12VDC cooling fan 9

[windell747]

Hi, I've designed a pressure bottle that houses some electronics for an underwater robot. Although I predict that the electronics will remain sufficiently cool without much forced convection, I would like to install a 4" 12VDC computer fan inside the bottle to circulate the air. The fan would pretty much run all the time the DC to DC converter would be in use so I won't need to switch the fan on and off. I am tempted to just connect the fan across the 12V output of the DC to DC converter. However, I would like to know if there is a better way than just connecting the fan to the output (i.e. putting a capacitor across the output). I want to remain simple, but i would also like to learn a little too since i'm an ME.

Thank you in advance for the suggestions!
windell

 

[VEBill]

A 12-volt should hook-up directly to the 12-volt supply. If you don't need that much air flow, you can run most such fans at a much lower voltage to reduce power consumption (if that is an issue). Capacitor probably not required, but you could check the PS with a 'scope to make sure.

 

[MacGyverS2000]

Depending upon the size of the enclosure (i.e., the amount of free air space), including a fan may do nothing more than increase power draw (and therefore more heat) with a negative benefit to cooling.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[VEBill]

The fan in my home theatre cabinet is a 12-volt fan running at about 3 volts - just enough to exhaust the hot air from the several satellite TV receivers. The current consumption is (I believe) much less than 100 mA. Power is thus a very small fraction of 1 watt. (For my application the reason for the reduced voltage was to reduce acoustic noise to basically nothing.)

I have seen fans used for similar purposes as windell's in military gear. Internal fans moving the air around.

Too bad the ocean full of cool water couldn't have been used more directly for cooling; but I understand the difficulties in trying to make such a design practical in the face of overriding considerations to deal with ultra high water pressure and ease of maintainance.

 

[itsmoked]

I'm surprised you don't do the common thing of filling the housing with oil which negates 99.999% of the leak problems and then the oil easily/automatically/effectively conducts the heat to the enclosure.

I totally agree with macgyvers2000 comment. In a small closed space a fan moves almost no air but rather becomes a static heater.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[ScottyUK]

If you are using a DC/DC converter can you run the fan directly from the incoming supply? It would reduce the losses in the converter if it doesn't have to run the fan.

----------------------------------

Sometimes I only open my mouth to swap feet...

 

[BobbyNewmark]

Sealed chambers tend to make for condensation problems. If a sudden temperature change makes the electronics the coldest thing in the box, that is where the water will condense. Dessicant bags are your friends.

 

[IRstuff]

A fan CAN work, if there's a clear path for the air to circulate. A circulation fan can substantially reduce the thermal resistance.

However, you might simply use passive heatpipes directly to the outer shell. They can draw large amounts of heat out, without the extra heat load of the fan.

TTFN

FAQ731-376

 

[OperaHouse]

The fan could be electrically isolated with 10 ohm resistors and a 220 to 470uF capacitor. Most fans today are brushless and require a minimum of 8 or 9V to operate at about 0.1A of current to operate the electronics. This might reduce some pulsing on the output of the power supply if your electronics is really sensitive. Doubt that would be a problem.

 

[windell747]

Great guys! Thank you for each of your valueable suggestions! Each of them adds a new facet or a new way to looking at my problem! I really appreciate the time that each of you have spent answering my questions.

I understand that the capacitor would prevent voltage variations across the dc to dc converter output, but what purpose does the 10ohm resistor play? Is it just to drop the voltage?

Also, would it be a good idea to put a diode in series with the input to the fan to prevent back current?

Lastly, for reference and documentation do you know if there is a typical application note or schematic that illustrates the wiring (yet very simple) for connecting the fan with the resistor and capacitor?

Thanks!

 

[logbook]

>Also, would it be a good idea to put a diode in series with the input to the fan to prevent back current?

No. There is no "back current".

The 10 ohm resistor just reduces the voltage to the fan and therefore reduces power consumption. The recirculating fan works well if the convection cooling dominates the conductive cooling. For example, if you have used small heatsinks on the pcb there might be indequate cooling on those components without a fan. The fan reduces hot spots (if they exist).

Fans draw very "unpleasant" spikey current waveforms. The capacitor reduces the current spikes from the power rail.

If the DC-DC converter is 70% efficient and your fan is a 1W load, the battery has to supply 1.43W. Therefore you put the fan on the battery and save the 0.43W. Also the DC-DC converter will have worse output ripple when loaded by a fan, even with a large smoothing capacitor.

 

[automatic2]

If heat is a problem, creating more is more of a problem. Your fan will add to the heat load not only from itself, but from working the power supply that much more. Adding resistors compound the issue. The very best way is to transfer the heat to the exterior media, likey byway of metallic conductors.

The only problem you could run into in adding a fan,is that the wattage of your supply may be tapped to the point where ripple increases to the detriment of the circuits. Additional capacitance may be required to mitigate this condition.

 

[MacGyverS2000]

Windell, another possibility might be heat spreaders/heatpipes, as suggested by IR... check out the possibilities from Celsia. They have amazing prices, do custom work, and it may make spreading heat to the outside cold sea an easier task. Completely passive, so no extra heat added, no power necessary, and with the right heatsink/fin design will possibly net you better cooling performance than a fan, even if it stays completely internal.


Dan - Owner

http://www.Hi-TecDesigns.com

 

[itsmoked]

How deep is this system going?

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[peebee]

Heat pipes are really cool, and you might even be able to get a free sample from a friendly vendro.

I suggest trying it as a coffee stirrer.

 

[windell747]

To answer Keith's question, the depth of the current generation of the system is 100 meters. A future generation is planned for 6000 meters so im sure there will be some major design changes.

 

[PeacefulKancer]

If this is an underwater robot as you listed, there is not going to be a lot of room for air to actually/physically circulate, is it? I don't know your design, but it may not even be beneficial.