duron.

Shit I hope the Duron still works! Will test it out tonight. If not I will have to order a new one as this block will not fit anything else.

[satanicoo]

Quite some sad statements they have made over there. How can they say a block is better than another if one is getting water with lower temp than the other? :shrug:

[Khledar]

Give me a break jaydee, OBVIOUSLY the block on my phase change setup is better than any waterblock for watercooling... I mean duh.....







On another note, that is certainly one of the prettier direct die setups I've seen - good on ya!


[EDIT: Missed an S - fixed it]

[jaydee]

Originally posted by Khledar
Give me a break jaydee, OBVIOUSLY the block on my phase change setup is better than any waterblock for watercooling... I mean duh.....







On another note, that is certainly one of the prettier direct die setups I've seen - good on ya!


[EDIT: Missed an S - fixed it]

Hummmm.....How about a direct die phase change block!

[Phant0m51]

I can just see the DIE cracking in half as soon as you turned it on...

[Arcturius]

Originally posted by jaydee116
LOL.

The Die itself is actually made from silicone from what I understand. It will errode, especially with the way I have it set up here. 6 jets powered by a HydroThruster 500GPH pump. I am sure it would take a while though. Once it is up and running and all looks ok I will leave it run for a long while before taking it apart again or untill it fails which ever comes first. Biggest concern I have now is the jets plugging up.

It's silicon--note the spelling (not trying to be a jerk, the distinction happens to be a pet peeve of mine).

My TbredA die seemed to erode pretty quickly, even with a wimpy pump and no jets, simply exhausting from a 3/8" hose barb.
It hasn't affected the chip's operation, though, as near as I can tell. The actual interconnects are at the 'bottom' of the die, and most of the thickness of the die is for manufacturing ease, AFAIK. I don't recall how many layers an Athlon die is made of, though.

[t00lb0x]

Glad to see you back JD, but I wish I would of seen that original thread, didn't know what really happened.