You might want to try a different (or more) flux on that, sounds like it's not fluxed up right.

the solder and the flux came in the same pack so I assume they where ment for each other. I tried lots of flux and little flux ect.... The flux burns off long before the solder starts to melt.:confused:

Do you put the flux on first? Thst what i am doing and it just burns up and leaves a little wet film on the block. then i put the solder on and it just peels off.

A very short Q
Flux?

Have you read the thread?

Also note I am trying this on Aluminum and not Copper. I think Copper will be easier.

I am installing my Copper block now and will test it agains my Aluminum. If the temps are within 1C I am going to drop the Copper all together as it is not worth it to make. I just took off the Aluminum block I made before with the Epoxy and it is in great shape. I smacked it around and tried ripping the top off by the connectors and all I did was dent it up a bit.:D I think it will be fine with that Epoxy. I will see how the Copper one does. I will send Joe the Epoxy versions aswell and we will see what he finds.

A possibly problem with epoxy is due to the water causing it to soften, though the extent of this varies with different water addititves and different epoxies. I would say copper would be a lot easier to solder, definitely try it on the Cu block.

This stuff I am using is supposed to be 100% water proof. And it says it can be used to fix radiators. I can't imagine any computer water cooling stuff effecting is any worst than a cars radiator. Those get hot to!!

Yes I have

To Lymmen

quote:
--------------------------------------------------------------------------------
Originally posted by Lymmen
A very short Q
Flux?
--------------------------------------------------------------------------------

Flux=Flussmedel

Maybe you should get a English to Swedish dictionary eh?

I already got one! :p
but in my dictionary it's stands "stadd i omvandling" not much to help :)

Err, aaaa, YES I did use plenty of flux.:)

I would still like the copper ones. I don't have a problem with the O ring and screws either. That way if Idecide to put a plexiglass top on I can. Sound fair? By the way been watching and the blocks look good. Let me know when I can get one for testing.

If you are going Copper you are probably better off using the Maze 2 as it isn't going to be much cheaper for me to make a Copper one. The tooling alone for proper Copper milling is expensive not to mention the stress on the mill is to much to risk right now as I cannot afford to be replaceing parts on it every 3-5 blocks. The first copper one I milled was sheer tourcher on it. I revised the code to make it easier but it takes 3 times as long now. I went from 13 minutes from Al to 1hr with Copper. Thats just the block that dosn't include soldering the top on, drilling and tapping the holes. Copper just isn't going to be worth while for me at this point.

Copper is MUCH harder to mill than I was let to believe. I really don't want to damage my Mill for just a .5C decrease in temps. I am not making much money on water blocks as it is so I don't see much reason to beat it up.

My main profits will be from the fan controlers. The mill can etch the PCB boards for them. Once I get the schematics the way I want the mill can make the boards and I can solder the parts in.

As of now I think I am going to stick with Aluminum as it is really easy to work with and I have had plenty of interest from people in them. I am trying to think up a copper insert design which I think would be the best. Aluminum seems to transfer heat faster and Copper absorbes heat better. So putting a copper insert in to absorb the heat from the core and the aluminumn to remove the heat from the copper insert may be the way to go if I can find a way to make the heat transfer from the copper to aluminum good.

I really should have started with the fan controlers first and blocks later.

Things are still under development as of now. Not going as well as I hoped. But what does.

Any question e-mail me at this address: jaydee116@attbi.com as my webhost is total crap lately. My site has been down for a few hours and my e-mail has not worked right for weeks.

I wouldn't bother with a waterblock with two materials. you'd have a 3 piece block then, so even more chance for leaks.

Just stay with the 2 piece all Al one

The copper insert would not go all the way through the block. So it would still just be the top where any leaks may accure. I doubt it will do any better but I am going to try it anyway.

It's not an unreasonable question Lymmen. From http://www.superiorflux.com/ comes the following quote...

I can tell you that JB Weld and Epoxy will NOT work in cases where there is a temp swing. I know this as I originally used JB Weld to afix 1/2" barbs to my 3/4" IN-OUT on my heatercore. I let the JB Weld set for 48 hours before I began leak testing. I leak tested with a Danner 5 in a loop with the heatercore for 24 hours with no leaks whats so ever. I then installed the system where it was running for 2 weeks with no problems. Then one day I found a drop...the drop increased in size....I took the system down and used some g00p on the fitting, problem solved. Next thing I know, about a week later the other fitting began to leak....simple g00p job later and everything was right again in my world. This was 6 weeks ago and no leaks whats so ever detected. g00p is designed to withstand expanding and contracting with varrying heat loads, where JB Weld clearly was not. A waterblock has even greater temp swings, therefor I would hafta say the only proper way to seal a WB is either with solder, or with a silicone O-ring that can withstand high heat.

I have been running the epoxy block for a couple weeks now and no probs. There are dozens of types of epoxies. You have to use the right one. The stuff I use is rated at 10,000PSI sheer strength and up to 200F.

JB WELD(not that I would use it) was designed to work with temp changes. It was originally made for cracked engine blocks!!!! You have to make sure your mixture is 50/50 and the area is very clean or it will not work right. JB Weld dosn't have the right consistancy. I need the liquid stuff so it sticks. I then clamp the top on with C clamps. I have been trying to get the top off but it will not.

Anyway I like the silicone idea as I use it all the time at my job. That stuff sticks like glue when it dries and it will take a lot of heat. And it is easily reaplied in the future.

I cannot design my block to withstand user error though. That just isn't going to happen. If the user screwes up they are on their own. Just like any other part in a computer or anything else. I will give a set of standards for each block. And if they are not within those standards it is all on your hands if it fails.

jaydee116

I remember watching a program once a while back about Boeing's new airliner, (757 or 767), can't remember which it was now, but was horrified to learn the wings are glued together!!!. no joke they were glued together :eek: . Never been in a 757 or 767 myself :D.

Still the point is if a commercial airliner can be put together like an airfix kit then certain epoxies will do the job after all those wings go through far greater loads and temp changes than a CPU water block.

Of course we don't know what they used and to be honest I still don't see Glue of any type as an acceptable engineering solution myself. I guess I wouldn't sleep well knowing nothing else but glue was holding my water block together. I personally don't see it as an acceptable solution for heatsink mounting either, (wish ATI wouldn't also :( ),

So to sum, yes it probably could work if you use the correct stuff mixed the correct way, but I'd rather not chance it just in case it wasn't the correct stuff mixed the correct way or found to go soft after 6 months of coolant contact etc. All copper blocks with brass barbs and soldered together air pressure tested to 2 bar = nights of restful sleep for me....... http://www.zytec.worldonline.co.uk/p...ons/sleep2.gif

Bladerunner, I'm pretty sure it was the Boeing 777 project, their brand new long range wide body. Anyway, I am very sure they had numerous chemical engineers, plastics engineers, and mechanical engineers working on the exact formula and Boeing spent many millions of $$ and many weeks on research before they implimented their glue design. Obivously jaydee this isn't a viable option for you given your goal of economy in your blocks. Personally I think a Al base with a screw down Al top and a silicone O-ring is the best option for you. I know I for one would deffiniatly like to see this design better than anyother short of soldering the top.

Yes, looks like it is back to the drawing board. I only have 1/8" from the channel to the outside edge og the block which isn't enough for an O ring or silicone. I will have to narrow up the channels to 3/8". Then my surface area and flow will decrease. probably raise temps some. I will go look for some proper sized O rings today and mill out a 3/8" channel block and see if I can work it out. I always wanted my tops to be removable anyway.

what you should do jaydee is turn off your pump, let your cpu get to 65C or so, then turn the pump back on, and see if any part of the bond is a little weak

Just got back from a meerting at work.:(


I am going to heat one up with a tourch to. Once I get my website semi finished and posted I will work on it. The Epoxy is supposed to be good to 200F but that dosn't mean the top will stay on.

wouldn't it be easier just to shove it in an oven?

"Also note I am trying this on Aluminum and not Copper. I think Copper will be easier."

Um. You cant solder aluminum. You have to use a special braze for aluminum...