This week Olaf and I went to the junkyard to salvage a pressure sensor; we found one in an '89 Cadillac.
4/21 - 4/27:
This week we began our design for the pressure sensor.
We met with the group and discussed the following project parameters:
- Pressure Sensor design finalized
- Drilling of lid discussed
- Piping and wiring of PVC tubing designed
- Cost analysis performed
4/28 - 5/4:
The breadboard came in this week. We've begun testing this week.
 |
| Air test - sensor 1 |
 |
| watertight |
We met with the group and discussed the following project parameters:
This pressure chamber will be used by the robotics team or, at the very
least, serve as a prototype for a more elaborate apparatus to be made.
So, the following considerations should be made that are relative to this idea:
- User Case scenario
- How long does it take to fill and empty the standing pipe?
- To calculate fill time, we can calculate inward flowrate against increasing opposing pressure
- To calculate drainage, we can use the volume flow rate
- Cost Analysis
- How much volume of water is in the standing pipe?
- What is the rate of water in LA country?
- With the above, we can calculate the cost per fill
- if substantially small, we can present fills per dollar, etc.
5/5 - 5/11:
Our sensor fried this week! It now operates as a switch. We found another switch, and a different type of sending unit. We began testing the new sender this week. The Arduino came in this week so we started on the code this week.
5/12 - 5/18:
We received a new sensor this week. 122PC15A strain gage transducer. Our difference amplifier design was slightly modified.
We are now using a wheatstone bridge to more easily measure fluctuations in voltage.
We began testing this sensor, made terminals, and researched the 122PC datasheet. Unfortunately, the datasheet says that the 122PC15 is not accurate when using liquid as a measurand. We did begin the testing for the 122PC15 though. This is Plan F - just in case all other options fail.
 | |
| 122PC15 Air test |
 |
| Sensor with grounded case - too much fluctuations in resistance |
We began lid construction this week during the group meeting.
5/19 - 5/25:
We recieved the pressure sender from glowshift.com. This sender looks a lot more promising.
Sensor Specifications:
GS-S01 uses a diaphram gauge while the GS-ES02 uses a microchip
GS-S01 Ohm Specs:0 psi - 3 ohms
14.5 psi - 17 ohms
29 psi - 34 ohms
43.5 psi - 51 ohms
58 psi - 68 ohms
72.5 psi - 85 ohms
87 psi - 102 ohms
101.5 psi - 119 ohms
116 psi - 133 ohms
130.5 psi - 147 ohms
145 psi - 160 ohms
GS-ES02 Ohm Specs:
0 psi - 0.5 ohm V
14.5 psi - 0.9 ohm V
29 psi - 1.3 ohm V
43.5 psi - 1.7 ohm V
58 psi - 2.1 ohm V
72.5 psi - 2.5 ohm V
87 psi - 2.9 ohm V
101.5 psi - 3.3 ohm V
116 psi - 3.7 ohm V
130.5 psi - 4.1 ohm V
145 psi - 4.5 ohm V
Tested the sender with air to see the resistance change -- it appears linear. Additionally we tested the sensor with large amounts of water above the sender, and also noticed the resistance changes as with the air tests.
This sensor will be used in the bottom leg of the wheatstone bridge. We decided on a wheatstone bridge that does not have a feedback resistor Rf tied to the non-inverting op-amp input.
This could be a result of two factors:
- not enough water above the sensor
- resistance change is too small to appreciably load the bridge
We began looking into multiple op-amp configurations so that we could set a gain on the sensor output. We settled on a dual op-amp configuration where the left leg of the wheatstone bridge will be tied into an inverting amplifier.
At the group meeting, we began testing this configuration with water.
5/26 - 6/1:
We tested the new dual op-amp configuration with air and made some notable results.
 |
| Dual op-amp configuration |
Turns out our bridge was never balanced.
In an attempt to balance the bridge, we switched out the fixed resistors and tried a configuration with 4 potentiometers and the sensor connected in series to one of the potentiometers.
Finally, we settled on a configuration that uses a potentiometer in the bottom half of the right side of the wheatstone bridge.
To ensure zero voltage across the bridge, we connect a voltmeter across the bridge and measure the output - we adjust the trimmer accordingly until the voltage reaches zero. We hooked the sender up to an air pump to measure pressure fluctuations.
Additionally, we just about finished the Arduino coding that takes an analog input, maps it in a range from 0 to 125 kPa, prints the output to the serial monitor, and prints a graph. A big thanks to: David A. Mellis, Tom Igoe, and Scott Fitzgerald for making the code available online. We just need to create a new variable for pressure and the proper computation that will convert the voltages to pressure values. We will do this when we conduct the water test.
6/2 - 6/14:
We tested the new wheatstone bridge with the trimmer. Success with air!
We see a change in voltage across the bridge when pressure is applied to the sender. The group met up at Mt. SAC for the water test. We wanted to get a baseline reading for pressure so we can determine the ratio of volt/kPa. We have success. We connected the sender to 30 ft of hose and observed the output voltage from the op-amp change as pressure was applied to the sender. The wheatstone bridge difference amplifier was successful.
We've continued to work on the project. We just about finished construction on the barrel and will be able to get the real water test soon -- barrel, pvc tubing, the works.
We were given a different sensor and attempted to get pressure readings from the sensor to analyze the voltages. Unfortunately, the sealed barrel can not withstand as much pressure as we had initially expected. The lid on the barrel failed at about 70kPa.
However, we have not given up. We soldered the circuit onto a PCB; it is ready to receive a two-terminal sending unit. We are not yet able to test the Arduino code; we are in the process of exchanging the sensor.
