Hi all. Ok, I will dive in here as well. I had a rather elaborate post in another thread, maybe you could make a link to it AC? I don't think I can insert links yet.
EDIT (Piwoslaw): Here is the link:
The Homemade Heat Pump Manifesto, post #808
Anyways, here is an explanation on what I said about floor heating being "self adjusting".
The effect of self regulating comes built in to the whole floor heating system. If you have a system that can supply a steady input temperature to your loops that is, and that the floor loops are not too long. Never push the lengths of the floor circuits. It is not obvious unless you think of it in the correct way. I will try to explain more in detail.
Let's use a normal concrete floor as an example. A floor like this will keep an almost fixed temperature if fed from an adjusted source. With adjusted source, I mean that there is an accumulator tank that has a higher temperature than the input at all times, and that the circulating water is mixed with a varying amount of water coming from the accumulator to make the feed water temperature very stable. Given that the input is very stable, and the loop length is not too long, the floor will stay very close to a constant temperature no matter how much heat it delivers.
As a theoretical basis, we will assume that the floor is constant temperature. Given this, we also know that the air temperature in a room with floor heating is quite close to the temperature of the floor, in this case we assume 20C floor temperature. Now, if the difference between floor and air is 1 degree when the outside temperature is 10C. Calculating the heat provided from the floor, you use this delta temperature in a formula that would look something like p = floor_area * heating_capacity_constant * thermal_difference. The first two does not change for a fixed room. So if the temperature drops to 0C, that would increase the thermal loss trough walls and ceiling and everything, reducing the air temperature, let's for the argument say that it is almost linear and that it now becomes 2 degrees difference from air to floor. What happes to the p above? It doubles. The floor will now release twice the amount of energy. Yes, the water leaving the floor will be ever so slightly colder, but the system adapts to that so that more heat is grabbed from the accumulator. This means that the temperature has dropped 1 degree in the room as a result of 10 degrees drop outside. If it drops another 10 degrees, there will be another 1 degree drop inside. Not a significant drop.
Comparing this to a system with 50degree radiators, a drop in temperature from 10 degrees to 0 degrees outside would demand the radiators to deliver twice the energy to keep the temperature up. The formula for delivered thermal energy would be about the same as above, but the difference would go from 30 degrees to 31 degrees. That would not double the P. Therefor you need to have each radiator or room thermally controlled via some regulator, which pretty much all radiators have. Heated floors does not need that, because of the huge area and low temperature difference between air and floor, an increase in temperature difference makes a huge impact on the amount of energy it provides. Many will try and sell you these wireless controllers that you should use for each room. That is imo a waste of money. If you can make the mixing temperature from the accumulator go up and down a degree or two based on the general indoor temperature, that should keep the whole house at a pretty steady temperature. This is why i say that it is self adjusting. It is this way by design. Now, some floors are not so efficient as a concrete floor. If there is a layer of wood or so above the tubes, the adjustment of the mixing temperature should certainly be done as the indoor temperature drop could be several degrees and change the temperature outside the comfort zone. Dispite this, the difference would be a lot smaller than an unregulated radiator system.
Regarding the spacing, this depends on a couple of things. Most floors here are made of wood. Solid, parquet (most people have this) or laminated. We have very few homes with a concrete floor, except for the basement that is, but usually we cover this with wood too. We are 13 persons per square kilometer, and we have loots of forrest, so pretty much every home is made of wood materials, from studs to floors and roofing. Top roofing is usually stone tiles of some sort, or dutch clay if you have the money for it. Anyway, as we usually have a wooden floor, and the beams are wood, and we usually have an under-floor of 22mm chipboard or similar, we lack the heat distributing layer that a concrete floor gives (we do use concrete on top of the 22mm chipboard if we put tiles, like in our bathrooms). So we use heat spread aluminum plates. The most popular brand here is uponor. There are two main dimensions used, 20mm and 17mm. The 20mm can be pulled for 120meter or so, while the 17mm should preferably not be much longer than 80m because of the reduced waterflow. Also, the heat spread plates are not as wide with 17mm, so that makes the area it can cover about half of 20mm. The bonus though, is that there will be less local difference. The thermal energy does not have to move so far from the tube to get to the center between two loops, making sure that you don't get the situation that your toes are warm and your heel is cold. Most installers use 17 wherever they can, but 20 in concrete. 6" equals to about 15cm. That seems to be close to the 20cm that our 17mm system uses. There is a gap of 10mm between the heat spread plates to avoid them making sounds, and the heat spreaders are 190mm. The 20mm is spaced about 30cm (12 inches approx I guess). There is a point that the efficiency is higher with a better heat spread, but it does not solely depend on spacing. How you spread the heat is also important. If you had a perfect heat spreader, you could easily get away with only one straight tube and a huge heat spreader, efficiency then should be phenomenal.
The product that you mentioned is also sold here, although it is not widely used. The reason for this is that placing something like that, you would have to protect it to avoid damage to the heat spreader while constructing, and that is unpractical. Placing heat spreaders manually goes quite quickly. You either get some flooring with channels cut into them for the tubes, or just make your own. Just ordinary pieces of wood paneling, cheapest sort should do. Fix them to the floor with just enough spacing that the heat spread plate with pex in it fits in the gap. Put some cheap fabric in there to avoid squeaking. Most squeaking in heated floors with this solution is the heat spreader moving up and down slightly. The rest is usually due to not gluing your chipboards, not using screws but rather nailing them down, and not putting a fabric between the chipboards and the beams the rest on. Was it Vlad that did this already? You can drop the aluminum foil, as the reflected IR will to a big extent be reflected back down by the aluminum heat spreaders. It is not a worth the effort, but more importantly, it creates a diffusion tight seal that might be a cause of condensation issues if the area below is humid. A heated floor should help the problem, but it is not ideal.
My best tip here is to use heat spreaders of some sort. Aluminum is cheap and has a low thermal resistance as long as it is not paper thin. The ones you refer to has thicker aluminum that usually used here for 20mm pipes (0.6mm thick), so they should do just fine.