Fan cooled chassis in an Equipto enclosed cabinet 2

[TJK1]

We are planning on having a large, fan cooled, chassis mounted in an Equipto enclosed cabinet. The manufacturer of the chassis prefers that it is mounted in a cabinet that does not have a back on it, for cooling reasons. But we need to keep the back on the cabinet for safety reasons. As a result, we are not sure if the circuit boards in the chassis will become too hot. We have the chassis running in the lab in an open rack, but we do not have a cabinet to try it in. The cabinet will have some louvers on all sides. I am not a thermal guru so I would appreciate communicating with anyone who could provide free advice on how we could determine analytically, or perhaps by a combination of calculations and some measurements if the temperature of the chassis will become too hot.

 

[EdStainless]

Companies manufature air conditioners meant for cooling electronics enclosures. They keep things cool, dry and clean inside. The other option, if you don't need things kept below room temp, is an air-to-air heat exchanger.
We have a lot of equipment that runs in such enclosures.
Run some searches, the only name that I remember is Lytron, but there are a bunch.

= = = = = = = = = = = = = = = = = = = =
Corrosion never sleeps, but it can be managed.

http://www.trenttube.com/Trent/tech_form.htm

 

[tkordyban]

Equipto will be very happy to sell you a door with a vent hole pattern cut into it to allow air flow. If you choose your hole pattern properly you should be able to satisfy your safety concerns. Holes of diameter of 0.25 inch will let air pass and keep prying fingers out.

 

[TJK1]

Tkordyban, the Equipto Cabinet we will be using already has louvers (narrow horizontal openings) in both sides and in both the front and rear doors. What I really need to know is how to determine whether or not they are adequate to prevent the heat from building up in the cabinet.

 

[MintJulep]

Direct test is easy.

Put thermocouples on critical components.

Put complete enclosure in a chamber held at highest expected ambient temperature.

Turn equipment on.

Wait for equilibrium condition.

Read the temperatures.

Can also do this at whatever (non worst case) ambient, and extrapolate.

There are a handful of analysis packages specifically for thermal analysis of electronic enclosures, and I'm sure that there are consultants willing to model your equipment.

 

[ko99]

For this, testing is a lot easier than a CFD model. In the meantime, here's a way to estimate the effect of the rear door:

1. Measure the system flow rate without the door. Construct a simple hood at the chassis inlet or exhaust, measure velocity profile with h.w. anemometer, average over hood area to get CFM.

2. Estimate the operating point of chassis fan(s). A single fan is easy: find the pressure on the fan curve at your measured CFM. For multiple fans use the fan laws for series & parallel fans.

3. Estimate the pressure drop through the louvers. Idelchek is a good source. He won't have your exact louver pattern, but use your judgement. Also use your judgement to estimate the effective exhaust area (estimate the air spreading between chassis and door).

4. Go back to the fan curve and add the door pressure to the operating point. In many cases, it will be pretty small and the door won't hurt the ICs significantly. If you don't trust your estimates, try it again, only be more conservative.

5. If the door pressure reduces the fan CFM significantly, there are ways to estimate the effect on IC die temperature. Let us know how it goes and I (or someone on this forum) can help.

ko (

http://www.ecooling.biz)

 

[TJK1]

ko99, in para. 2 above you say estimate operating point of chassis fan(s). Please define what operating point is. Also, in para. 3, you say I should refer to Idelchek for pressure drop thru louvers. Where can I get Idelchek info. When I did a Web search on Idelchek, I just got the names of some companies that apparently use Idelchek methods for some of their work.