I have a test valve setup running with two Rainbird CP075 valves and two 3D printed passive valves, the PEEP I designed and a PANDA PEEP I printed here in PETG. A BenchDuino prototype board with an Arduino Mega 2560, display and lots of switches controls a Keyes 2-channel 5 V relay board that controls the Rainbird valves. A BME280 pressure sensor is used for monitoring. An old heavy duty 13.8 v linear power supply powers the BenchDuino and valves. Air is supplied by a compressor next to the setup with a 3' hose.
There are flow restricters between the Rainbird valves and the PEEP valves and on the outlet of the PEEP valves.
A simple program cycles the valves and monitors the P sensor. I will add to the program later but I wanted to get the valves cycling. Now I'm running with a 1 second fill and 1 second exhaust.
The pressure sensor is mounted in a cap in a T with the wires brought out through the cap. The first two attempts at sealing the wires with silicon and latex caulk both failed at around lower than spec pressure.? My third attempt is using JB weld and I'll install it later today.
Initial testing was run with the initial inlet pressure spec of 50 psi. The compressor could not keep up with this and the P sensor seal blew out. Subsequently the spec was reduced to 20 psi and even that has been hard for the compressor to keep.
I have about 16,000 cycles as of this minute.
First observations:
The Rainbird valves sound like airhorns when they open at great than about 10 psi. You couldn't possible sleep with them in a room.
Sealing the P sensor wires is much tougher than I thought. I t have high hopes for the JB weld approach.
It will take some time to size and balance the two the flow restricters to reduce the load on the compressor yet provide enough flow and back pressure to open the PEEP valves.
At low pressures the compressor pressure regulator valve is highly dependent on the tank pressure. Unfortunately, as tank pressure drops the regulator open so the viscous cycle drains the tank.
Bob?? KD8CGH
