My second attemp was to move to micro valves. I will need normal off valves since the final apperture should be stable also when not supplying a voltage. I milled again channels (three at the same time, middle block) and put a silicon sheet in between (left). The top acrylic I had 3 inlets and 3 outlets. In the middle three inlets to control the valves.

The top plate I countersinked the control port of the valve a bit that the silicon could be sucked in.

As basic setup I used an Arduino Nano controlling a mini pressure pump (in vacuum mode) and a pneumatic valve (not shown in picture) and try to pump some water. The setup shown was one of some different tires I had.

Results; In general it worked .... a bit. The main problem were:

• depending on how tight I screwed the screws the valve was not tight enough and the varuum sucked the water all the time or the valve was to difficult to open, that I even could not operate it with mouth sucktion.
• In the beginning I used 3D printed hose connecters (i.e. the blue one at the water reservoir) but they had not been as tight as well as they broke when I had to remove the hose.
• I tried 6 different valve design / dimensions and countersinks. I also ordered different silicon material as well as one set of liquid silicone.
• The athmosperic pressure sensor range was way to small (or my pump to strong :-)). I tried to order another one without going in the costly range.

I think I need some more inspiration how to make it work. It should be possible this way, but I did not really understood the effect chains and how the design parameter influence the behavior exactly.