Channel Hopper said:
If the air can be kept circulating past any cooled component (taking into account the resultant rise in temperature by further cooling) then the chance of water precipitating out as ice is reduced.
The idea is to freeze the LNB and lose the heat elsewhere, if there was enough heat at the cooling site to actually effect condensation (and therefore freezing) then it would be really inefficient and you wouldn't want that either. The peltier suffers from that (hot right next to cold) and it is more difficult to rip the heat away and lose it with a remote heat exchanger without heating the air near the LNB to a degree. I think butchering a small freezer or fridge / ice box could be the most practical for this.
BTW.. The thread post 2
http://www.physicsforums.com/showpost.php?p=846711&postcount=2 describes the ‘use’ of freezing onto an element to trap vapour in another dehydration process
demonstrating what I have said about condensation/ freezing- that is it is that good at extracting moisture out of the air. The process that is the main thread demonstrates the same problems and the air flow to a dryer element. That is that ventilation (in open air) is not anything you would want to do (unlike ventilation in a house for example) but a drying system in an enclosed environment is as I said earlier, if there is no vapour in the air it will not result in ice build up (or condensation). It doesn’t say if they are using a cooler at the uplink or this is a passive system for general dry air/ minimal condensation. I have been involved in engineering/ chemical projects that, although not directly radio electronic cooling related, give you the experience of the kind of other problems that will arise trying to cool any object like that.
The worst issue may be condensation and ice in the feed horn/ wave guide and on the antennae (and down the connected coax). Until you try it you don’t know, the gains may out way the problems or have to be dealt with.
Lagging:
Dr Dish & RD100‘s.… cooling with freezer spray.. A far better test and result is to use lagging… when you use freezer spray to freeze a water pipe, you lag the pipe- you can’t do it without lagging - the lagging increases the freezing massively and retains it. If Dr Dish had lagged the LNB and pushed the freezer spray through a small hole instead, the temperature would have been far lower. The same goes for a proper setup and lagging will reduce condensation and freezing air bourn vapour as well as make it far more efficient.
Divibi
The noise reduction gains we are after are in the order of the equivalent of 10-20-30% (40-50% would be fantastic but doubtful and certainly not those massive gains you have mentioned- we wish)… say a 1m acting like a 1.2- 1.3m or a 1.2 like a 1.5 to be realistic (in actual signal to unwanted noise wise- not signal strength- that shouldn’t alter with this). From our point of view in the UK there is a planning limit of 1 x 1m & 1 x 60cm dish per property now, so maximising a 1m rather than just getting bigger ugly dishes is attractive for a lot of people as well as bigger enthusiast’s ground based dishes gaining. Bigger dishes also have their problems too, weight, wind effect increases and needing 36v and far stronger gear all round etc.. this can help avoid some of this so it might be worth a bit of trouble.