Assignment 8, Part 3: add flow control and test your device

From Course Wiki
Revision as of 16:28, 4 November 2019 by MAXINE JONAS (Talk | contribs)

Jump to: navigation, search


The next step is to set up the reservoirs, tubing, and valves to control the flow through our microfluidic device. The fluid from two large reservoirs will be connected to the PDMS device with thin tubing. We also need a mechanism to drive fluid flow through our device. While some devices use carefully regulated air pressure to drive flow, we will be using a peristaltic pump to pull fluid through the tubing. A pinch valve (that does exactly what you think) will allow us to choose which reservoir the pump will pull fluid from.

Schematic of flow setup

The solenoid pinch valve has two slots for tubing – one is normally open (NO), which will be used for the regular (low salt) medium, and one is normally closed (NC), which will be used for tubing connected to the high salt medium. Applying 12V to the solenoid will switch the flow from low to high salt. We will use a short length of flexible silicon tubing that can be easily pinched by the valves, and connect it to the device and reservoirs using a stiff less-expensive tubing made by Tygon.

Pinch valve operation schematic

Assemble the fluidics caddy

  1. Grab one of the L-shaped black acrylic fluidics caddys and secure it to your vertical P14 post.
  2. Secure two 50 mL conical tubes to the leftmost side of the acrylic using two routing clamps, a 1/4-20 screw and a wing nut.
  3. Secure the pinch valve to the acrylic nearest the P14 post using a two line routing clamp.
Fluidics caddy

Using a data acquisition card to control pinch valve and LEDs

To oscillate the flow between two fluid reservoirs, we need to switch on and off the solenoid valve at a particular rate. Let's get MATLAB to do the work for us by interfacing the valves with a lab computer. While we're at it, wouldn't it also be nice to turn on and off your LEDs using a computer signal? A data acquisition or (DAQ) card allows us to send and receive electronic signals from the computer. These are powerful cards, however, we'll only use them to provide a digital output of either 0 or +5V to control whether something is on or off.

Unfortunately the +5 V signal from the computer cannot provide enough power to run our LEDs at 1A of current, nor can it provide the +12 V necessary to control the solenoid pinch valve. Should we give up and go home? No! Let me introduce you to the cornerstone of digital electronics: the Transistor. We won't go into detail about how these incredibly useful and versatile circuit elements work, but this video goes through an awesome explanation if you're interested. For our purposes, we'll implement a transistor in the following way, where it will act like a voltage controlled switch:

Using a transistor as a switch.

By sending a 5V signal to the gate, the transistor effectively "closes the switch" and allows current to flow from the drain to the source, turning the circuit ON. Sending a 0V signal in turn, prevents current from flowing from the drain to the source, effectively acting like an "open switch", turning the circuit OFF. We will implement three of these circuits: one for each LED and one for the solenoid valve.

LED control circuits


The LED control circuits are implemented as shown in the above figure. Note that:

  • P1.0 and P1.1 are the signals generated by the DAQ.
  • When the DAQ signal is +5V, the LED will turn on. 0 V from the DAQ turns the LED off.
  • The 1kΩ resistor connecting the gate to ground is there to prevent the transistor from switching on accidentally when the DAQ is disconnected.
  • Each LED is connected to its own channel on the lab power supply, so the blue and green brightnesses can be controlled independently.

Solenoid valve control and power management

To control flow through our microfluidic device, we will use essentially the same circuit to open and close the solenoid pinch valve as we did to turn on and off the LEDs. The circuit is shown below.

Solenoid pinch valve control circuits.

There are two small modifications that we need to make compared to the LED control circuits. Namely:

  1. There is a diode in parallel with the solenoid valve. This prevents voltage spikes caused by switching and inductive load on and off.
  2. We'll use pulse-width modulation (PWM) to reduce the power consumed by the valve. Supplying the solenoid valve with +12V for long periods of time will cause it to heat up significantly. This can have adverse effects like heating the media in the tubing which can cause bubbles. We want to avoid bubbles at all costs. The valve requires +12 V to turn on initially, but requires much less to stay on. Our solution is to use the DAQ switch the circuit on for 100 ms (to engage the solenoid valve), then switch on and off the power to the valve at 10kHz with a 17% duty cycle (to keep the solenoid engaged, but with lower power).
Solenoid valve PWM.

Build the DAQ interface circuits

On an electronics breadboard, build the three control circuits detailed above.

Note the following things:

  • Use an N03HDL transistor, found in the rightmost of the west drawers. Check the part number since we have several different transistor models floating around the lab that look the same but behave differently.
  • The diode needed for your pinch valve circuit can also be found in the west drawers. Notice that the silver band around the diode indicates the cathode. Pay attention to the diode's orientation when connecting it to your circuit.
  • Use the two channels of the lab power supply to power each LED circuit (in independent mode).
  • The pinch valve will use the +12 V output from the Diablotek computer power supply (below). Also shown below is a Molex connector that you can use to connect the Diablotek power supply to the breadboard.
  • Use the cable provided (see figure to the right) to connect your breadboard to the DAQ. Each wire should be labeled based on the pin it is connected to in the table below.
DAQ connection cable
Signal Name Signal Location Pin wire color
Blue LED P1.0 Red
Green LED P1.1 Orange
Digital GND DGND Black (bundled with Red/Orange)
Solenoid pinch valve CTR.0 +Green / -Blue or -Black
Digital GND DGND Blue or Black (bundled with Green)

Test your circuit

  1. In MATLAB, initialize the OsmoticShocker
      foo = OsmoticShocker;
  2. Test your control of blue and green LEDs using the commands:
    • If your circuit is working properly, only one color will be on at a time (i.e. the command foo.GreenOn will both turn off blue and turn on green.)
  3. Test your pinch valve control using:
    • You should be able to hear the valve switch on and off, and see the plunger move.
    • Make sure the Diablotek power supply is switched on.
    • It's best practice to leave the valve in the off (low salt) state when not in use.

Test your flow using fluorescent dyes

Initiate flow through the PDMS device

We need to test that the flow in our microfluidic device is behaving like we expect it to. To visualize the flow, we'll fill each fluid reservoir with a different fluorescent dye. One reservoir will contain Rhodamine B, which is excited by green light, and the other will contain fluorescein, which is excited by blue light.

  1. Use the 50 ml conical tubes to hold 15 ml fluid reservoirs: one with Rhodamine B and one with fluorescein.
  2. Mount your device onto a microscope stage, and insert the inlet tubing into the reservoirs.
  3. Record which valve state ("high" or "low" salt) corresponds to which fluorescent dye (fluorescein or Rhodamine).
  4. Connect the outlet tubing and waste tubing to the peristaltic pump.
  5. Use a 50 ml tube and tube rack as a waste container for the outlet, and tape the waste tubing in place.
  6. Insert the flexible silicone tubing from one reservoir into the open slot within the pinch valve, making sure that the silicone tubing is fully seated in its groove.
  7. Use Matlab to open the high salt valve and insert the other inlet tubing into the now-open slot. Again, wiggle the tubing back and forth to make sure it is all the way back into it's slot.
  8. Turn on the peristaltic pump to initiate flow through the tubing.
  9. When one side of the "Y" is filled, flip the valve to the "low salt" state, and fill the other side.
  10. Once the device is filled, you may want to turn down the flow rate from the pump. Flowing too slowly will cause stalling in the pump, but too quickly will disrupt the yeast cells in our later experiments. A value around 20 is probably a good place to start.

Set the camera exposure and gain and test your setup

For each valve state ("high" or "low" salt),

  1. Open the valve for that state.
  2. Turn on the LED that excites the fluorescent dye that is flowing through the device. Increase the LED current to (but not more than!) 1 A.
  3. Set the exposure time using the appropriate command below to get a bright (but not saturated image):
foo.SetBlueImageParameters( gain, exposureInMicroseconds ) 
foo.SetGreenImageParameters( gain, exposureInMicroseconds ) 
  1. With the LED still on, flip the valve to its opposite state. You should see the fluorescence decrease. If you can't see a difference in fluorescence between the two valve states, it's time to do some debugging! Some good questions to ask are: can you see droplets forming at the outlet tubing (demonstrating that there is flow) when either valve is open? Are there any bubbles blocking your tubing? Are your exposure settings sufficient?
  2. Repeat the above steps for the other fluorescent dye and its corresponding excitation source.
  3. Close and open the valve a few times to make sure you get the expected behavior.

Record some test data

Once you're sure everything is working, take some test data.

  1. Record a step response movie ("low salt" first for 10 s, then "high salt" for 60 s) using the following commands:
  2. In addition, record movies for at least two oscillation frequencies (between Rhodamine and fluorescein), including 2 min and 1 min.
  3. Don't just take the data and go home. The point of this exercise is to make sure your device is working. If your data isn't what you expected for both illumination colors, it's time to troubleshoot!

  1. For each oscillation movie, plot the average frame intensity vs. time for each illumination color. Also plot the valve state (1 or 0) vs time. Subtract the minimum value and divide by the range from your average frame intensity so that it oscillates between 0 and 1.
  2. Plot the average frame intensity vs. time for your step response movie for each color. Again, normalize your data so that the intensities vary from zero to one.
  3. Did your fluidic device behave as expected? Why or why not? Reflect on any challenges you faced in the lab and what you might do differently in following experiments.
  4. Note: This question is moved to Assignment 9 for Sp2019 Fit the low and high salt data to the following functions and plot your fit results on the same plot as the data (question 3).
    $ y_{\text{low salt}}(t) =\begin{cases} e^{-(t-t_{0,1})/\tau_1}, & \text{if $t>t_{0,1}$}\\ 1, & \text{otherwise.} \end{cases} $
    $ y_{\text{high salt}}(t) =\begin{cases} 1- e^{-(t-t_{0,2})/\tau_2}, & \text{if $t>t_{0,2}$}\\ 0, & \text{otherwise.} \end{cases} $
  5. Note: This question is moved to Assignment 9 for Sp2019 Record your estimates for each fit parameter, $ \tau_i $, and $ t_{0,i} $, of your fluidic system.
    • What physical aspect of your experiment do each of the above fit parameters represent?
    • In Assignment 10, we will be measuring oscillations with periods of several minutes. Can we assume an ideal step response in that case? Why or why not?

Clean up

  1. Remove the inlet tubings from their reservoirs, disconnect them from the pinch valve, and drain the device of all remaining fluid.

Global Tree.gif Make sure to dispose of liquid and solid waste as follows:
  1. Disconnect the tubing and dispose of the device (PDMS, coverslip, blunt-tipped needles) in the sharps waste container.
  2. Tubing can be disposed in the regular (non-sharp) biohazard waste container.
  3. Any other used needles (even blunt ones) and connected syringes go in the sharps waste.
  4. Any unused fluorescein and Rhodamine B from their reservoirs can be kept for other groups, but you may empty the mixed fluorescent dye waste into the designated waste bottle on the wet bench.


Back to 20.309 Main Page