[TowerTalk] Ladderline - what are the facts??/

[jimlux]

Kevin Normoyle wrote:
> Looking at this document "Heat Dissipation in Electrical Enclosures" and 
> a nice graph on heat dissipation vs total surface area for airtight 
> enclosures,
> it would seem to me that:
> 
> 1) assuming .1 db loss minimally in any balun design makes sense when 
> designing an enclosure. That's roughly 35 watts at 1500 watts transmit.
> (I would even want to be able to tolerate twice that, designwise)


If you wanted to test your design, a 60 or 100W lightbulb makes a decent 
test source.

> 
> 2) 35 watts would need probably 2 sq ft of surface area for heat 
> dissipation, to prevent the internal temperature from rising more than 
> 100 deg F over ambient. Assuming airtight enclosure. Radiant heat transfer.

ACtually, you'd probably get significant convective transfer.  If you 
could fill the enclosure with oil or pot it in something thermally 
conductive (alumina loaded epoxy is common) you'd be better off.

> 
> 3) Balun enclosures are typically .5 sq ft of surface area or less. So 
> they don't dissipate the heat, leading to thermal runaway. Which is why 
> there's so much anecdotal information about failures.

Agreed.  Thermal design doesn't feature into most amateur designed 
products, because the "worst case" actually doesn't occur very often.


> 
> Fixing that:
> Passive air flow would seem sufficient. But it means that Baluns need to 
> be vented.

Venting isn't a panacea.. aside from issues with vents plugging up, you 
need to have significant thermal differential to get decent flow.

> 
> I've seen a lot of discussion about just how much loss there is, but it 
> seems to me that the real issue is correct enclosure design.
> 
> Can we agree that the real problem is that QRO Baluns require venting, 
> rather than airtight enclosures?


More properly, anything designed to handle kilowatts needs some 
attention to thermal design.

> 
> I thought the info that unpainted aluminum is bad for radiant transfer 
> is interesting. I may paint my DXE enclosure white and drill a hole 
> pattern for venting.

White isn't the best color for radiation: black is.  But, then, what you 
really care about is the emissivity of the coating at the blackbody 
temperature of the device, and what it looks like in the visible 
spectrum might not have much to do with the emissivity in the far IR.  I 
note that thermal radiators on spacecraft (where there is no convective 
heat transfer) are pretty shiny silver teflon.


>