Passive Water Cooling Emergenzy FAN System
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
Passive Water Cooling Emergenzy FAN System
I was thinking of making a fanless ( or almost ) water cooling system - putting copper pipes on my room floor ...
What I need it a easy way to monitor the water temp and when need arises start the 120 MM fan that sits on the radiator.
Anyone have any ideas?
What I need it a easy way to monitor the water temp and when need arises start the 120 MM fan that sits on the radiator.
Anyone have any ideas?
most fan controllers would work; like a DigitalDoc5 with a sensor in the water.
There is however, a better way- make a large reservoir, like a 5 gallon bucket. With 5 gallons of water to heat up, any changes in water temp will occur very slowly; so even if you can't cool a processor under full load, it will take hours of full load to heat up the water too much.
There is however, a better way- make a large reservoir, like a 5 gallon bucket. With 5 gallons of water to heat up, any changes in water temp will occur very slowly; so even if you can't cool a processor under full load, it will take hours of full load to heat up the water too much.
-
- Posts: 587
- Joined: Fri Aug 01, 2003 10:45 pm
- Location: North Billerica, MA, USA
- Contact:
I'm not at all sure that most controllers would do what you are after, as they seem mostly intended to adjust the speed of a running fan, not turn a fan on and off. Some do have switches that will do a power cycle, but those are manual, and you implied that you wanted something automatic...
Options as I see it,
1. Digi-doc as suggested earlier, it does do automatic fan power control, but may be overkill for your needs.
2. If you are using Windows (you have my sympathy...) look at Speed Fan, I think it will give you software fan power control, however beware of the question of what happens to the temps and fan control if you get a BSOD or other crash. (I require that my fan controller NOT be dependent on software that runs on the main system)
3. I've found a couple of circuits here and here
that will give both on / off and temperature related speed control with a relatively simple setup. I prefer the first circuit. (Note that these will NOT allow 3 wire RPM monitoring, I had a thread over in Fans and Controls where we came up with an additional bit of circuitry to enable that.
4. You can purchase thermostatic switches of different sorts and ranges from places like McMasters, the challenge would be finding one that would fit your application, and figuring out what temperature you wanted it to turn on at.
Gooserider
Options as I see it,
1. Digi-doc as suggested earlier, it does do automatic fan power control, but may be overkill for your needs.
2. If you are using Windows (you have my sympathy...) look at Speed Fan, I think it will give you software fan power control, however beware of the question of what happens to the temps and fan control if you get a BSOD or other crash. (I require that my fan controller NOT be dependent on software that runs on the main system)
3. I've found a couple of circuits here and here
that will give both on / off and temperature related speed control with a relatively simple setup. I prefer the first circuit. (Note that these will NOT allow 3 wire RPM monitoring, I had a thread over in Fans and Controls where we came up with an additional bit of circuitry to enable that.
4. You can purchase thermostatic switches of different sorts and ranges from places like McMasters, the challenge would be finding one that would fit your application, and figuring out what temperature you wanted it to turn on at.
Gooserider
-
- Posts: 48
- Joined: Sat Apr 26, 2003 3:34 pm
- Location: Wisconsin USA
If you've got enough surface area on your tubing, I doubt you'll need a fan. It never hurts to be safe, though.I was thinking of making a fanless ( or almost ) water cooling system - putting copper pipes on my room floor ... What I need it a easy way to monitor the water temp and when need arises start the 120 MM fan that sits on the radiator.
I'm working on something very similar. I'll be using about 16 feet of 3/4" copper tubing attached near the floor of my computer room. I've got most of the parts and will fire up the torch tomorrow to solder it all together. I'll post pictures in a couple of days.
My first attempt at a passively cooled radiator was the Tower of Cooling Power. It has about five feet of 3/4" copper tubing with fins attached. With no fan it keeps the CPU temperature at 41-44 degrees C, depending on room temperature and CPU load. The thread includes a test I did with a fan pulling air through the radiator. It didn't change the temperatures much. I expected more of a difference.
I've used SpeedFan on a server with Windows 2000 Server. Anytime the OS is not running the fans default to full speed. Very nice program, and it's free.2. If you are using Windows (you have my sympathy...) look at Speed Fan, I think it will give you software fan power control, however beware of the question of what happens to the temps and fan control if you get a BSOD or other crash.
I'm using the second circuit that Gooserider mentions to control a couple of fans on my desktop comp. One pot sets the minimum cold fan speed and the other pot sets the fan speed at operating temperature. It works great. Highly recommended.3. I've found a couple of circuits here and here
that will give both on / off and temperature related speed control with a relatively simple setup.
Thanx you guys - this helped alot...
As I dont know much about electrics I just want to clarify something about this powercycle issue you mentioned. Im not quite clear on what that is.
Is it not enough if I have a fan controller that feeds the fan juice at a too low voltage for the fan to start as default. And then when some idicator or other tells it to - raises the voltage above the fans start limit and thous the fan starts?
And on another note im also trying to figure out a way to remove the PSU fan and putting in somekind of watercooling block - but im a bit scepticall if I even should try this. But those fanless PSUs are damn expensive - might buy one anyways though.
Im looking forward to your promised pictures joesgarage11 hehe
As I dont know much about electrics I just want to clarify something about this powercycle issue you mentioned. Im not quite clear on what that is.
Is it not enough if I have a fan controller that feeds the fan juice at a too low voltage for the fan to start as default. And then when some idicator or other tells it to - raises the voltage above the fans start limit and thous the fan starts?
And on another note im also trying to figure out a way to remove the PSU fan and putting in somekind of watercooling block - but im a bit scepticall if I even should try this. But those fanless PSUs are damn expensive - might buy one anyways though.
Im looking forward to your promised pictures joesgarage11 hehe
You could feed it too low voltage, though that would be kinda wasteful.
Overall, I wouldn't worry about it a lot. I unplugged my radiator fans yesterday (dual 120mm sized radiator) and even in a poor passive cooling set up, fins horizontal and the fan shroud still on, the water temps went to 20C above ambient. If you actually have a passive cooling optimized setup 5 to 10C would be easy.
A watercooled PSU is possible; BladeRunner at www.zerofanzone.com has done it the hard way; the easy way, though less effective, is to stick some copper pipe on the MOSFET heatsinks; I'm probably going to do this to my 120mm Fortron; the fan insures room for the pipes and it won't be too expensive if I screw up.
Overall, I wouldn't worry about it a lot. I unplugged my radiator fans yesterday (dual 120mm sized radiator) and even in a poor passive cooling set up, fins horizontal and the fan shroud still on, the water temps went to 20C above ambient. If you actually have a passive cooling optimized setup 5 to 10C would be easy.
A watercooled PSU is possible; BladeRunner at www.zerofanzone.com has done it the hard way; the easy way, though less effective, is to stick some copper pipe on the MOSFET heatsinks; I'm probably going to do this to my 120mm Fortron; the fan insures room for the pipes and it won't be too expensive if I screw up.
-
- Posts: 210
- Joined: Sat Jul 19, 2003 1:40 pm
- Location: Kortrijk, Belgium
-
- Posts: 587
- Joined: Fri Aug 01, 2003 10:45 pm
- Location: North Billerica, MA, USA
- Contact:
Its a fancy way of saying Turn on / Turn off... The idea is to 'cycle' through all the settings of the power switch.As I dont know much about electrics I just want to clarify something about this powercycle issue you mentioned. Im not quite clear on what that is.
I've never tried it, so I don't KNOW that it would cause a problem, but I wouldn't recomend it. I'd consider it not unlikely that running a voltage below the starting value through the fan coils might generate enough heat to be a problem. IMHO you'd be better off to choose a controller that switched between completely off (0 Volts) and enough voltage to be sure the fan will run.Is it not enough if I have a fan controller that feeds the fan juice at a too low voltage for the fan to start as default. And then when some idicator or other tells it to - raises the voltage above the fans start limit and thous the fan starts?
-
- Posts: 587
- Joined: Fri Aug 01, 2003 10:45 pm
- Location: North Billerica, MA, USA
- Contact:
Its a fancy way of saying Turn on / Turn off... The idea is to 'cycle' through all the settings of the power switch.As I dont know much about electrics I just want to clarify something about this powercycle issue you mentioned. Im not quite clear on what that is.
I've never tried it, so I don't KNOW that it would cause a problem, but I wouldn't recomend it. I'd consider it not unlikely that running a voltage below the starting value through the fan coils might generate enough heat to be a problem. IMHO you'd be better off to choose a controller that switched between completely off (0 Volts) and enough voltage to be sure the fan will run.Is it not enough if I have a fan controller that feeds the fan juice at a too low voltage for the fan to start as default. And then when some idicator or other tells it to - raises the voltage above the fans start limit and thous the fan starts?
Zhentar, quick note: be careful with this. If the heatsinks are live, that means your entire watercooling setup, and probably your case, will be live too. I don't think I have to go into why that's dangerous.A watercooled PSU is possible; BladeRunner at www.zerofanzone.com has done it the hard way; the easy way, though less effective, is to stick some copper pipe on the MOSFET heatsinks; I'm probably going to do this to my 120mm Fortron; the fan insures room for the pipes and it won't be too expensive if I screw up.
Semm
Dakhor, yes. In some PSU's the large heatsinks are electrically charged. And if they are charged they may have a very dangerous charge in the neighborhood of several hundred volts DC.
EDIT: I'll share one of my experiences with a live PSU. This was many years ago, way before the SPCR days. We had a machine at work that had a squealing fan in the PSU. Swapping a fan out of a PSU is a pretty simple thing, but I was being lasy and kept putting it off. When I finally got around to fixing it the guy who used that machine told me not to worry about it, since the squealing had stopped. I looked around the back and sure enough, the squealing had stopped because the fan had stopped.
Now this was only a Pentium166 system, but the PSU was the only exhaust and the heat had definitely been building up. The top of the case was so warm that you couldn't resy your hand on it.
This system was tied to the corporate satellite system, and if it was turned off it would take an hour or more to get the connection back up properly. That fact, combined with it being in the middle of the work day, my concern about the heat, and my general laziness, all contributed to my decision to just swap the PSU fan out with the machine on. It was only a 55watt PSU after all.
Ya'll see where this is going, don't you?
So I pull the case off, unmount the PSU while leaving it on, and take the cover off the PSU. While I'm fidgeting in there trying to unscrew the fan the side of my hand comes to rest on one of the heatsinks in there.
The electric shock traveled instantly up my arm and caused every muscle in my arm to contract. The resulting convultion made my own fist make contact with my own face, hard enough to give myself a bloody nose.
And that was only a 55 watt AT-PSU. I can only imagine the pain a 400 watt could inflict.
EDIT: I'll share one of my experiences with a live PSU. This was many years ago, way before the SPCR days. We had a machine at work that had a squealing fan in the PSU. Swapping a fan out of a PSU is a pretty simple thing, but I was being lasy and kept putting it off. When I finally got around to fixing it the guy who used that machine told me not to worry about it, since the squealing had stopped. I looked around the back and sure enough, the squealing had stopped because the fan had stopped.
Now this was only a Pentium166 system, but the PSU was the only exhaust and the heat had definitely been building up. The top of the case was so warm that you couldn't resy your hand on it.
This system was tied to the corporate satellite system, and if it was turned off it would take an hour or more to get the connection back up properly. That fact, combined with it being in the middle of the work day, my concern about the heat, and my general laziness, all contributed to my decision to just swap the PSU fan out with the machine on. It was only a 55watt PSU after all.
Ya'll see where this is going, don't you?
So I pull the case off, unmount the PSU while leaving it on, and take the cover off the PSU. While I'm fidgeting in there trying to unscrew the fan the side of my hand comes to rest on one of the heatsinks in there.
The electric shock traveled instantly up my arm and caused every muscle in my arm to contract. The resulting convultion made my own fist make contact with my own face, hard enough to give myself a bloody nose.
And that was only a 55 watt AT-PSU. I can only imagine the pain a 400 watt could inflict.
Last edited by Rusty075 on Mon Oct 06, 2003 11:43 am, edited 1 time in total.
Ok thanx -- Ive got a Seasonic Super Silencer 400W , I dont know if the heatsink is live or not but my father can check that out propperly if I dont find some info about it...
I guess I could buy a new one - wich might be better suited for watercooling. After all the Seasonic is mighty quiet anyhow and can be used in some other comp.
/Dakhor
I guess I could buy a new one - wich might be better suited for watercooling. After all the Seasonic is mighty quiet anyhow and can be used in some other comp.
/Dakhor
-
- Site Admin
- Posts: 12285
- Joined: Sun Aug 11, 2002 3:26 pm
- Location: Vancouver, BC, Canada
- Contact:
In case people have forgotten, we do have a passive water cooled system article here: Boxing & Watercooling to Silence.
Then there's Ron's Water Cooled HDD Silencing Enclosure. That's actually a part of a massive passive water cooling setup that uses a bunch of copper pipes on the floor of his garage. The article is at overclockers.com
Then there's Ron's Water Cooled HDD Silencing Enclosure. That's actually a part of a massive passive water cooling setup that uses a bunch of copper pipes on the floor of his garage. The article is at overclockers.com