My psu is uber-loud..

[Eddie3dfx]

I have a 700 watt ocz that sounds like a vortex.

I have 2 options.

1. Replace fan with a 44cfm yate loon and risk overheating.. or

2. Modify it with a waterblock :)

So then I stumbled onto this page..

http://www.webx.dk/oz2cpu/pcmod/water-psu.htm

What say you?

http://www.webx.dk/oz2cpu/pcmod/psu-pc400-comp1.jpg
http://www.webx.dk/oz2cpu/pcmod/psu-pc400-out.jpg
http://www.webx.dk/oz2cpu/pcmod/water-psu01.jpg
http://www.webx.dk/oz2cpu/pcmod/water-psu05.jpg
http://www.webx.dk/oz2cpu/pcmod/water-psu07.jpg


Now here is what the inside of the OCZ looks like

http://i38.tinypic.com/frreu.jpg

http://i37.tinypic.com/262vexd.jpg
http://i37.tinypic.com/i1b3vq.jpg

Thinking of doing something similar with copper.. followed by soldering a flat copper piece on top so they are all attached, along with a flat hardrive block mounted on top.

[rocket733]

Personally I've never been a big fan of messing around with high voltage/amp electronics myself.

[Zehnsucht]

I dunno... Watercooling and 240V...
You are sure it wouldn't be easier and cheaper to buy one of SPCR's recommended supplies?

[tcorbyn]

I remember seeing a long time ago that a PSU manufacturer was working on a watercooled power supply. It may have been OCZ but I cannot remember. If they have not been able to make one it is obviously for a reason! If I were you I would replace the stock fan with the most similar fan you can get (obviously quieter) and leve it at that. You might be able to get away with the Yate Loon fan if you have the PSU in a separate compartment to other components (like in the TJ07). If you want a totaly silent PSU im sure I have seen some that use heatsinks instead of fans. Personally the most annoying thing for me when I an trieing to build a TOTALY silent PC is the PSU, but who knows, If you dont mind I would try out that watercooling a PSU on an old one if you have it and take some temp readings (with the case on). Remember to be sure to post them here! :up:

[MomijiTMO]

Koolance was the one.

I would just buy a bigger capacity psu. More load/max load = more heat = faster fan = NOISE.

[Chruschef]

I honestly don't believe that the stock PSU fan is better than a 44CFM yate loon...

My first step would be to check how the fan thats currently in your PSU is performing. A lot of the times companies through PoS cheap fans into the PSU that just get the job done, to keep prices low.

My second step, if the fan was actually performing better than a 44 CFM yate loon, then I would just return the PSU, or sell it. Messing around with something producing those voltages is NOT a good idea in my books.

[tiro_uspsss]

Eddie3dfx

only problem is that those heatsinks are live, so this will conduct power to your loop

please don't stick your finger in res to check level while running PSU......instant Afro with burned finger...... results

?? the heatsinks are live? lol wut? I *highly* doubt there is electricity being conducted to the heatsinks. The conduction of electricity produces heat - if those HS's were 'live' it would only add to their 'workload' :stick:

links to proof, otherwise utter bs :down:

[HotGore]

Interesting idea. I still would push some air through it.

[Boyu]

Get a bigger PSU, maybe your PSU is working to hard so the fan was spinning fast to handle the heat.

[link1896]

?? the heatsinks are live? lol wut? I *highly* doubt there is electricity being conducted to the heatsinks. The conduction of electricity produces heat - if those HS's were 'live' it would only add to their 'workload' :stick:

links to proof, otherwise utter bs :down:

okay, proof.

http://www.fairchildsemi.com/ds/FD%2FFDH44N50.pdf


the flange is connected to the drain. The drain is connected to B+, see a typical circuit HERE (http://www.onsemi.com/PowerSolutions/content.do?id=1331)

B+ is the rectifier and filtered mains

The flange of the mosfet (aka tab) is connected to B+, the mosfet can be directly mounted to the heatsink with NO insulating pad to reduce the thermal resistance between the mosfet and the heatsink.

[Sparda]

Eddie I use that ocz gsx 700 in one of my rigg before and change the loud blue fan to zalman f3 fan. Hook it up with Q6600(~3.86Ghz) ,7900gt(vmod) and bunch of other stuff and use it for half year without any problems before I got a better psu.
iirc the blue fan is rated 2200rpm according to johhny guru review. I think change it to S-Flex F or G(horizontal position+bearing) might give you a little bit peace of mind. Btw the fan is load control so the zalman fan will not spin at full speed all the time.

[kevinbo03]

They are most certainly live. Both times I've been zapped by a naked PSU, it was those damn fins.

In my defense, they weren't plugged in, they do hold a charge after all! :p:

[xTKxhom3r]

do it man be xtreme!!!

[Zehnsucht]

?? the heatsinks are live? lol wut? I *highly* doubt there is electricity being conducted to the heatsinks. The conduction of electricity produces heat - if those HS's were 'live' it would only add to their 'workload' :stick:

links to proof, otherwise utter bs :down:

Because the use of an isolator would lower the heat transfer ability significantly. Depending on what type the transistor is the cooling area is either GND or VDD.

[Jah]

If you check the pictures on the top where the PSU is modded you can that he is using plastic washers and some kind of thermal pad (most likely isolating) between the V-regs and the sinks to electrically isolate the IC:s from the sink. And yes it seems to be common practice to connect to the backplate of the IC:s to power.

IF YOU ARE NOT 100% SURE OF WHAT YOU ARE DOING DONT MESS WITH MAINS VOLTAGES!
Mains voltages may kill you or your children...

So judging from you questions don't mess with the PSU.

[T_Flight]

?? the heatsinks are live? lol wut? I *highly* doubt there is electricity being conducted to the heatsinks. The conduction of electricity produces heat - if those HS's were 'live' it would only add to their 'workload' :stick:

links to proof, otherwise utter bs :down:


You do know that KaptKrunch is one of our electrical guys here don't you? He can literally tell you how to design circuits, and has helped me before, and I have a pretty good understanding of electical stuff, but KaptKrunch has alot more than a good understanding of it, he was literally telling us which circuits would be able to power 15 fans at one time through one dial, and made suggestioons on which one to buy to handle it.

He has also posted diagrams on here to speed controllers. Full electrical diagrams.

I've been through and A+ class back years ago, and yes...if you want to get lit up real good just go in there thinking they are not live. You will get your proof instantly. They also stay live sometimes for a long period of time once powered down, and the smart ones will discharge the caps (the safe way), and understand what they are doing beforwe messing with what is basically a AC/DC power amplifier. These units won't just knock the ever lovin' crap outta 'ya, they'll kill 'ya and knock you accross the floor and leave you for dead.

[HuffPCair]

God one error and a leak and BOOM goes well about everything, even your eye brows probably.

[Eddie3dfx]

I honestly don't believe that the stock PSU fan is better than a 44CFM yate loon...

My first step would be to check how the fan thats currently in your PSU is performing. A lot of the times companies through PoS cheap fans into the PSU that just get the job done, to keep prices low.

My second step, if the fan was actually performing better than a 44 CFM yate loon, then I would just return the PSU, or sell it. Messing around with something producing those voltages is NOT a good idea in my books.

The psu has a 110 cfm fan.

[T_Flight]

If it has that high of a rated fan it's probably not spinning at full speed, but if it is, then that's pretty bad. That means the PSU is horribly inefficient, and is getting hot causing the fan to spin up higher and higher. Also, if it's heavily loaded it's gonna spin faster. That is why it's always a good move to go with a higher rated PSU than you'd ever use. The less you load it, the more efficient it is, the cooler it runs, and the slower the fan turns.

I have an OCZ EliteXStream 1000w and run a OC'd i7 system with a GTX280 in it OC'd, and the fan never spins uip any faster than it does when it's first turned on. You can almost count the blades going around. That PSU only gets slightly warm on one side...barely detectable above room temp even when it's beeing worked very hard with a full load on the CPU.

Silence is soemthing I don't worry about because it's impossible to attain, but that thing I can stick my ear against the fan and can barely tell it running. 4 inches away it's inaudible.

It's a very strong, and very efficient PSU. I am very picky when it comes to PSU's. I want the cleanest power, the most efficiency, and the coolest running, with as close to zero ripple as I can find, and that was the one I found that had all of that.

To use my favorite word...It simply ROCKS!

[cegras]

I've dealt with that thing before.

Just get rid of it.

[Zehnsucht]

The less you load it, the more efficient it is,

That is wrong.
Power supplies have the highest efficiency around 50% of their rated output. Beyond and below it drops off.

http://www.silentpcreview.com/files/images/corsair_psu/efficiency.jpg
(notice that the x axis is reversed to what you'd expect)

[articu]

i have a 600w OCZ stealthXstream and had the same problem. Apparantly its not uncommon for the thermostat to go, hence the fan being at 100% all the time. I just snipped the wires and am running it off a molex now.

[Eddie3dfx]

I think I'm going to see if I can stuff a 140mm fan in there.