As you might know, I am slightly obsessed with autodosing. I have tried more designs than I care to remember. Sometimes an idea gets stuck in my head, and a few days later you find me going through scrap parts trying to assemble yet another monstrosity.
Some designs have worked out really well. Dosing N and K via the waterpump method, and P via the test tube method is extremely simple and reliable. There have been a lot of attempts for micro dosing that for different reasons did not work out long term. Dosing diluted solutions lead to undesired growth in the containers. Timer driven syringe dosers did not last. Peristaltic pumps are expensive, and one of the two I have tends to strip the plastic rotor and stops working.
There was another design
which seemed to be promising. I spent many, many hours trying to make it work and never could. What would happen was that either the "refill hole" was too small and the tubing would not refill, or it was too large, which lead to little droplets entering when all the solution was pushed out, and when the pump stopped, the droplets would again prevent the tubing from refilling.
Still, this seemed like a real smart idea, so I kept playing around with it, and I think finally found a bit different design which seems to work better. The trick to making this work is to have a restriction ("small hole") while the pump is pushing the fertilizer out (which keeps the tubing free of more solution), and having open flow when the pump turns off, so the tubing refills readily.
After many hours of thinking I came up with a somewhat redneck, but simple and cheap solenoid valve. I used a piece of latex glove, wrapped around the end of some air tubing with a rubber band. That went into a plastic syringe. As the air pump pumps, some air is diverted to blow up the latex bubble, which blocks the flow through the syringe, therefore restricting the flow and keeping the tubing clean.
Okay -- here is the design: