Very nice work.

Were the heat sinks in you PSU live? Or did they have no connections to them at all?

Forgive me if this was mentioned earlier and I missed it. What make/model power supply are you working with?

JFetting,

I'll consider it... but after I remembered I had to use oil... it worked just fine...

I could have kicked myself for not remembering... :mad:

ozzy7750,

I don't think any of the heat sinks were "live" not anchored to any traces. I will put a meter on it after it's done without the main block to make sure. Only one of the regulators was not insulated from the heat sinks. So that's where I think there might be a problem. Not sure yet.

Krazy,

Not sure at the moment as I'm not home... I'll check and post the model number tonight.

Krazy,

It's a PowerTek 500W Switching Power Supply (Model Number ATX500D). It normally comes with two fans... I removed one previously and will probably undervolt the remaining one.

ozzy7550,

I checked it and there was some voltage potential between the "fingers"... So I was going to check it further and had just plugged it back in to get readings when I heard the worst sound you can hear... "*POP!*" ... and my UPS clicked off.

There was a small burn spot on the back of the circuit board... I can only suspect that a small piece of loose solder got on it and ZAP... dead PSU.

So, I'm off to the net to locate another one of the same make and model. :cry:

thats no good!!

So you dont think it was anything to do with the blocks?

When you say there was some potential there, are you talking 1V DC or are we talking big AC voltage? The problem being that that will then charge the whole system, ie if you have AC on the blocks, then you exposed CPU block will also be charged, and the way my system is, it is very easy to touch both water and my blocks. I need to pull my psu apart and check it out some more.

Bummer! :(

This place has a 500watt "titanium" version for about $20:
http://www.buypcdirect.com/dept.asp?dept_id=24-004

Here's another place with something looking similar for about $19:
http://store.yahoo.com/dealsurprise/po50dufanatx.html

Perhaps this?:
http://www.dealsonic.com/po50atxp4amd.html

Looks like a nice inexpensive (and well-laid out) PSU to work with. Do you know how clean the voltages are? I might look into getting one of these babies.

I ordered the EXACT same power supply (as far as I can tell) from us-depot for about $30 after shipping.

Since the old supply is dead anyway I'm going to isolate that one regulator that was grounded to the heat sink. Maybe that was the cause. But, first I'm going to do some measurements on the aluminum heatsinks and see what kind of voltage they have on them before I alter ANYTHING.

Also, I really do suspect I got sloppy and left a very small sliver of solder on the board. I'm going to double check EVERYTHING this time around.

Now that you have the power supply blocks made, do you really need to unsolder the MOSFETs to install the blocks? That looks like a big risky step that needs to be eliminated.

When you get the new one, I'd try a few things before opening it and voiding the warranty.

Reach inside with a small screwdriver while the PSU is running through one of the air intake slots or something and short both of the heatsinks to ground. If the new PSU doesn't blow up because of this, you can be pretty confident that it is not designed to have any kind of voltage across the sinks. If it goes out with a bang, that handy dandy warranty sticker is still intact and it's DOA in my book.

UberBlue,

yeah, it would be nice to do without unsoldering those... however, there are several large components in the way of the screws that hold them to the heat sinks. So, I am REQUIRED to remove them to change out the sinks.

Krazy,

I don't think I'll short them with a screwdriver... I think I'll check them with a meter through the vent first. If there is significant voltage present on one or both, I will just make sure and isolate that last regulator as that's the ONLY connection either of the "fingers" had.

Again, I am convinced that I messed up with a sliver of solder across two traces or something similar as there was a black soot spot on the casing and the board. I cleaned the area and it was still shorting. So it's fried.

The new PSU has shipped and should be here in a few days. Meanwhile I'm going to start on the HDD water block drive cage. I have some old dead drives around to use as "dummy" drives for the construction so I don't risk my new drives. :D

MMZ, actually all of the transistors/fetts are conducting to the block, in order to totally insulate them youll have to get some plastic screws to go along with the shims.

kinda funny, I didnt realize that either till now:p

Jon

i think they will probably have a plastic insert so that the screw doesnt touch the voltage regulator.

If the heat sinks do carry potential, they will for a reason, ie they will make up part of the circuit. By isolating that, the whole thing may not work.

JFetting,

ozzy7750 is right... all but one regulator (5 out of 6) had an insulating pad behind them as well as a plastic insert for the screw to go through... only one is actually conductive to the heat sink and just one heat sink.

ozzy7750,

That heat sink is NOT connected to ANY circuit, it is only connected to that regulator. I figure it is for a reason, but I don't know why.

All I know is that if I isolate that regulator, it should NOT affect the circuit as it would dead end in the original heat sink anyway (open circuit).

Again... I'm going to double check all this with my meter when the new one arrives.

No PSU yet... so I worked on the Hard Drive Cage solution today.

First cut out my templates. Then I sprayed some "3M Super 77 Spray Adhesive" on the plate copper and and, viola, easy to follow lines. You could, after gluing on the templates, score the lines to make it visible even if you rub the paper off.

Here are the plates for the Ati All-in-Wonder Radeon 9700 Pro solution...

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0096.jpg

Here, I have made my "tooling plates" out of 3/16" aluminum (greyish-beige plates) and have them secured down along with clamping blocks for one of the Hard Drive Cooler plates.

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0098.jpg

I have one 7/64" hole drilled through the all the plates and 5/16" hole milled through the copper only. This will allow me to secure the copper to the "tooling plates" without drilling into my table.

Here I've drilled and tapped all four corner holes and have secured the copper plate. I'm now checking to make sure my copper is square in relation to the table by first aligning with one hole's center mark...

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0099.jpg

Then moving the table in the "X" direction to another center to see how far off I am...

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0100.jpg

Only being off by about 0.02" is not bad... so, I loosened the bolts holding the copper and lightly tapped the copper so it turned to line up. Then double checked this again. "Dead on!" :D

Next from the second hole's center mark, I moved the table in the "Y" direction until I reached this center...

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0101.jpg

Not off at all... that I can see. A more accurate way to do this would be from the corners of the template itself as those points are farther apart. I will do this with the Video Card solution plates as the measurements on those are MUCH more critical.

Here I've set the depth of my 3/8", 4-flute endmill to 5/32" by putting a 3/32" drill bit under it for accurate measure. (The plate is 1/4" thick)

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0102.jpg

That's it for today... I'll cut the channel tomorrow. Peace... :)

Full system drawing for viewing...
 

One more image to tickle your imaginations... :D

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

What cad software do you use?

I like very much the look of your drawings. :)

Maybe it would be a good idea to hack the old ps case so its ready and you can just transplant the guts into there and have it ready without ruining both cases.


Jon

nicozeg,

I don't use CAD software really. I use Microsoft Visio 2k. I use it so much at work it's like second nature to me and I've been doing orthographic drawings since I was 6 years old.

JFetting,

Yeah, I am going to hack up the cover of the old one as that's the only place that this will protrude. I will probably use the power connectors off the dead one too, as I will be making custom cables just for this system configuration.
__________________

More to follow later tonight... gonna go eat supper now. :p

timelord, do you thinks its a good idea to split the flow just before getting to the cpu block?

i guess that i would prefer having the cpu block as a flow splitter, and sending 1 line to the video block, and the other to the psu + hdd blocks.

thats my guess :)

MMZ_Timelord:

If I may be so bold as to make a suggestion, I'd recommend tapping a small hole in the bottom (which is the highest point) of your heatercore and putting in a little needle valve and 1/8" stretch of hose. You can dump that back into your main loop up in the top part of your case. Doing this should let you clear the air out of the tanks of the core really easily; otherwise I suspect you'll have to pull it out and shake it a bit...

-J-,

I want the split right at the PSU the way it is because the flow to the CPU block is straight through the "T" at the PSU block... only the 90° branch on the "T" actually goes through the PSU block and then on to the Video Card Solution and then the Hard Drive Solution as these are VERY restrictive and I want the highest flow through the CPU block and then immediately back to the reservior.

pHaestus,

With the pump I'm running I actually have not had a problem with trapped air in the heater core. However, I will check it when I fill the system and see if it's necessary. Thanks for the suggestion.
__________________

Today's progress:

I started by pre-drilling the entrance hole and then using my 3/8" endmill, I bored it out. Then from that point I started milling my passages in the "Y" direction of my table first.

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0103.jpg

Here I'm on the third passage...

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0104.jpg

Quite often, usually about 5 or 6 times, I had to clear the debris from the channels as it was quite thick. I then added some 3-in-1 oil back on the endmill and the copper and ...

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0105.jpg

then I could continue milling with a clear channel... like this...

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0106.jpg

Whew! Last of the passages in the "Y" direction! :)

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0107.jpg

Here I've already done the exit passage, two of the bottom "X" direction connector passages, one of the top "X" direction passages and working on the second.

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0108.jpg

Okay, all the passages are complete...

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0109.jpg

Here is why I made the "tooling plates"... :D

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0110.jpg

Here is one finished plate after cleanup and sanding off the template paper next to the other one to be done with it's template in place.

http://pages.sbcglobal.net/mmz_tl_01...s/Dscn0111.jpg

That's about it for today... I'll probably do the other plate tomorrow and then get them soldered to their 1/16" back plates utilizing the "oven technique". :p

Until then... take care everyone! :dome:

just got to say that i love milled blocks, im stuck with my drill press :(

Milling copper looks like good fun :(

Just thought you might want to know. When I was looking at building my own res, I called GE's help line about the S-II kitchen and bath stuff. They told me that it isn't rated for continuous submersion. I don't know about your experiences in the past, but I've used some of that DAP aquarium sealant (it's also 100%) silicone in a res and it's held up well.

JoeKamel,

Yeah, I've changed it since... it never got solid enough to make it's own gasket. So I have 100% silicone (Clear, high tempreture stuff). I made a mold and used it to create a 1/16" thick gasket which I can remove and re-install, turned out very reusable. :D

Ahh, good good. Sorry, just thought it would SUCK to have to see all that machining and other work go to waste cause of $0.15 worth of gasket.

hey timelord, a simple question, with what you stick the templates to the copper?

Very nice work! What type of milling machine are you using to make that thing? And the obvious, how much did it set you back if you own it?

I might have to hold off on trying to work up enough cash to buy a drill press...

Regardless of that, that's an extremely well made block. Should be a very, very clean solution once finished.

-J-,

Quote from post #44...

Fyber,

This is basically the mill I have. Yes, I own it. I purchased it about 4 years ago.

"Invest in a good tool and it will last a lifetime" - My Grandfather (Gramps) who taught me to work with wood when I was 4 years old.

Keep in mind if you buy a milling machine of this type, you will invest another $300 to $500 or more on milling tool accessories. (i.e. - rotary table, endmills, collets, hand tools, calipers, machine grade T-squares, etc.)

I bought it because I can use it for other things than just computer water blocks. I just enjoy working with my hands as I don't have much opportunity to do so where I work.

I finished the other block just like the one I did above... had to have a pair. They are in the oven as we speak. It just got "pre-heated" so I'm going to go and flow some solder on it.

Pictures to follow later.

DUDE awsome work with that kind of mill.

BTW i just figured out that we are neighbors (sourt of). I live in Palo Alto!

Waiting for the rest of this to be done! maybe see it in person eh?