So I've mentioned a project previously, in which I'm controlling the opening and closing of a number of cheap water valves. I am not the best at analog circuitry, but it's been fun learning along the way.
My inspiration started with this video, but I'll be building it a bit smaller due to costs. I also have some "interactive" ideas to build on top of it.
I'm making a smaller "solenoid latch" board, which is an I2C slave, an 8-bit latch, 8 hexfets, bridging between 12V and 5V power levels. Each hexfet switches one solenoid of the "hold current to open" variety.
At this point, I've got the switching done, and am testing an implementation of the I2C slave now. The PCF8574 seems cheap enough and straightforward for my purposes. If I have to, I'll add a second layer of latching so I can tell all my boards to invoke their next set of switches independently of loading the data.
My goal is to drive the whole affair from one or two Arduinos. One will manage reading images from files on an SD card and pouring the images. A second one will drive all the interactivity I have in mind.
Comments welcome, I could use any sort of good ideas.
How much current do you need to drive each solenoid? Are you sure you can't find
a single I2C or SPI latch with high current outputs? Some of the 6x595 octal relay
drivers can sink 300-400mA and have a voltage ratings of 30V or more (See TI and Allegro). IIRC TI and Allegro also have a number of other drivers that can do higher currents.
I'd like to keep price below $15 each. I want the DC at 12V each or less, so I am able to simplify the power supply issues as much as possible. It needs to be fast enough for bit-banging visual patterns in the water flow. It needs to be small enough I can rack them together without too much extra tubing work. Of course, the pulse-on, pulse-off types are attractive from a power usage standpoint, but I think I can support pulse or hold types just with changes to the software.
There is a difference between sinking 400mA on one pin, and sinking 400mA on every pin of a device. Many will just pop due to heat dissipation if you try to do that. And my current experimental solenoids are 600mA hold types, anyway.
I was afraid of that. I don't think you'll get the sort of short on-off performance you'll need for that sort of application out of normal irrigation valves. They tend to take multiple seconds to turn on and off; the solenoid controls a very small flow of water which changes the pressure on a larger diaphragm, and THAT switched the main flow. I don't know where you'd get the high-speed water valves of the sort you'd need, and I rather suspect that high power draw is going to be part of their nature, since (I think) you'll have to have electrical force completely overcome water force.
