Time to help the fanless Zen?

cloneman · Post by **cloneman** » Sun Apr 08, 2007 11:38 pm

I just got a Fortron Zen Fanless 300W for my system... and lo an behold I get a free 9800 Pro from a friend. Of course this drastically changes my power draw (previous card a GF4 MX440)

Time to consider helping the fanless PSU with a fan, or moving it outside the case? The top seems quite warm, I would estimate 55 degrees. This is after ~ 10min VGA load.

Specs:

P4-2.8 (Scythe Ninja) (Passive - 120mm (BQE's) Antec case Fan @ 12V)

Radeon 9800 Pro - cooler throws hot air outside

2 Hard drives, 1 DVD drive.

I know the Zen is supposed to shut down if it reaches critical temp.... but I don't think I want to take it there.

Bluefront · Post by **Bluefront** » Mon Apr 09, 2007 1:44 am

300w is plenty for that system, and the Zen should handle it without sweat. That PSU is designed to be cooled by case exhaust airflow, without a dedicated fan. Of course helping it out with careful design, won't hurt anything. First off.....guessing at temperatures is not so good. I'd get a small external temp sensor device for a more accurate reading. I have two systems using a 300w Zen without problem, but I made certain that case airflow could cool the Zen ok. Take a look at this system......the Zen is mounted partially outside the case, so there is plenty of passive airflow over all sides of the thing. It never goes much over 40C, no matter where you measure it. That system uses a P4-3.4NW, and had an X800XP in it for a while. There are no exhaust fans in the case....only the lower intake fan. My other Zen-based system is similar. Link

cloneman · Post by **cloneman** » Thu Apr 19, 2007 8:10 pm

Depends on the side I put my finger on. The area near where the cables come out is _scorching_ after some gaming ( I can't leave my finger on for more than 3-4 secs. The top is also very hot. The sides and bottom are ok, but then again, those are probably just farther from the heat sources...

I added a fan on the ninja - although this liberates the Zen from abosorbing heat from the ninja, I'm also concerned it's counter-acting the effect of air being forced to go through the zen as the casefan pulls. (look at pic for explanation).

Note that the top the the Zen near the red area is also pretty hot

jaganath · Post by **jaganath** » Fri Apr 20, 2007 12:56 am

If you remove the fan on the CPU H/S and the airflow "short-circuits" as planned (not certain) then this will detrimentally affect CPU cooling, I would imagine. One way to help cooling of the Zen might be to mount a fan externally on the rear of the PSU, is this at all possible?

cloneman · Post by **cloneman** » Fri Apr 20, 2007 1:13 am

Yes, removing that fan would raise my cpu temp, but that's are not a concern.

I'm thinking of just putting the zen (at least partially) outside the case like Bluefront suggested - I would need an extension cable for the motherboard connectors though... do they even make extensions for P4 connectors?

amjedm · Post by **amjedm** » Fri Apr 20, 2007 2:44 am

How about attaching a 80/90mm fan with Blu Tack to the Zen - the side where the cables come out from? If you undervolt it shouldn't make much difference to the noise?

Bluefront · Post by **Bluefront** » Fri Apr 20, 2007 2:47 am

Mounting an external fan on the Zen, defeats the whole fanless PSU thing. Your case is negative pressure only, from what I can see. This pulls air through the Zen, back into the case. I have never tried such a thing with my Zen systems (I have three including the newer 400w model). My cases are all positive pressure, with the case airflow going through and over the top of the Zens, out the rear. The 300w zen has a rather small opening at the rear.....my mounting the PSU partially outside the case, gives more openings for exhaust (or intake) on the outside, and definitely makes for a cool-running Zen. You surely can make or buy extensions for any cables from a PSU.......I usually just splice wires if any of mine are too short.

Gojira-X · Post by **Gojira-X** » Fri Apr 20, 2007 2:57 am

Cardboard on the underside surface of the Zen should stop the airflow short circuit. This would mean that the rear case fan pull airs through the Ninja (ducting may help).

You could then mount the fan formerly used for the CPU near the Zen, adding valuable cooling.

EV10 · Post by **EV10** » Fri Apr 20, 2007 2:05 pm

Bluefront wrote:Mounting an external fan on the Zen, defeats the whole fanless PSU thing.

I don't think so... When you mount an external fan, you can choose any you like. You can take a 120mm 800rpm fan and set it on 5V to spin at very low speed, completely noiseless unless you really try to hear it, yet creating some controlled airflow. Modding a normal PSU this way is bad for it, but it's just the kind of help a fanless one needs.

Bluefront · Post by **Bluefront** » Fri Apr 20, 2007 3:07 pm

None-the-less........the Zen was designed to run fanless, and it can be run fanless. But if you want it to run both fanless and cool, you are going to have to start thinking differently about case airflow. This sort of PSU is made for a positive pressurized case, even though it will run in negative pressure. IMHO, the Zen will run ok just sitting outside a case with no fan, and this is a better solution than sucking through the Zen, pulling heat back into the case.

EV10 · Post by **EV10** » Sat Apr 21, 2007 3:45 am

Is there difference between air running left to right or right to left? In theory, a little, but practically it's important that it runs.

Though I find the whole negative pressure difference only case a bit undesirable idea. First, heat actually makes air expand (and not negligibly), which helps to increase the pressure. Second, forced intake helps to cool all the components, as opposed to pure exhaust fans, considering the CPU fan airflow.

IsaacKuo · Post by **IsaacKuo** » Sat Apr 21, 2007 4:14 am

EV10, I'm afraid you don't the true effect of hot air expanding in an open chamber. It can indeed affect pressure--but ironically it can actually slightly decrease pressure in some places. It has to do with the "stack effect" used by chimneys. Hot air expands, and so hot air weighs less. As a result, hot air in a tall column can weigh less by enough to reduce the pressure at the bottom. This negative pressure sucks air into the base of the chimney.

In practice, the stack effect is very weak for something like a computer. Nevertheless, it's possible to fanlessly cool a computer using a negative pressure stack effect case.

As an engineering matter, it's actually much simpler to design a negative pressure case where air intakes cool all components. This is why most home electronics and game consoles use exhaust fans rather than intake fans. You simply place an exhaust fan anywhere you want, and then every other opening becomes an intake. The engineers then merely need to place any hot components next to an air opening.

With a positive pressure case, the engineering isn't quite so simple. Airflow from a fan is complex to model. A fan "pushing" air pushes air very directionally, whereas it will "suck" air uniformly from every intake.

The basic upshot for this particular case is that a single exhaust fan WILL cool all components. Sure, air will enter the case via the Zen PSU--but air will also enter the case via all other intake openings.

There are advantages and disadvantages to both negative pressure and positive pressure approaches. I personally use mostly positive pressure cases, but I don't shy away from a negative pressure approach when it's better for the situation.
---------------

Oh--my advice for this case is to try using a small partition between the PSU and CPU heatsink to force air through the PSU rather than "short circuiting" via the green airflow lines. I know that's what you desire, but this short circuiting air will actually be bypassing the PSU components that aren't in the rear. This partition need not extend all the way to the front of the PSU. Anyway, it might help. It might not. It's worth a try, IMHO.

jaganath · Post by **jaganath** » Sat Apr 21, 2007 4:35 am

heat actually makes air expand (and not negligibly), which helps to increase the pressure.

Really? even at the small 10-20C delta usually found in PC cases?

BrianE · Post by **BrianE** » Sat Apr 21, 2007 10:11 am

IsaacKuo wrote: In practice, the stack effect is very weak for something like a computer. Nevertheless, it's possible to fanlessly cool a computer using a negative pressure stack effect case.

I might have missed something at some point, but did you finally get that "chimney" project to work?

(Sorry for the OT.)

IsaacKuo · Post by **IsaacKuo** » Sat Apr 21, 2007 10:55 am

I abandoned my "chimney" project because the stack effect was far too weak given the low amount of heat my old Pentium 3 hardware was making. Also, I've sort of set up a tradition of single fan builds.