A general guide about using fans and examples of setting up the hardware needed to create a breathing fan apparatus.
Overview:
The goal of this project was to get some familiarity with a variety of fans, test their capabilities, and build a “breathing object” that mimics the rhythmic nature of a creature breathing.
Using Arduino, a power supply, and a darrington array we can change which fan is rotating and the direction of the air flow. By placing two fans back to back and placing a bag over one end, we are able to both push air into the bag and inflate as well as pulling air out of the bag to deflate it. With the Arduino looking for an analog value, this could later be attached to another application or control board for synchronization or choreography.
Research:
There are a variety of fan types based on different types of motors. The first question to answer is whether the project calls for an alternating current (AC) fan or a direct current (DC) fan.
AC examples:
For any sort of larger project or more required airflow AC fans will be a better solution since they are created for standard wall power. Standard window box fans are easy to come by, but have discrete settings (1, 2, 3, OFF) instead of a full range and must be powered on at full speed. Additionally, there are examples like blowers, inline duct exhaust fans, or inflatable fans which can move a large amount of air very quickly.
DC examples:
With a few exceptions, DC fans are going to be smaller, have less airflow, and are primarily meant for cooling computers. That being said, it is easier and cheaper to get an array of DC fans since they are designed in a compact form factor and simple wiring ideal for interfacing with any single board computer, in this case Arduino.
Both AC and DC fans are designed to spin one direction. Multiple design elements determine the direction of spin, including the coil windings for the motor in AC motors, any diode placement in DC motors, and the design of the fan blades themselves. For this prototype we overcome the limitations of direction by pairing fans together. By placing two fans in the opposite direction back to back, we can control which direction the air flows through both of them by turning either one on and off.
We ended up selecting three Arctic F14 140mm PC case fans and six Arctic F8 90mm PC case fans.
Version 1: Single set of fans
In this first version, we use an ULN 2803A Darrington Array IC and the Arduino to control a pair of the F14 Arctic fans. By varying the value of our PWM output, we can adjust the speed of the fans individually. It is worth noting that fans don’t start from 0 as they require a minimum start current to overcome the friction of the blades around the housing. For our computer fans, the minimum RPM starts at 800 (at 5 volts power) and goes up to 1400 when all 12volts are supplied. The fans also do not cut off immediately when turned off, but rather wind-down. Both of these are the results of the fans being moved by brushless motors. This also means that there isn’t a perfectly smooth behavior to the breathing in and out, but rather that there are momentary “hitches” when switching directions.
The wiring is shown above. We used 12V 2A power supply attached to the Arduino’s VCC pin, and ran the power directly into the positive side of both fans. The ground side of each fan was placed on a pin of the Darrington Array. The nice thing about the 2803A is that it does allow for analog values. By utilzing Arduino pins 10 and 11 which allow for PWM, the Arduino triggers the 2803’s internal transistors to let the power flow in analog increments, from about 40% up to 100% (5V to 12V sent to the fan). Each fan has a maximum 0.16Amps, which is well within the range for the Darrington array.
While fabric was attempt as a material, the weight of it resisted the range and fluidity of movement which we hoped for in the project. A plastic bag worked perfectly, but for a larger scale something like thin plastic sheeting would be best as it can be formed into any shape and heat welded together.
Version 2: Set of three fans
The second version used the six F9 Arctic fans in pairs of two. One bag was then cut and placed over the end of the whole group of fans. The wiring for this experiment was the same as for the first prototype, but used 6 pins instead of two to accommodate all of the fans. All three fans facing the same direction were turned on simultaneously, while the three opposite fans were left off. While a large bag worked well enough, it would be more interesting and effective to split each pair of fans with its own inflatable surface. Despite attempts to fill sections of the bag at different times through a variety in fan behavior, the bag inflation looked fundamentally the same each time.
Take aways:
- Fans work well in pairs for moving air in multiple directions:
Since the fans are incapable of spinning both directions, any sort of inflation requires a pair of fans for bi-directional airflow. As long as the fans are the same type (and have a flat face to mount against each other) pairing them up works well.
- Any fan setup require sufficient space for airflow:
If the device was placed facedown on the table, it severely limited the ability of the fans to inflate the bag. Any installation would require sufficient space behind or beneath any fan setup.
- Fans on the inside (nearest the material) should always be faced to blow air onto the material:
That way when the second fan pulls the air out of the material/bag, it doesn’t accidentally pull the material into a spinning fan blade.
Code for both versions of the project is available here.
Reference Projects:
Zimoun – 2015
Zimoun – 2014
Roosegaarde – Flow
Matthew Plummer – Apifera window installation