The sampler has now been used in the field for a study in nitrate concentrations. Part of the deployment process included placing the collection bottles in a chilled container, such as a insulated lunch box with ice packs. This adds to the complexity and cost of deployment. However, we needed to be sure that this was required. We then conducted 12 and 24 hour experiments to see if nitrate concentrations deteriorated without being chilled. The 12 hour experiment revealed that we did not need to chill the samples while the sampler was deployed. We are still waiting on the 24 hour experiment results. Nitrate seems to be pretty stable. That is, it is not changing over to N2 or something else too quickly. This would not be the case for studies of other nutrients such as phosphate.
The unit has run as expected. The filter over the pump housing was easily cleaned. The 4,000 mAh battery from Voltaic systems easily ran the system over 24 hours, dropping about 1/4 power during that time. The system could be deployed for up to about 3-4 days at this rate.
Deployment in small brook.
Deployment in an even smaller brook.
Screen over pump housing before cleaning.
Well almost. I realized a useful feature would be a button that, when pressed, would run each pump for 10 seconds. This would allow for cleaning with deionized or distilled water and also to check that all pumps are functioning. From the time a current is applied to the pump to the time water makes it out of the end of the ~3′ tube (1/4″ inner diameter) is about 1-2 seconds.
Check out these photos of the complete system after its first successful run in the bathroom sink.
The complete system.
Detail of the pump-head. You can see the the screws and nuts used that perfectly hold on to the pump’s housing. Click on the photo for a bigger view.
The bottom of the pump head. To help keep the pumps running, a drain sleeve will be cut and placed over the bottom and maybe top openings of the housing to prevent sediment from potentially clogging up the pump.
The pump-head housing is a 4″ PVC pipe cut in half. This allows for easier drilling of the pump retainer screw holes and getting the tubing into place.
So after working out some current issues, the unit is working as expected and I have moved it into a housing. It’s almost completely sealed except for the headphone jacks I’m using as pump connectors. While it should be fairly water resistant, it is by no means waterproof.
The guts of the unit are complete. Headphone jacks allow for the easy exchange of pumps. A basic test script shows all pumps are running as expected. The next task is to place the unit in a housing that will protect it from the elements.
The sampler is moving along well. The setup reads the dates and times from a config file on micro sd for each of the 5 pumps to run. Time is kept by a real time clock and the pumps are micro submersible off the Web for less than $5 each. Next steps are putting together all 5 pumps and then packaging.
Get the source from github.
Work on my thesis this summer will require that I be in more than one place at a time! Getting pricing on commercial/scientific portable autosamplers shows that there is plenty of room for more affordable options. That is why I will be developing a simple autosampler to collect and store water samples from river streams. The initial sampler will be arduino powered. 3-5 submersible pumps along with 3-5 collection bottles and a real-time clock should do the trick.
The logger Revision 3 of the Eco Sensor Network is headed out for its first field test. The R3 features on-board temperature and relative humidity sensor as well as logging to micro-SD. This particular unit will be recording fluctuations in tree stem size using a dendrometer and a 16 bit ADC bridge.
The oven controller assembly is complete. The device allows for the controlled heating of a toaster oven. Simply plug the toaster oven into the orange cord, turn the device on, set the rate at which you want the temperature to increase, place the thermocouple inside the oven and hit the start button. The logic is a bit off at the moment, but it works fairly well.
The final product. Oven controller.
Arduino UNO R3
2X16 LCD Screen
Thermocouple and Shield
Eastern Software Group develops low-cost electronics and equipment for scientific use and provides software for general scientific computing.