Well I got the #rotor area all done except for breaking the walls between holes... also haven't cleaned out all the chips yet so it's harder to see the pattern.

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0136.jpg

I also started to drill the pilot holes for the 0-80 screws... (the bit is 3/64" in diameter :eek: )

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0137.jpg

I have not finished all the pilot holes as I have various chores to do to prepare for my trip... I will monitor the forums while I'm gone if anyone has questions. :dome:

that looks great

have you considered that maybe the gpu zone is excesively restrictive?

i mean, you got quite a wide path before it, and suddenly it reduces to 5 1-2 mm channels.

maybe increasing the number of holes so you have more channels for the water to flow?

apart from that it looks great.

BTW would you be so kind to share the height of the base pads for the mem and the core?

im looking foward to make a r9800 block too

-J-,

I want it restrictive in that area only BECAUSE of the core. Also, I want the flow restricted in these blocks as the main flow needs to go through the CPU block.

I'll do some calculations on the cross sectional area of the GPU passages around the #rotor columns and the inlet. I think you will find that they are VERY close.

As to the height differences between the GPU and the RAM...

It was mentioned by BladeRunner in a previous thread that dimension varies because of IC and solder ball height. The variance is quite wide making the dimension useless for all cards.

I could post the dimensions for my card... but this measurement will vary from manufacturer to manufacturer and from model to model. The variance is still to big, even of two identical model cards from the same manufacturer...

You'd be better off taking the measurement yourself. :dome:

Ahh... so now we find the reason why copper is in short supply... MMZ_TimeLord's got most of it :p

Very nice to see m8, love the hdd structure, and also nice to watch someone else put in the hours on a mill, takes some dedication to do all what you've done so far

Not been around here much recently as I'm sorting a unit atm :cool:

Thanks BladeRunner! :D

I won't be posting much to this until the 21st or so... but I will be monitoring the forums from afar. :dome:

Cheers!

Well, I'm still stuck in training here in San Diego... ;)

Hope to get some good work done when I return next week. :D

Cheers!!!

I must ask, what training are you doing?
I am in Atlanta right now for some sun cluster training (normally in houston).

I'm in San Diego for Voice over IP Soft Switch and Gateway training. (I'm in Telecom) :D

No kidding, I bet that thing isn't light. :D

Looking good man!

Another silly idea come to me mind, does anyone know how well zinc xfers heat? I ask cuz looking at your original Alum. heatsinks on the voltage regs I am wondering why not use Alum. tubing on the orginal heatsinks by using an "aluminum solder" such as Alumite or similar brand (98%zinc/2% Alum soldering rods). Theres a guy on e-bay sells small packs of the stuff for like $5-10(US). I have ordered some before and the stuff works great, I just don't know how well it xfers heat.

-MC

Well... nothing got done yesterday... so today I did a little.

First I milled a shallow run on the passages that were at right angles to the table... and some small start around the upper portion of the #Rotor area. Also (after breaking 3 bits in the tiny holes) decided to put two of the larger cap screws in the area that the passage crosses over to the back plate.

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0164.jpg

Then I got fed up with the template... (3-in-1 oil was breaking down the 3M adhesive) ... so off it came (after I marked the corners of the block outline :D )

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0165.jpg

Then I finished milling the right angle passages to depth...

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0166.jpg

I'm going to try and get the rest of the water passage side done tomorrow.

Cheers! :D

The same happened to me while working with paper templates. The reason is that oils in spray cans have some solvent that loosen the adhesive. Is better using some thin oil applied by drops and the templates will last a lot more.

nicozeg,

My 3-in-1 oil is not in a spray can... it's in a sqeeze can with an applicator tip.

I do know that it's one of the oils itself that breaks down the adheasive. I may have to go to an adheasive that is oil resistant. :shrug:

i'm extremely impressed so far! keep it coming :)

Hmm, I also use 3m super 77, and switching oil just worked a lot better. Maybe you can try motor oil.

Edit: I remember another trick; use some water soluble adhesive, it resist oil very well. ;)

nicozeg,

Hmmm... I'll pick up some SAE30 today and give it a shot. Unless you have another recommendation.

That thing about the water base adhesive could work.

Is something I haven't tried with oil, but I use it a lot when dremeling small parts.

The one I use is UHU STICK, yes, the same thing kids use in school!

First do some light sanding on smooth surfaces, then generously rub the bar on the copper, so it build a good sticky layer. Quickly apply the printed paper and press it on the surface. let it dry a couple minutes. Bonds very hard if you made it OK, and wash out easy with water.

nicozeg,

I will see what works over time, but for now the 3M works long enough to get the blocks done. :dome:

Today's episode is on finishing water passages and #Rotor area cleanup...

First I moved the block to my rotary table so I could get the angled passages milled straight...

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0167.jpg

Here I've finished the long taper passage into the #Rotor section...

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0168.jpghttp://pages.sbcglobal.net/mmz_tl_01...s/Dscn0169.jpg

Here I've completed the exit side of the #Rotor section... taper is quicker, but this is after the #Rotor section, so it's not as vital to have a gradual taper.

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0170.jpg

This shows the use of a 1/16" endmill to create a "connector" passage around the "corners" of the #Rotor grid so that flow is not TOO restricted. :p

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0171.jpg

Side passages are done and most of the #Rotor pins are completed. Only small trimming with a X-Acto knife left to do...

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0172.jpg

Here is a false color image to show the pins a little clearer...

http://pages.sbcglobal.net/mmz_tl_01.../Dscn0172a.jpg

Completed block... ready for milling on the reverse side for RAM and GPU height and keepouts for the SMCs around the RAM.

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0174.jpg

That's about it... until tomorrow then... Cheers! :dome: :p :cool: :D

looking good :)

nice work man, keep it coming :)

definately looking good, whats the outter dimentions of that peice of copper?


Jon

Well, way back at the beginning of this work log I started on the Power supply solution and blew the supply in the course of events.

I ordered another one and they sent me the wrong supply... I called them... they sent me the wrong supply again ... with an RMA for the first wrong one. YEESH!

So I ordered what looked like an identical supply (by Rhycom)... The general layout is the same... but the board is indeed a little different. :(

Two shots of the Rhycom supply with the cover off...

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0175.jpg


http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0176.jpg

Here's a photo comparison of the labels and casing covers... looks like the same actual manufacturer to me. :p

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0180.jpg

After desoldering the Regulator ICs... the first heat sink appears to be almost identical to the one I made for the PowerTek supply... so no modification needed here...

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0183.jpg

The second one's holes were a little different on two of the ICs... so they would mount like this unless I move the holes over...

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0184.jpg

So... that's what I did ... I moved the two holes over 1/4" and tapped them... also... the holes for the mounting screws that go through the board are completely wrong now. I had to drill and tap those as well.

Ultimately this means that my original solution will work... but it will not look as good with all the extra holes in it. I may end up doing a new design down the road, but for now... it's okay.

More to come tomorrow... CHEERS! :D :p :dome:

Ahh you are coming up against the same types of problems I often have, in my experiance The exact same item often isn't the same in layout / design. I had this ages back with my GF3 cooling. The original I used the four smaller holes around the GPU to mount the GPU block, these holes were part of the nvidia reference card layout. Making a replacement, mk 2 Asus cooler, on the new Asus gf3 card, (that was the same model as the first one), these holes were missing!. It's annoying and will make it hard to make any unified product to fit universally.

Had similar to this just now in fitting my 9800 pro cooler to the Asus 9800xt. ok so this is a different card version so it's to be expected but many slight component placement differences on the card, some very subtle even though the cards are the same basic size & layout meant it wasn't a quick swap over. Took me the best part of a day with a dremel to modify the cooler enough to fit including the fact the overall pcb ram chip height was a little less.

Bladerunner did you mention a few pages back or in another thread about antec psus? Im thinking about adding my CWT 550w to my list of things to w/c and its exactly the same as an antec truepower 550w on the insides. Do you know if the layout is good inside?

I have the same thought to water cool my powersupply... and i know i will probobly fry one so i bought 3 at the same time...
if i don't end up using it now... i'll be building another machine in the future and use it then...

I just bought 3 pretty cheep 450 watt PS. only $50 each CDN

Looks good so far...
before you dismatle the PS will you be doing a voltage check accross the heas sinks to see if there is any stray current running through it?