I've took the basic shutters application, and made it run based to temperature. The project is that I want to heat water up to 70 celcius as efficiently as possible, using solar panels. So, 0-100 celcius is the range. As you can see, the shutter has 3 points where I want it to stop. Fully open, in the middle and fully closed. Since this is a forced circulation system, I took almost all the measurements at the circulator, as I think this is proper and safe enough for the circulator.

My question is, I've managed to use it as desired most of the time. The problem is, that in the 55-67 temp section when temp is falling, If I connect a Nand(edge) and that to the open position, this operates as supposed  and opens the shutter. The problem is, that it does the same when the temperature is rising (exceeds 67 temps) and this is a problem. How can I make it recognise that when temperature is rising I do not want it to operate?

[snip]

I did it your way and simulated a power line failure at 73c and one at 30c. Now it doesn't decide what to do. up or down? I added retentivity and made it worst. I must add, without affecting the other 2.

The problem with this setup is that it changes value at 67 once again and the NAND edge sees it as a change once again. Basically now it affects everything above 67 and below 55

Above 70 roll down until middle I2 or lowest I3.

Between 55 and 67 in the middle I2

Below 51(45) roll up until the top I1

I have succeeded,

Above 70 roll down until the lowest I3

Between 55 and 67 and when the temp is rising, roll down from the top to the middle. I2

Bellow 55 roll up until the top I1

Also power line failure

at 70 and above it goes down to I3

Between 55 and 67 it goes down to I3

Below 51(45) roll up until the top I1

I4 exists because if they @ up I must reset it somehow without them interfering

Q1 is activation of the pump but I have no problem there. 

Maybe I shouldn't be too greedy I think

I did a simpler thing I added all the limit switches to an OR so they have equal rights except I energised the middle one only when the temp is between 54 and 68 So it doesn't interfere with the 70 and 51 (45)  temps

Also if you look closely, what you propose above is almost exactly what I did

I used 

up-down_pushbutton_control_of_shutters_v8

example as guide

Your point is valid.

Above 70 should go to the bottom and that is it.

I think I understand what you want to do, that is to add timing restrictions to this and forget the limit switches.

I though it too, but, I am sceptical. If you add a time constant in a variable system, I mean the sun alone, (not  minding the water temp fluctuations) then where should it be your sweet spot? In the high intensity, or in the low intensity? And what happens to the other? 

My thoughts were

Below 55 you want full exposure because you want hot water

Between 55 and 67 things get serious so you better be cautious and set it to the middle

Above 67 you're gonna melt the flowmeter, the pump will freeze and all the cool stuff ! so you want the least exposure possible. To the bottom.

Between 55 and 67....

Below 55 ...

These are my thoughts

If you can do it better than me, by all means and I will fit the rest

I Just want to make a presentation on what I presume is almost the final edition.

I want to make an automatic shade for my solar panels to avoid overheating of the forced circulating solar heater. (This arose because I overestimated the power of the sun and basically I damaged the insulation of the boiler. Also because I want to put the boiler under the solar panels, to upgrade the insulation for the winter)

And control the pump, secondary.

So the main temp is circulator temp (AI1), as it must be the cooler in the system, and AI2 in the collector, (before the boiler) works as a comparison , obviously you don't want to cool your medium. This is situated in the middle left.

 Actuation Control for the pump is situated in the top middle. Timing things have been added because if the circulator stops you must operate it once in a while, to take measurments. Also not in the night.

In the top left are the redundant logo controls, operate I4 and it goes into manual.

The program works this way

Situated in the bottom left I1, I2, I3 are limit switches, activated when the shutter passes over them. Obviously, when all the way Down (I3) Both I1 (top) and I2(middle) are active, and this did not work, so I've put time restrictions. The limit switches are active by default of the machines, so the on position is actually the off. There is one trigger active b010 in I2 when temp is 53-68c because I did not want it to stop in the middle, during a power failure, but rize all the way to I1 top, and generaly this trigger if applicable must be active before

situated in the middle right, circulator average value b062 kicks in, takes a reading and lowers it in the bottom (I3),if over 70 (trigger b055) and middle (I2) if 55-67 (trigger b039). Trigger b052, if applicable obviously propels it to I1 etc. But I wasn't content at just that and put two mathematic instructions (derivatives) situated in the middle, just to make sure that if temp fluctuates more than enough I delete or reinforce middle I2 or b039 trigger because I want formula 1 realm performance. 

In the bottom centre are information and failure messages. In bottom right some makeshift timing restrictions because I wasn't content with the factory delays.

In the top right is the activation circuit of the shutter. Explains itshelf.

Testing

Obviously the time sampling of b062 average value is the most crusial factor for testing, I like 15 seconds followed by 10 seconds for the b007 average value of I2(active when 53-68). Obviously speed of the shutter is instantaneous.

You position your eyes on the result situated in top right Activation circuit.

Simulation.

You operate only I1 I2 I3 where are the limit switches for the shutter and basically AI1 circulator temp.  Give it hell and winter.

This is hopefully the final version of the above. It is way simpler since I found some limit switches that are now normally open. The main difference in testing is that  sensor I1 top is depressed when I2 is pressed and so on. All the rest run the same Time order is not to be changed, nor is the duration. This, compared to the previous one, is child's play.

Also I added redundant circuits in case something fails and an interface.