The device is made for use in Cal State LA CEA-CREST research of how flow in around coastal area would affect the growth and breeding of mussels. The device is a Turbine based meter that would directly measure the velocity of water passing over it. This is a program sponsored by NSF.

Step 1: Parts Required

Here is the list of parts required to recreate this Flow Meter

Non-electronics parts:-
1. 5cm Dia. with 12cm long PVC Tubing

2. Propeller with mounting from GlobalWaters.com

3. Any Water TIght Compartment that is bigger than 6 x 15 x 5 cm

4. 12 oz. of Aquarium Type Sealant.

5. Zip Ties

6. A metal plating with a screw stud to mount the device onto the mussel bed.

Electronics Components:
1. Capacitors:-
1 x 0.01mF
1 x 1.0mF
1 x 10nF

2. Resistors :-
2 x 100ohm
2 x 10kohm
1 x 1kohm

3. 1 x EL-USB-3 Voltage Datalogger

4. 2 x 9V batteries

5. Multicolor roll of 14 Gauge Single Strip Wires( could use ribbon wires too)

6. 1 x 5x3 cm copper plated board.

7. 2 x Connector for 9V Batt

8. 1 x Melexis MLX90217 (Or any other Hall Effect sensor would work)

9. 1 x National Semiconductor LM2907

10. 1 x 7810 power regulator rated at 10V 1A max.

11. 1 x 2 pin switch (optional but recommended)

Step 2: The Board

Here, the first step would be to connect all everything as shown in the diagram below:-
( All positive lead should be connected to pin 3 of the 7810 power regulator and all ground to pin 2)

Step 3: Step 1 (continued)

After all components is connected, it should look something like the three little board on the top corner of the picture. (Notice how i made the first prototype with three different boards to make it easier to swap and test out different designs.)

The board on the left is the power regulator, the 7810, it is connected by two wires that runs to the main board, which is the middle board that contains the LM2907 and some resistors and capacitors that controls the maximum amount of revolutions that the circuit can take (which is around 830Hz) and also the ratio between the frequency to the corresponding voltage. The board on the right is the board that translate the mechanical movement of the fan into pulses that the LM2907 understand. It also contains a 10nF Capacitor and a 5.6kohm Resistor that reduces the noise in the circuit and also to protect the MLX90217 from any voltage spikes that might occurs when water do get into the circuit.

Step 4: The EL-USB-3

The datalogger that we picked were made by Lascar. It is a USB based voltage datalogger that has its own built in power supply that ensures that data would be safe even if the main 9V batteries running the Turbine meter is depleted. Furthermore, with a battery life of a year with continuous usage, it would safe alot on buying new batteries.
With a maximum of 32000 data points and a variable recording time, (from 1s to 24hrs), we either have a good idea of how speed changes in a small time frame or just take a sample at a long time interval that could keep the data logger running recording data for days. Furthermore, data downloading is easy with the USB connector, just pop it out of the box, download data, then pop it back into the box.
Future data logging would most likely be a wireless bluetooth type that would either be continuous live feed of data or keep the data in the device and then download data via bluetooth.
For the the current data logger, all data would then be downloaded into the computer and with the downloadable program that can be found on Lascar's website, data can be viewed in either raw data points or in graph form.

Step 5: Fan Enclosure

In the making of the fan with the enclosure, the fan that was bought from GlobalWater.com was mated to a aluminum rod that was fabricated in the machine shop as shown in the picture. As can be seen, the fan mount was screw on the side to ensure stability. And the front screw was there to ease maintenance. The small little black dot is the sensor MLX90217 that is placed near the fan to detect the magnetic field change that occur with the spinning magnet that is embedded into the fan fin.
This method is a good way to ensure no mechanical failure could arise due to the fan spinning and also it reduces the amount of parasitic loss that could follow in a mechanical sensor, i.e a switch. As can be seen is the black box that stores all the electronics.
A hole is drilled through the box cover and the enclosure for the magnetic sensor to pass through and be able to detect the magnetic flux. Aquarium type sealant is used to seal the hole where the sensor passed through and around the base of the enclosure to the box to ensure proper sealing from seawater.

Step 6: Testing

After the sealant is dried out, usually about 23 hours, the whole device could be tested in a wave tank with a simulated mussel bed to ensure that the device is recording data as it should be and the fan should show signs of velocity increasing and decreasing as waves hits the mussel bed.
The relationship between the revolutions and the velocity is found to be V = 0.0644 f - 0.0303 with a linear relationship with a +-1%. Which is good considering the data needed is only a rough data for a general idea of how flow speed affects the growth and production of mussel larvae.
For each device to be properly calibrated for more accurate measurements, a steady state flow test should be done for every one of the device. But it is not really require as the fan bought was made using a injected mold method that would ensure each fan would be the same.

Step 7: Testing 2

Here you can see the calibration of the frequency to the average flow velocity

Step 8: Final Test

Here is the fan enclosure being rigged and tested in a simulated mussel bed and we can constructive and destructive waves forming and the fan moving along with it.