Last Modified: 2014-07-11
When working with pressurized systems, it is often desirable to have a security group to prevent the system to undergo an over-pressure issue. Such an element can also be used to fix the working pressure inside a tank by continuously injecting gas into the tank and letting the excess pressure go out by the security group.
I propose here a method for the construction of an overpressure valve whose threshold is set electrically. The overall concept is to place a powerful neodymium magnet inside a variable magnetic field generated by a coil and to use the force acting on the magnet to close a hole from which air flows out of the pressurized tank. Technically speaking, the system can as well be used for negative pressure (vacuum) systems by pulling the magnet instead of pushing it. As I am mostly interested here in positive pressure system these days, I will focus on the push mode. Also, I will assume here that the air can be sent to the atmosphere and do not require any treatment.
A prototype assembly is given on Figure 1. A more advanced version for 3D Printer owners is given at the end of the document. A coil made from more than 1000 turns of 0.35 mm is wrapped around a plastic case and a 6x10mm neodymium magnet with a .177" shooting ammo steel ball is placed inside it. The flexible is conventional 4/6 mm silicone tubing. The steel ball should be slightly larger than the internal diameter of the tubing to have a good closure.
When no current flows into the coil, the magnetic field is null and the magnet make no obstruction to the air which freely leaves the case by a small hole drilled on the side. The pressure inside the tank is almost the atmospheric pressure as the pressure drop is extremely small unless the compressor has a very large flow rate. After some threshold current i0, the force acting on the magnet is sufficient to overcome friction and the magnet moves toward the flow path and make obstruction to it. The coil pushes the magnet with a force F(i) but, on the other side, the air pushes on the magnet by the effect of the positive pressure inside the tank. As long as the pressure force will be bigger than the magnetic force, air will escape by the hole on the side of the case but, when the pressure force gets below the magnetic force, the valve closes and pressure can start accumulating again until a steady state is reached.
Actually, things are a little bit more complicated because it does not act as a whole-or-nothing system: as the steel ball starts to obstruct the air path (but before it is firmly pressed onto the flexible), the friction of the air on the various surface begins to create a pressure drop which is log-proportional to the air flow rate. But this is not really important to get the device working. What is important is that we have some mean to tune the pressure inside a tank connected to an air compressor using an electrical current. This is usually a good thing because it allows process control and feedback operation by connecting a pressure sensor on the tank.
Finally, I have made some tests using a 200 mbar aquarium pump, the electrovalve and a manometer to monitor the pressure inside the tank. The current was increased by small steps and the pressures recorded. The recording was performed by first increasing and then decreasing the current to check for hysteresis patterns. The procedure was repeated twice and the results are given on Figure 2.
There is clearly a threshold current of about 50 mA required to put the magnet into motion. Beyond that threshold, the pattern is relatively linear which is good news because linear actuators make feedback easier. The only issue was that the steepness of the relationship seems to have changed from one cycle to the other. Also, it was difficult to have large pressure inside the tank without using high currents which made the coil heat up a little bit.
Apart from that, the system works really fine and is extremely easy to build. So, why don't you try too?
For those of you who have a 3D Printer you can download the 2 parts required for the model. Take care that models printed with the filament method are not airtight along the Z-axis. So you will have to put a lot of silicon onto it and use Teflon tape for the screw. First put some silicone on the cap and let it dry for 24 hours. Then, screw the cap into the frame and don't forget to apply a lot of Teflon tape on the threading. Finally, and depending on the resolution of your 3D printer, you may have to slightly reduce the XY scale of the cap by 1 or 2% to make it fit the casing.
You may also like: