Artificial pancreas

Discussion in 'The Projects Forum' started by maliha tamkeen, Jan 9, 2016.

  1. maliha tamkeen

    Thread Starter New Member

    Jan 6, 2016
    10
    0
  2. GopherT

    AAC Fanatic!

    Nov 23, 2012
    6,060
    3,823
    Look closely at the breadboards to the right-hand side of each photo of a breadboard. Make sure your circuit is the same. Also, make sure of your resistor values. Don't just look at the color codes. Make sure of the values by measuring them.

    Finally, make sure your transistor is the same model and oriented correctly.
     
    cmartinez likes this.
  3. wayneh

    Expert

    Sep 9, 2010
    12,144
    3,054
    Wow, with no photos or diagrams of your actual work, we're supposed to diagnose your problem?

    Here are some things to get started:
    Does your pump work when powered directly?
    Are your batteries fresh?
    What MOSFET are you using? (Check and double check the pinout. It's very easy to get it wrong, or inserted backwards into the breadboard.)
    Do you have a multimeter to check some voltages? If so, measure some your nodes and report what you find. The MOSFET gate voltage especially.
    Also, check to see if you can see an ohms variation across your sensor.
    Divide and conquer - figure out what is working and what is not.
     
    cmartinez likes this.
  4. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,292
    1,255
    I think there may be an error in the circuit. The 100k and 300k resistors should be swapped. Then of course the sensor must be in the acid.
     
  5. maliha tamkeen

    Thread Starter New Member

    Jan 6, 2016
    10
    0
  6. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,292
    1,255
    There is a schematic on the page called help.
     
    cmartinez likes this.
  7. maliha tamkeen

    Thread Starter New Member

    Jan 6, 2016
    10
    0
  8. cmartinez

    AAC Fanatic!

    Jan 17, 2007
    3,574
    2,543
    After seeing the diagram:

    artificial-pancreas-circuit-diagram.jpg

    Wouldn't it be best if PWM were to be used instead of trying to control the voltage at the mosfet's gate? In fact, I have very serious doubts about the practicality of the circuit above.
     
  9. wayneh

    Expert

    Sep 9, 2010
    12,144
    3,054
    I was going to question that, because the MOSFET has a threshold voltage of just 2.5V, but the schematic is for a 12V supply. With just 6V, I think you're right. There's no way to pull the gate high enough otherwise.
     
  10. wayneh

    Expert

    Sep 9, 2010
    12,144
    3,054
    It'll work. Just mentally replace the sensor with a wire, or a switch. In fact, that would be a good test for the TS to run. If that doesn't trigger the MOSFET, he can rule out the sensor. The 100k resistor could probably be eliminated. Just absorb it into the pot.
     
  11. cmartinez

    AAC Fanatic!

    Jan 17, 2007
    3,574
    2,543
    But wouldn't it be better to use a traditional bi-polar transistor instead? Just wondering, I'm not the expert here.
     
  12. maliha tamkeen

    Thread Starter New Member

    Jan 6, 2016
    10
    0
    I would really appreciate if someone suggest the solution.
     
  13. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,292
    1,255
    Removing the 100k is a good idea.
    So @maliha tamkeen why don't you try that?
     
  14. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,292
    1,255
    He probably can't get enough base current to turn on a BJT into saturation.
     
  15. cmartinez

    AAC Fanatic!

    Jan 17, 2007
    3,574
    2,543
    Right... how about a darlington then?
     
  16. sailorjoe

    Member

    Jun 4, 2013
    361
    63
    First, use a spare wire and put a short circuit across your mosfet, from drain to source. Does the pump turn on?
    If that doesn't work, then there is either something wrong with the pump or with power to the pump. Fix whichever it is and let us know.
    If that works, then remove your spare wire and use it to put a short circuit across your conductivity sensor. Does the pump turn on?
     
  17. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,801
    1,105
    With the circuit shown there is a good chance that the FET will not turn on fully and will overheat. It really could benefit from a snap-action gate-driver. I suggest you research 'comparator' and 'hysteresis'.
     
    sailorjoe and cmartinez like this.
  18. wayneh

    Expert

    Sep 9, 2010
    12,144
    3,054
    I don't see a reason the avoid the MOSFET. Getting voltage should be easy, it's current that is lacking.

    It's possible it might work best without either of the bias resistors. The pot at ~500K should pull down the gate enough when the sensor goes "open". Then the sensor should have no problem getting nearly full 6V onto the gate when it is immersed.
     
    cmartinez likes this.
  19. atferrari

    AAC Fanatic!

    Jan 6, 2004
    2,648
    764
    Mr Maliha is not telling what tests were carried out nor their results.

    Or just a clear picture of how the circuit was implemented.
     
Loading...