Squirty Boy

O, I have a scrounged AC transformer I was hoping to power this with. It has 16 (?!)  ~12v secondary windings, which I had chained all up to give me around 30V to rectify and regulate for powering the Raspi. However it seems the over-all current supplied by these many taps is just barely insufficient to power a bare Raspi 4. I am able to measure the transformer output current to about 1.5A, but this is not quite enough. I can see on the scope where it dips as the RASPI tries to boot, sucks too much juice, and then shuts down for a second in a lamentable waltz for watts. I have tried disabling various components on the RASPI with no luck... So I am going to have to order a new transformer to power this box.

I am pleased with the ability of the diodes and beefy 3905 to convert 30Vac to a clean 5Vdc. The ripply DC hitting the regulator IC is ~40V!

I suspect the plethora of taps is reducing the total current of this transformer; The wires in the secondary windings are tiny, which I believe is what limits the available current of a transformer. If I was using something better suited, such as a 1:5 transformer with a single beefy secondary winding, I probably would not be having this problem! Sadly, sometimes recycling old parts to drive vastly over-powered microcomputers does not work.

Now to see where I screwed up in the design. Its obvious I should spend more time measuring things out in PCBnew to make sure everything matches the footprints in the datasheets. Even so, it looks like only minor changes will be necessary!

I spent another few hours fiddling with kicad and I think I have a workable PCB ready. I will upload the files when I remember where the appropriate place is.

I am going to sleep on it and then order a batch of boards and bits in the AM!

I have done a successful proof of concept for using a triac to switch the AC voltage for my valves! Now to design the PCB

10 more minutes of duckduckgo-fu has led me to believe I want to use Thyristor chips instead of MOSFETs. While it is a bummer that these are 100x+ more expensive, it seems to be much more appropriate than running two MOSFETS in reverse bias.

I will research this more tomorrow.

I have created a first sketch in KiCad of how I imagined the circuit would work, and uploaded it as a ZIP archive. First I added a power section which will take the 24VAC from a 120VAC transformer and then rectify filter and regulate to 5V to power the RasPi, then I added a MOSFET relay control section which will actuate the solenoids in the water valves. I need to learn more about how to control AC with MOSFETS, so if anyone has any good guides or circuit examples it would be greatly appreciated.