Turning on the SAS
Calibrating the SAS
To get the most accurate calibration volume the pump should be run twice. The first time to prime the tubing with water, and the second time to take the calibration measurement described above.
Repeat this test for Pump B (PB) as well. Once finished the proper calibrations can be entered into the sampler settings under the PUMPCAL menu (See Image 6). Navigate to the PUMPCAL menu using the left cursor. Select PA and update the calibration number to what was found in the test for that matching pump. Then select PB and update the calibration number there. Choose ENTER SET and hit enter to save the calibration settings. It is a good idea to check the success of the calibration setting by running a sample volume on both pumps and seeing if the volume set on the SAS matches the pumped volume. If further adjustment is needed to improve the pump accuracy the calibration settings can be adjusted. A larger calibration number will decrease the sampling time and volume, and a smaller number will increase the sampling time and volume.
Setting the real-time clock
Once the calibrations are set, the time on the sampler should be set. From the top line of the PUMPCAL screen press the left cursor on the IR remote to navigate to the TIME SET screen (See Image 7). The first row is the time in HH:MM:SS and the second row is the date in DD/MM/YY. Change the numbers using the up and down cursors to match up with the time and date. Be sure to give yourself time on the seconds setting so that you can select ENTER SET and press enter at the appropriate time to synchronize the real-time clock (RTC) on the sampler with real time. If you return to the top of the TIME SET screen and navigate to the left or right to get back to the STATUS MENU you will see the newly set time and date shown.
Choosing sampling mode and volume
The sampler can operate in two different modes: Once and Daily. Once mode will run Pump A and Pump B at a single given time and date and then wait until a new time and date have been set to run again. Daily mode will run Pump A and Pump B at a given time every day until turned off. From the STATUS MENU use the right cursor to navigate to the SETTINGS menu (See Image 8). Under MODE you can use the up and down cursor to switch between ONCE and DAILY. The next line lets the sample volume be adjusted (both pumps will be set for the same volume). The up and down cursors can be used to increase or decrease the volume in increments of 10mL with an upper limit of 900mL. Once the mode and volume have been chosen select ENTER SET and press enter to save the settings. If you return to the STATUS MENU you will see the newly set sample mode and volume settings shown.
Setting pump sample times
From the top of the SETTINGS menu use the right cursor to navigate to the PUMP A menu (See Image 9). The first line (T) is the time setting for an alarm to go off and activate Pump A while the second line (D) is the date. These options will always both be available, but the date will only be displayed on the SETTINGS menu for Once Mode. Use the up and down cursors to adjust the time (HH:MM) and the date (DD/MM/YY) and then select ENTER SET and press enter to save the Pump A alarm time. At the top of the PUMP A menu use the right cursor to navigate to the PUMP B menu (See Image 10). Following the same steps as above set the alarm time and date for Pump B, and save by selecting and pressing enter on ENTER SET. If you return to the STATUS MENU you will see the newly set pump times and dates listed if in Once mode, and only the newly set times if in Daily mode.
Initiating sampling mode
After calibrating the pumps and inputting all of the settings, the sampler needs to be in a sleep mode to save battery life and initiate sampling and temperature logging tasks. From the top of any of the menus use the left and right cursor to navigate to the INITIATE menu (See Image 11). Select PRESS ENTER, and press enter on the IR remote, and the sampler will go into sleep mode. The time, date, temperature, battery voltage, and sample code will be logged on the microSD card when the pumps turn on. If either of the pumps are set to go off within 60 seconds of sampling mode being initiated "Wait..." will be displayed on the screen until the pump(s) start. When the pumps are activated the screen will read “Pump A” or “Pump B” and the pump will run to collect the sample. If the timing for the two samples to be taken overlaps, the screen will read "Both Pumps" while both pumps are running (See Image 12).
Waking up from sampling mode
While the SAS is in sampling mode no changes can be made to the sampler settings. To wake the sampler up from sampling mode the reed switch on the face of the sampler must be activated. To activate the reed switch press a magnet against the acrylic faceplate and move it back and forth in front of the white reed switch (See Image 13). After 2-3 seconds the sampler should come out of sleep mode and display "Interrupt" on the OLED before displaying the STATUS MENU. The reed switch can be activated in the lab, field, or underwater. If the reed switch doesn't seem to be working, reorient the magnet and try again. Be sure to move the magnet across the face of the reed switch, since a static signal won't wake it up.
Deploying the SAS
Samplers to be deployed should have calibration settings for the two pumps completed, all O-rings lubricated with synthetic lube, fully charged batteries installed, and all securing screws and nuts tightened (no more than hand-tight) to ensure a waterproof seal. Take care when lubricating the O-rings to be sure that they are clean and there is no dirt, hair, or other debris fouling them that might break the waterproof seal.
To prevent any change to the carbonate chemistry of the seawater sample the Tedlar sample bags should be dosed in advance with a saturated mercuric chloride (HgCl2) solution. The volume of the HgCl2 used should be 0.05 to 0.02 percent of the total sample volume (Dickson et al., 2007). A hypodermic needle can be used to inject the HgCl2 through the membrane in the top of the Tedlar bag valve (See Image 14).
Be careful not to puncture the Tedlar bag by pushing the needle too far into the valve membrane.
The sampler can be deployed at any depth up to 180 feet of seawater (See Image 15). It is possible the sampler may be able to be deployed deeper, but 180 feet represents the maximum depth that the design has been tested to. The settings for time and sampling can be made with an IR remote in the lab, in the field at the surface, or underwater. If an IR remote is to be used underwater it should be kept in a waterproof bag to prevent failure of the remote. Once all settings are completed and the sampler has been initiated the Tedlar bag valves should be opened one half turn to be ready to sample.
The natural buoyancy of the sampler should be offset by attaching two pounds of weight to make it slightly negatively buoyant. The sampler can be secured on the substrate by attaching extra weights to the bottom which will maintain an upright orientation (See Image 16), or by securing it to underwater structure (See Image 17) in any orientation. The intake tubing for the sampler pumps should be unobstructed when deployed which is why it is important to make sure that the sampler is not lying sideways in the sediment which could allow debris to be taken in with the water sample and cause clogging in the sample tubes.
Collecting the samplers and transferring the samples for analysis
If the samplers are set to DAILY mode the Tedlar bags will have to be replaced in the interim between samples. To replace the bags, close the Tedlar valve on the completed sample. Remove the Tygon tubing from the valve stem. Carefully take the neoprene cover off of the Tedlar bag.
Be careful handling the full sample bag and avoid handling the bag by the valve as that interface on the bag is the most likely for failure.
Once the filled Tedlar bag is safely removed place a new closed Tedlar bag dosed with mercuric chloride in the protective neoprene bag. The Tedlar sample bag should be placed in the folded contours of the protective bag with as few creases as possible. The valve of the Tedlar bag will come out of both holes in the neoprene to secure the neoprene bag flap closed. Reroute the Tygon tubing through the slits in the neoprene cover and attach to the stem of the Tedlar bag valve (See Image 18). Open the valve a half turn and leave the sampler to take its next daily sample. This procedure must be repeated for every extra sample taken on Daily mode.
Once the samples have been taken and no others are needed the sample bag valves should be closed and the samplers returned to the surface. Care should be taken to store the filled sample bags away from sharp objects and direct sunlight. Assessment of the length of time the Tedlar bags can be stored or left in the field without affecting the carbonate chemistry of the sample is still underway. Preliminary results comparing seawater samples in Tedlar bags to the industry standard borosilicate glass bottles (Dickson et al., 2007) show samples to be stable in Tedlar bags in salt water or air for up to two weeks without altering the chemistry significantly. Before storing the SAS between deployments rinse it in freshwater and rinse the pump tubing by running the calibration command while submerging the SAS pumps. This will clear out any salt or sediment residue and prevent clogging.
Analysis of the carbonate chemistry parameters of dissolved inorganic carbon (DIC) and total alkalinity (TA) can be done directly from the sample bag. If the sample must be transferred to another storage container use caution to minimize gas exchange and splashing in emptying the bag into another container. If this is necessary, be sure to collect more than the necessary sample volume to allow for overflow of the sample into the new container, following normal sampling guidelines laid out by Dickson et al. (2007). One way to carry this out is to attach a length of tubing to the tedlar bag valve, place the tubing in the bottom of a clean borosilicate glass bottle, and gently squeeze the bag until all of the sample has been transferred into the bottle, avoiding splashing and bubbling by keeping the end of the tube at the bottom of the bottle. The bottle can then be sealed and analysis can be run from the bottle as is convenient.
Repeated deployment of the samplers without exchanging or charging the main batteries is possible but not recommended. The optimal working voltage is between 10 and 12 volts. It’s not recommended that the samplers be used with less than a 10 volt charge as the sample volume can become more variable at lower voltages and eventually will cease to take samples at all as battery charge is further reduced.
To retrieve the sampling data from the microSD card first remove the battery packs or power connection, and then remove the microSD card. Download the two text files, dataLog.txt and sampleParam.txt. The dataLog file has all of the sampling time and date settings, the battery charge and temperature at the time of sampling, and a sample code for each of the pumps to identify their status (0 = Waiting to pump, 1 = Initiating pump, 2 = Pump running, 3 = Pumping finished) with a line for each sample taken in the following format:
The sampleParam file has the settings for the sampler including mode, calibrations, pump volume, and time and date for scheduled samples in the following format:
The sampling parameters on the microSD card (sampleParam.txt) are constantly referenced by the sampler program. If the sampler is on when the card is removed the programming on the Teensy microcontroller may need to be reset. Be sure to remove the battery packs before removing the microSD card from the SAS. For the same reason, after downloading the text files from the microSD card, be sure to plug the card back into the microSD slot on the circuit board before plugging the batteries back in. If the batteries are plugged in without the microSD an error message will be displayed saying “MicroSD not detected!” and the sampler STATUS MENU will not be displayed. If, after putting in the microSD card, the sampler settings are different than what was programmed, the Teensy should be reset.
To reset the Teensy remove the batteries, remove the circuit board cover, and press the small pushbutton next to the microSD card slot on the Teensy microcontroller (See Image 19). Plug the battery packs back in and the parameters set on the microSD card should be restored on the sampler.
Safely Shipping the SAS
The SAS is built to withstand deployment on reefs, but the jolts and rough behavior experienced when mailing the SAS or even in having it stowed in luggage in a plane's cargo hold can prove too much for some of the less supported components. To prevent SAS breaking in transit, as with other scientific equipment, some precautions are recommended. To prevent over pressurizing the sealed SAS housing and compromising its watertight seals, be sure to ship the SAS without the top/main O-ring. Even with the O-ring removed the acrylic face of the internal armature can still create an airtight seal with the housing endcap, so make sure the locknuts on the face are not tightened down all the way. Putting packing peanuts, styrofoam, or even cloth stuffing in the body of the SAS and then installing the armature on top will both pad the base of the SAS, and also add cushioning for the armature to be installed nearly all the way, and immobilized with the locknuts, but not closed completely allowing for unpressurized travel.
Troubleshooting the sampler
The motor is running in the wrong direction.
The motor shaft will turn clockwise or counterclockwise according to the order its positive and negative wires are connected to it. If the motor is spinning in the wrong direction switch the order of the wires on the KK connector that connects the motor to the circuit board and that will make it turn in the opposite direction.
The motor isn’t running.
The OLED can turn on with only one battery pack attached, but the motors cannot. If a motor is not firing at the time the alarm is set for but the OLED is displaying the “SAMPLING Pump...” message, look at the battery packs and make sure they are both fully inserted and clicked into place.
The IR sensor (or remote) isn't working to move through the menu.
The IR remote is powered by a single coin cell battery and requires very little energy. If this is the first time using the remote make sure there is a battery inserted and that there’s no plastic covering on the battery preventing it from powering the remote. Look at the wires connecting the IR sensor to the board. Since the connectors on the IR sensor are a simple slide on Dupont connector it’s possible they aren’t fully pushed on and may be preventing the IR sensor from working. Check both the DuPont connectors and the KK connectors to make sure they’re fully attached and then try the IR remote again.
What if the Auto Sampler resets the time every time I remove the batteries?
The real-time clock (RTC) on the SAS is powered by the coin cell battery when a power source (i.e. battery pack) is not attached. Normally once the time is set the RTC will continue counting down at all times and keep track of the time regardless of having a battery pack attached to the sampler. When the battery packs are plugged into the sampler they must be plugged in one at a time with the battery pack below the power connection to the circuit board plugged in first (See Image 1). If the battery packs are plugged back in out of order the momentary lack of grounding will cause the RTC to reset. Try resetting the time and removing and replacing the battery packs in the proper order and see if the time still defaults to an alternate setting. Another possibility is when the voltage of the coin cell battery falls below 3 volts the RTC is no longer powered and the sampler will be unable to retain the set time and date requiring the time and date to be reset every time the power source is reconnected. Replacing the depleted coin cell with a fresh battery will solve that issue.
The sample bag didn’t fill up all the way.
The Tedlar sample bags can hold a 1 liter sample. If the bag is less full than expected, make sure that the Tedlar bag valve is only opened one half turn. Opening the valve too far can allow the water to be pumped out of the valve head instead of into the bag once resistance starts to build up. Also make sure the calibration step was done to adjust for variances in the pumpheads pumping rates, if not, complete the pump calibration process and set the new calibration number for the pump. Other reasons that could cause the sample to be incomplete is issues with a clogged pumphead or bad tubing. If the deployment site is particularly silty or the SAS is surrounded by sand the pumpheads can get clogged up and stop working. For this reason it is a good idea to examine the SAS after each deployment to see if the pumpheads need cleaning. Some of the tubing provided with the pumpheads by the manufacturer has been found to be faulty and can sometimes seal or weld shut after sitting in the pumphead for an extended period of time. If the tubing has welded shut it should be replaced with similarly soft silicone peristaltic tubing.
Both pumps are firing at the same time.
If anything is bridging the gap between the power pins for the motor a misfire can occur. To avoid this check and clean off the circuit board with a soft brush or compressed air. This will help remove any hair, dust, or debris that might create a connection to bridge that gap. Also check the MOSFETs on the circuit board and make sure they're not pressed together (See Image 20). If they are, pull them apart slightly to establish a gap and avoid connecting the two pump circuits. Test the sampler using the RUNCAL menu to see if both pumps still activate at the same time.
Dickson, A.G., Sabine, C.L. and Christian, J.R. (Eds.) 2007. Guide to Best Practices for Ocean CO2 Measurements. PICES Special Publication 3, 191 pp.