Electrical thermal question

[Racecardriver16]

Its been a few years since I had to do any thermal analysis. I have acclimated myself with the software. However, I have several generic electrical components (self contained video encoders, solid state PC, splitters etc) that I have power usage for, but not idea how much of the power is rejected as heat.

Is there a way to estimate this? Or should I just do the analysis based on max power in as max heat out? Obviously things like relays, etc I wouldn't do that.

 

[MiketheEngineer]

Obviously some of the enrgy is used to do the things those things do. But if I were sizing an A/C or any cooling unit - I would probably just add them all and be done with it - and have slightly conservative system!!

Easiest and quickest way - but don't forget that this is from a Structural guy - who remains lazy and always conservative.

 

[MintJulep]

Obviously some of the enrgy is used to do the things those things do.

And when they do those things they get hot. Then that heat is rejected into the space.

Energy in = Energy out

 

[Racecardriver16]

My initial plan was to assume 100% of input power was converted to heat, but introduce known duty cycles to bring that number closer to reality.. unless I could come up with a solid estimate of actual power used to create heat for each component.

 

[IRstuff]

Relays burn power in only one state, at worst. If they're latching relays, then neither state consumes power. In the former case, then, you could assume that 50% are in one state, and 50% are in the other.

As for the other compoents, if they're on, they're consuming. There would be no obvious duty cycle to consider.

TTFN

FAQ731-376
Chinese prisoner wins Nobel Peace Prize

 

[Racecardriver16]

Yea, obviously a non latching relay consumes power on the coil side, but dissipates little energy in the form of heat from the load side.

It is a tracking/video system (PTZ), and when it is not in motion, only the video stuff should be making heat. The stepper motors are holding, but the controller and hardware put off a lot less heat than when moving. Start/Stop is probably highest current, then moving, then stationer. The system is well balanced (CG) so hold current should be low.

Anyway, so it sounds like there is little argument to starting my calculations based on 100% of power consumed is converted to heat for video switches/splitters, ethernet hubs, etc etc?

The analysis is only looking at each sub COT part, inside a ruggedized box. Not going to board level or anything. Its been tested on the bench, but they need analysis to back it up-> and I plan to add solar loading in my analysis. I do not have access to bench test data to verify my results. Stoopid, but political

 

[MintJulep]

Power consumed is usually not anywhere close to nameplate load.