Introduction: FLUIDIC RATE SENSOR
Have you noticed that when you move a water hose from side-to-side the water jet lags the hose direction and aligns with it when the motion is stopped. Determining the angular deflection of the water jet at the output of the hose would provide a measure of the angular rate in this sideways direction.
This Instructable demonstrates this principle by constructing a 'Fluidic Rate Sensor' using 'Odds and Ends' available in my 'Home Lab'. The fluid here is 'Air'.
A simple method of testing this 'Gyroscopic Sensor' without the use of standard test equipment is also presented.
- An old CPU fan
- Mosquito repellent bottle (empty and well cleaned)
- Ball point pen with uniform rear tubular section
- Two small bulbs from a series decorative light string
- Scotch-Brite scrub pad
- Few electronic components (refer to the circuit schematic)
Step 1: HOW IT WORKS
The two slides give provide a schematic of the physical layout of a Fluidic sensor and the theory behind the physical phenomenon.
In this design 'Air' is the 'Fluid' which is sucked through a Nozzle using a small CPU-Fan. The air-jet impinges on two heated bulb-filaments forming the position-sensor. A Reference-Bridge is formed by two resistors.
Both arms of the full-bridge so formed are fed with a voltage V+.
Under steady state conditions the air-jet cools both bulb-filaments equally, the bridge is balanced and the output-voltage is zero.
When an angular rate is imposed on the physical system, the air-jet deflects and one of the bulb-filaments is cooled more than the other. This provides an unbalance to the bridge leading to a output-voltage.
This output voltage when amplified provides a measure of the angular rate.
Step 2: CONSTRUCTING THE SENSOR
FOLLOW THE STEPS
- Select two bulbs with similar resistance from the light-string. (Two bulbs with 11.7 Ohms resistance selected)
- Carefully break the outer glass exposing the bare filaments.
- Keep the CPU-Fan ready and check the air-flow direction at a supply voltage of 5 V. ( It is necessary to determine this as the fan needs to be used in a suction mode)
- Cut out the bottom of the mosquito-repellent bottle with a sharp knife.
- Cut away the top of the bottle-cap just exposing the front tubular portion.
- Disassemble the ball-point pen and cut away the bottom end. This should provide a uniform tube which would form the nozzle for the sensor.
- Insert the tube into the bottle-cap.
- Make two small holes in the bottle-body as shown in the picture. This should be suitable for fixing the bulb-filaments diametrically opposite to each other.
- Fix the cap, push the tube to a suitable length just short of the bulb-filament holes.
- Now insert the bulb-filaments into the holes and align them such that the filaments just enter into the periphery of the tube-end as shown. Fix the bulb-filament body to the bottle-body using hot-glue. (As symmetric a placement as possible should be attempted.)
- Fix the CPU-Fan to the rear of the bottle-body (bottom) using hot-glue at the edges. The fan must be mounted so that one of the flat portions is parallel to the plane of the bulb-filaments.
- Make sure the fan blades rotate smoothly and when powered air is sucked out form the rear so forming an air-jet through the pen-body-tube..
The basic sensor unit is now assembled and ready for testing.
This Instructable was made possible by a peculiar circumstance of matching parts:
Selecting parts for this Instructable was done from the 'odds-and-ends' in my 'home-lab'. The CPU-Fan size exactly matched the mosquito-repellent bottom diameter. The ball-point pen rear portion as a tube was a tight-fit into the bottle-cap tubular portion and the step-shapes in the bottle-diameter were suitable for fixing the bulb-filaments. A partly fused-out decorative light-string was available. Everything matched exactly!
Step 3: INITIAL TESTING & CIRCUIT SCHEMATIC
Initial testing was carried out by providing a 5V supply to the CPU-Fan and the voltage excitation to the bulb-filament half bridge.
An Android phone running the 'AndroSensor' application was kept beside the Rate-Sensor hardware and both were rotated in a sinusoidal fashion by hand.
The 'AndroSensor' GYRO graphical display shows the sinusoidal rate pattern. Simultaneously the low-level bridge output is monitored on an Oscilloscope.
+/- 5 mV signal was observed for +/- 100 deg/sec rate.
The electronic circuit amplifies this by 212 to provide the output signal.
Problem & Solution
The output had a significant noise-level even at zero-rate. This was diagnosed as due to unsteady air flow in the system. To overcome this a circular piece of Scotch-Brite was inserted between the fan and the bulb-elements and another at the input tip of the ball-point pen tube. This made a lot of difference.
Referring to the schematic:
5 V is fed to the CPU-fan
5 V is also fed to the 68 Ohm - Bulb - Bulb - 68 Ohm series combination. capacitor C3 filters the motor interference to the bulb-Filaments
5 V is also filtered by an inductor-capacitor combination before providing this as a supply to the OP-AMP
The MCP6022 Dual Rail-Rail OP-AMP is used for the active circuit.
U1B is a unity gain buffer for the 2.5 V reference supply
U1A is a 212 Gain Inverting Amplifier with a Low-Pass-Filter for the sensor-bridge signal
Potentiometer R1 is used to null the full-bridge formed by the potential-divider and the sensor-series-chain at zero-rate.
Step 4: SIMPLE RATE-SENSOR TEST SETUP
Standard Rate-Sensor test equipment includes a motorized 'Rate-Table' providing programmable rotation-rates. Such tables are also provided with multiple 'slip-rings' so that the input-output signals and power-supply for the unit-under-test can be provided for.
In such as setup only the rate-sensor is fitted on the table and other measuring equipment and power-supply are placed on a table by the side.
Unfortunately, access to such equipment is not available to DIY enthusiasts. To overcome this an innovate method using DIY methodology was adopted.
The primary item available was a 'Rotating Side Table'
A tripod stand was fitted onto this with a downward looking digital camera.
Now, if the rate sensor, power-supply, output-measuring-devices and standard-rate-sensor could be mounted on this platform. Then the table could be rotated Clockwise, Anticlockwise and to-and-fro to provide different rate -inputs to the sensor. While in motion all data could be recorded as a movie on the digital-camera and analysed later for generating the test results.
Having done this, the following was mounted on the table:
Mobile-phone-power-bank to provide 5V supply to the rate-Sensor
A digital multi-meter to observe the output-voltage. This multi-meter had a relative-mode which could be used of zeroing at zero-rate.
An Android phone OTG mode Oscilloscope using the 'Gerbotronicd Xproto Plain' hardware and 'Oscilloscope Pro' Android application from 'NFX Development' to observe signal variations.
Another Android phone running the "AndroidSensor' application by 'Fiv Asim'. This uses the phone inertial sensors to display the pitch-rates. Using this in the z-axis gives a reference-value to test the Fluidic-rate-sensor under test.
Test were carried out and some typical test cases are reported:
CCW Z:+90 deg/sec multi-meter -0.931 V, Oscilloscope ~ -1.0 V
CW Z:-90 deg/sec multi-meter +1.753 V, Oscilloscope ~ +1.8 V
Scale factor based on average of these two 1.33 V for 100 deg/sec
Sinusoidal test Android Phone reference p-p 208 deg/sec, multi-meter cannot respond correctly, Oscilloscope shows 1.8 Sec period, p-p voltage 2.4 Div X 1.25 V/div = 3 V
Based on this 1.8 Sec period corresponds to 200 deg/sec p-p
Scale factor 1.5 V for 100 deg/sec
Step 5: SUMMARY
FAILED TEST METHOD
Initially a method of mounting sensors, oscilloscope and reference-rate-sensor on the rotating table and observing data, manually or using a camera from the side was tried. This was a failure due to blurred images and insufficient response time for a human observer to record values.
TAKE HOME OBSERVATIONS:
The Fluidic-Rate-Sensor constructed for this Instructable serves the purpose of demonstrating a the concept it set out to do. However, the sensor needs to be built with better precision if has to serve any practical purpose.
The DIY method of rate-sensor testing using a rotating-table with all equipment and power-supply on the table-top is recommended for use by the Instructable community.
Participated in the