If you didn’t use the saddle valve to connect to the house water, then you’ll need to insert another valve in the tubing between the house and the solenoid intake port. Do not put a valve on the outlet side of the saddle valve. You must control the flow from the inlet side of the solenoid. Say a valve like this.

With the Zener in place, the 470R + the 0.47uF together have to drop the remainder of the mains voltage (approx 110V). If I remember right, most of that drop is actually in the reactance of the cap.

Thus with everything in the neutral, VCC = HOT = Triac gate voltage. It might actually work putting the 470R + 0.47uF in the HOT, but that means VCC is going to be ~110 volts below the triac gate, so that ATTINY pin driving the triac gate would see a +110 volt potential. Very little current (hopefully), but still a big potential.

Actually had one guy respond saying he’d done that (run 110v potential on the ATTINY pin). It’s way, way out of spec, but appeared to work ok.

Now… you could just use a triac with an opto-isolator to eliminate that voltage differential problem, but then what’s the fun?

If you want more LED’s I’d just add another independent grouping of 3 pins. Going to a 4 pin charlieplexing leaves too much dead time on each LED.

As for the contents… Never really got a good configuration. Latest setup I did was with small SMD LED’s hand soldered onto the wires and then dipped in white glue. The blob of white glue does a pretty good job of diffusing the LED glow as well as giving a bit more mechanical strength to the connection. The big LED’s and thin wires didn’t survive the “let your 3 year old play with it test”, but the SMD + glue so far has held up pretty well.

I’m just starting out with the Arduino, but saw your firefly project and thought it offered a nice variation on the many charlieplexing guides out there.

I’m using a vanilla Uno board though, so I’m having a bit of trouble understanding the pinout from the tiny you used and I get an error so I think I’m missing some library functions you call.

Any help gratefully received.

PS Do you plan to obscure the contents of the jar in any way? as I think just seeing a moving bunch of glows would be prettier than obviously being a bunch of LEDs

The ~9VDC coming in is regulated down to 5VDC on the controller board. The sensor sends back an analog signal proportional to the supply voltage at the sensor. The higher the analog signal voltage, the closer the object.

The fact that your Cat Faucet works, but only measures 3-4″ says to me the sensor is still working, but… either the supply voltage it is seeing is low, or analog distance signal is having problems getting back to the controller board.

So, at the controller board, the sensor should connect with a Green (signal), a Red (5Vdc), and a Black (Ground) wire. Do you have 5Vdc between the Red/Black at the controller board?

If you do, then make sure that 5Vdc is making it out to the sensor. Take the cover off the back on the sensor, the circuit board is a T shape. With the cover off, looking at the back of the circuit board, with the leg of the T facing ‘up’, the pins from left to right are: Signal, Ground, and then Power. (http://www.jenrathbun.com/Electronics/images/SensorSpec.jpg) So you should have 5Vdc there between Ground and Power.

If you’ve got 5Vdc at the sensor, then for some reason, the signal isn’t making it back to the controller. Unplug the green wire from the connector on the controller board. Place something 3-4″ from the sensor face and read the signal level at the sensor. Then measure the signal level again at the controller end (where you unplugged the green wire). It should be the same voltage level. Then finally plug in the green wire to the controller board and measure the signal again. it should stay about the same voltage level. (read the signal level on pin 20 of the AVR processor when the green wire is connected to the controller board).

Certainly the 3-4″ distance isn’t right. I can think of several things that might cause this problem. First, check the control box for water or past evidence of water. Corrosion on the circuit board(s) in the control box (especially on the bottom of the circuit board) could be a cause of this behavior. Second, check the connector where the sensor connects to the circuit board. If the cable to the sensor got accidentally stressed, I’ve seen it break some of the wires inside the cable right at the connection point which can also cause the short distance issue. Third, it is possible that water got inside the sensor itself, You might take the two screws off the back there and check for any water/corrosion damage.

This should help narrow down what needs to be replaced, ie either the sensor or the control board.

LED flash rate suggests that sensor is slow to react. Must place hand 3~4 inches away before LED begins more rapid flash. When hand is removed the the LED doesn’t slow back down until about 15 sec later.