hey, yeah, got a question? does anyone know if you can get hypodermic needles in the US like that, cause I don't think you can do that here. If not, any ideas for alternatives?

edit: and i do mean in a legal way

go to a doctor, veterinarian, hospital,...
(most chance with a vet or hospital because they do operations, doctors don't do operations so i guess they dont have such advanced stuff... 100% sure you'll find some in a hospital :D )

Dieter: Short answer, yes :) (for me it works, have a look at the graph)

Long answer: I have no idea how it would compare to any other waterblocks. What I can say is that it walks all over my modified taisol heatsink (80mm papst fan). But then "c'est normal" as we say here!

FYI, cpu is an Athlon 1800XP and mobo is k7s5a - running @ 133/133


grvydude: (wild guess) you sure you can't order online?

edit: (oops) forgot to mention that the temps are for full load running burnK7, as reported by MBM5 for the cpu in-socket sensor

So the highest temp of that graph is after a long while of 100% stress?

(I don't see the point of a cpu temp/water temp curve)

well i understand where to get it, I know many doctors, many in my family, thing is they say it is illegal to sell hypodermic needles without a perscription here in the states... Anyone to back it up or tell me another source? (once again a legal source in the US)

The graph gives you the cpu temperature in relation to the water temperature. Ideally you would want the cpu temp to be as close as possible to the water temp. Here you can see that this is not the case (and never will be, no matter what waterblock you use).
The graph shows that the difference between these two temperatures is not constant :)

At the moment, my radiator is just a large bucket of water. From the graph, you see that if the water never rises above 30°C, then the CPU will never rise above 42°C (this is pretty much the case now)

If I used a really efficient (large) radiator instead of a bucket, then the water wouldn't go much above the air temperature (17-18°C), so the CPU would remain below 35°C.

That is illegal?? Some anti-drug thing or something :confused:
Just explain them its for a waterblock :p



Lolito_fr what is the max temp your water/ cpu gets on 100% stress for a while, and with which roomtemp?

30° water and 42° cpu?

yes, unless I get a proper rad :)

Well that isnt bad I would say, knowing its your first try (it is your 1st try is it :p)

Nice work man :)

Keep us posted as the rest of your system unfolds :)

[Political rant warning]
For those living in Europe, who aren't aware of this sort of thing... Our government here in the land of the not-so-free United States decided many years ago that we were to stupid to be allowed to make our own health care decisions and initiated a "War on (SOME) Drugs". As part of this war against readily produced, non-patentable, low profit margin substances, the decision was made that except for a few limited effectiveness drugs, it would be necessary for the average citizen to get the governments permission to care for any serious illness. This permission is given out (for a healthy fee) by government licensed people who are allowed to put "M.D." after their names. Part of this restriction includes a strict prohibition on sale or possession of hypodermic needles by anyone not having a gov't license or a permission slip from a gov't licensee. (Have to do something to encourage addicts to kill themselves of by making them re-use dirty needles)
This helps to ensure large profits for pharmeciutical companies, and helps to keep the crime rates high enough to ensure that police forces get plenty of funding and protect insurance company profits.

What this means for those wanting to duplicate lolito_fr's design in the US will need to either get the needles from an illegal source, or convince a friendly M.D. to give them permission to build computer hardware....
[/political rant]

Gooserider

Its a crazy world... I cant get needles to make a waterblock because a small number heroine addicts use them to kill themselves with:rolleyes:

I'm sure there are alternatives... I have seen coffee straws used before... what about turkey basting needles? Or are they too big...

You could always ask a heroine addict if you can have their old needles :D

EDIT: ebay is always a good source for illegal sales of banned goods :D

http://cgi.ebay.com/ws/eBayISAPI.dll...&category=3179

Located right in my home town too... how convienent and cheap, but you will have to find someone with diabetes to order them for ya.

EDIT2:http://cgi.ebay.com/ws/eBayISAPI.dll...&category=1210 doesnt require a medical problem to bid :D

The rant is based on fact... however... the applicator I got for my liquid arcrylic glue has a hypodermic needle tip that is flat (read not usable for injecting into people)... I'm sure these could be procured in other sizes as well... just look around a bit. :dome:

Like this and this and even this.

So don't say you can't get them legally... just say you have to look. :D

Looks like I may just have found a market for exporting the things lol:D

7$ a piece, compared to 0.06€ = nice markup :drool:

My first set of needles came with an inkjet recharge kit; for what its worth :)

Yes, recharge kits are another source, I was only showing other sizes... I don't think the diameter is the same as the one's you had. I didn't find the size reference as I was just making the point that they COULD be found legally.

check out lmno engineering's website ( http://www.lmnoeng.com/ ), for the formulaes to calculate the flow restriction of an annular outlet, or more specifically, to convert an annular (ring) "pipe" to a round one.

You'll find that the gap has an equivalent hydraulic diameter, hence my comment. You need to keep the exit velocity under the inlet velocity, otherwise the inpingement jet won't form.

I know it seems counter intuitive, when you look at the corresponding surface areas, but you can't beat the physics of it.

That's because if the exit velocity was higher then the inlet velocity, you wouldn't have enough pressure to smash the boundary layer, right?
(if the water gets away too fast you cannot create pressure)

Well Happy New Year guys :)
Thx for the link, bigben2k. Looks like I need to brush up on my physics!

Exactly. ;)

Been thinking about this. I see your point BigBen2k, about the equivalent hydraulic diameter being too small. Makes sense actually, seeing as the (predominant?) friction losses arise from contact between the fluid and the pipe walls. (hydraulic diameter is defined as four times the ratio of cross section to wetted surface)

However, the equivalent hydraulic diameter is only used to determine the pipe losses. In this case, the "pipe" is the length of insertion of the needle in the cup. At roughly 1.5mm, this constitutes a very short pipe, IMO. (even if the hydraulic diameter is only 1.8-1.25=0.55mm)
I haven't worked out the theoretical losses (any volunteers?), but I should imagine that they would be considerably lower than the head loss due to the nozzles themselves.
Therefore I don’t believe that the head loss due to this "restriction" would impact the overall flow rate a great deal and I can't see how this is suddenly going to stop the jet from impacting the base of the cup :confused:

So yes, in theory maybe the cups are too narrow. In practice though, I wouldn’t like to use smaller diameter jets (my fingers are too big!?).
The other option would be to increase the cup diameter. (dumb technical considerations include drill size here…) This would reduce the density of the nozzle array, and the velocity of the water rising out of the cups would suffer, reducing the secondary cooling effects.

Edit: Impact of extra losses due to cups, and other obstructions on the water return path was determined (approximately) here :
http://forums.procooling.com/vbb/sho...4&postcount=20

How did you fix the needles in the block? did you glue them in or did you use silicone? How did you manage not to close the inside of the needles while fixing them?

Yea, except that a box of 100 hypodermic needles, enough for about 4 blocks, costs like $6. Where as ONE of those needles costs $7. So you can go the hypodermic route and pay 7 bucks for 4 blocks, or "look harder" and pay $175 just for the needles for each block.

For the U.S. people at least, the easiest source is McMaster-Carr. Is there anything they don't sell?

Nice link. 1ft might be enough for a block, and the prices are reasonable. Excellent selection of sizes as well:)

for Grassi: I made the holes for the needles a little small, that way there was no need for glue or silicone. Just a nice push fit. Anything that does get in the needles can be cleared out with a small drill bit.

The hardest bit is making the holes in the base plate and the polycarb so that everything lines up. If I had to do it again, I'd space the holes out a bit more. 0.4mm between cups is a bit too close for comfort when you're making this by hand:D

yeah, i was wondering if someone would point out the price or not, i had saw that myself when i looked at those sites, and 7 dollars for one needle, no thanks, at that price it would be cheaper to straight out get the cascade... anyways, groth, thanks for the link, that is very helpful

Those prices are PER FOOT... not too expensive really...

One foot ... for one block... $7-12 ... I'd pay that.

i was referring to when people were saying use the refiller needles, sorry for any confusion, no the tubing idea is great, ill probably us it actually, thanks again.

My problem is, that the block was plannes without the needles, and as for that I have to glue them into the holes. My dad said, that I should try some sort of Loctite (I don't know if it is available in the US too). Its some sort of superadhesive (translated by google ;). Should I try this, or some sort of silicone?
So you say, if something gets into the needles, I should use a drill....

Am I correct in assumeing that stainless steel and copper wont produce a battery effect?

If so, would you be able to use those McMaster tubes for both the cups and the needles?

For example, you could use the 8988K78(0.013"od/0.007"ID) for the pins, and then solder or ASepoxy 5560K69(0.020"od/o.016"ID) tubes to a 2mm cpper plate, and that would give you the cups. That way if you dont have tiny drill bits, you can still try this.


Would such an idea work?