rodent touch circuit

Discussion in 'The Projects Forum' started by burtS, Jul 26, 2011.

  1. burtS

    Thread Starter Member

    Mar 6, 2008
    29
    0
    My object is to create a ‘touch’ circuit in which a rodent standing on a stainless steel floor and licking at a stainless steel spout will serve as a switch and be detected.
    Restrictions:
    Prior research has shown that rodent detection threshold is about 0.5ua. Current flow through subject must be less than that.
    Unit is to be powered by a well regulated 24 vdc power supply.
    Output must be above 3v and below 30v to be detected by the input of my digital interface.
    Circuit should be as simple as possible (!).
    My solution using a Darlington is in Figure 1. Can't remember why I wanted to reduce voltage but I did. Touch 1 and 2 are contact points for the subject. It does seem to work reliably but may not be optimal or wrong in some basic way that I’m not aware of. Also: with this circuit current flow to the interface is 10ma.
    Your comments/criticisms would be appreciated!
     
  2. wayneh

    Expert

    Sep 9, 2010
    12,100
    3,034
    You need a comparator, or maybe an op-amp configured as a comparator. This will allow a very tiny current to appear as continuity and switch the output from the no-rodent to the plus-rodent state. The transistor idea is right, but the op-amp has dramatically better circuitry to accomplish a much higher current gain. Yes, it has more pins, but is actually far simpler to use once you've learned it. An enormously useful too for lab work.
     
    burtS likes this.
  3. wayneh

    Expert

    Sep 9, 2010
    12,100
    3,034
    Here's an example of a comparator circuit I used to detect the presence of the load (in your case, a rodent). It would turn on the light by just using my fingers as the load. A rodent's tongue is more sensitive than my finger (so is MY tongue) and this circuit may need modification, but it's an example.
    Picture 1.png
     
    burtS likes this.
  4. ErnieM

    AAC Fanatic!

    Apr 24, 2011
    7,386
    1,605
    My pick would be a MOSFET or a CMOS gate. The fet could run off your 24V, the CMOS has a 15V max I believe.

    You make a grounded plate for the rat to stand on. Connect the fet gate (or gate input) to the water, and add a meg-ohm plus pull up resistor.

    I used the CMOS gate idea once to make a rain sensor. A single raindrop would be enough to trigger the circuit.

    I don't think you captured your work 100% on the schematic, but if the rat and the digital interface both agree it works then that's your seal of approval.
     
    Last edited: Jul 26, 2011
    burtS likes this.
  5. wayneh

    Expert

    Sep 9, 2010
    12,100
    3,034
    Missed that! If it ain't broke...
     
  6. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Oops - the MOSFET is a good idea, but most MOSFETs have a limit of ±20v Vgs (voltage from the gate to the source terminals) and many are lower; they'd get zapped at 24v.

    Another thing, with 24v in, < 0.5uA means that there will have to be > 48 megOhms resistance from +24v to ground.
     
  7. Jaguarjoe

    Active Member

    Apr 7, 2010
    770
    90
    Some microcontrollers, like the simple Picaxe, have a "touch" input. A small change in capacitance upsets an oscillator and provides a digital output. Being capacitive, no steady current flows.
     
  8. #12

    Expert

    Nov 30, 2010
    16,261
    6,770
    I am so impressed with the rats ability to detect current that I doubt the .5ua requirement...like maybe it's a typo. Did you really mean that, Burt?

    Wookies pointer suggest the use of lower voltage to stay within the limits for a MOSFET and to make it easier to get resistors in a useful range.

    The MOSFET circuit would be about as simple as the original in post 1. Make R1 = zero ohms, change R2 to 22 meg0hms, put 10 to 22 meg ohms in series with the gate, and change to 10 or 11 volts.

    JagJoe points out that proximity detectors can be done in several ways, but a single transistor and a couple of resistors is so simple that it's hard to argue for a more complicated circuit.
     
  9. KMoffett

    AAC Fanatic!

    Dec 19, 2007
    2,574
    230
    Can post (later today) a circuit used in our Psych Department's operant training lab for a rat-touch water dispensers. Works with a MED Associates 24V system.

    Ken
     
    burtS likes this.
  10. ErnieM

    AAC Fanatic!

    Apr 24, 2011
    7,386
    1,605
    Good point. Years back we had an inverter circuit driving FETs to just 15V that would occasionally, usually overnight during an endurance test, blow the FETs and anything around them. The fix was to drop the drive down to 12 or 10 volts. The lesson was don't run them anywhere near the max. I've heard manufacturer's only sample test a lot for this parameter as it is such a device killer.

    As this is such a minimal current anything like a zener to limit may draw too much leakage current. Best way is to make a resistor voltage divider for 10-12V to limit the voltage on the pull up resistor.
     
    burtS likes this.
  11. KMoffett

    AAC Fanatic!

    Dec 19, 2007
    2,574
    230
    These were designed to operate solenoid valves for a touch fountains. Outputs also went to MED Associate's data-acquisition/control interfaces. The two attached circuits were drawn up when I needed to troubleshoot some problems with them. These were designed by someone else...before me. But, they have work reliably for maybe 20 years.

    Ken
     
    burtS likes this.
  12. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Gee Ken,
    I hate to be a killjoy, but the input is a transistor that has a 3.3k resistor from emitter to ground - and the rat's resistance determines the rest of the current. If you had a really low resistance rat (say, soaked in saltwater wearing golden slippers with a gold tooth), you could get up to ~8.48mA current flow. That's nearly 17,000 times as much current as the rat can detect (0.5uA).

    If the current needs to be limited below .5uA and the supply voltage is 24v, then the resistance between the supply voltage and ground needs to be > 48 megOhms to get below 0.5uA current; that includes the resistance of the rat, which at this point is an unknown.

    I'm assuming that the rat resistance will be pretty low; < 20k Ohms.

    The only way I see this working is to use a FET or CMOS input comparator with a very large resistor to ground to keep it from "floating".

    [eta]
    I'm having trouble finding even a FET-input comparator or opamp that has a max Ib specification less than 0.5uA over temp.
    If you don't know what Ib is, that's the input bias current. Good intro on that in this PDF:
    http://www.analog.com/static/imported-files/tutorials/MT-038.pdf

    Now, the AD549 comparator is low enough, but they are mighty proud of them:
    http://www.analog.com/en/all-operat...mplifiers-op-amps/ad549/products/product.html
    AvnetExpress carries the JH suffix for ~$25.43+shipping.

    But it would help a great deal if BurtS can reconfirm that the rat can sense as low as 0.5uA - are you certain it's that low? Microamperes? 0.000,000,5 Amperes? Or is it 0.5mA, which is milliamperes; 0.000,5 ampere.
     
    Last edited: Jul 27, 2011
  13. KMoffett

    AAC Fanatic!

    Dec 19, 2007
    2,574
    230
    Like I said, these are not my circuit designs. The 6.2V Zener and the 3.3K just set the emitter at a 20V reference point.

    Ken
     
  14. KMoffett

    AAC Fanatic!

    Dec 19, 2007
    2,574
    230
  15. someonesdad

    Senior Member

    Jul 7, 2009
    1,585
    141
    I wonder about the 0.5 uA spec too -- the usual human threshold to detect current is around 1 mA (which is why leakage current specs are often set to 500 uA).
     
  16. wayneh

    Expert

    Sep 9, 2010
    12,100
    3,034
    It's doable with the right op-amp. But yeah, I wonder if the OP meant 0.5mA instead of µA. I don't think his transistor solution would be working otherwise. And rats would really hate chewing on foil.
     
  17. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    The article Ken linked to in #14 mentions commercial units that have 50 MEG inputs, which agrees with what I was basing the > 48 MEGs on as per the OP.

    The circuit they used had a 741 with 10 megs on top of that. That could result in up to 2.4uA current.
     
  18. KMoffett

    AAC Fanatic!

    Dec 19, 2007
    2,574
    230
    The sub uA current is what you want. You don't want the rat to sense the current. What would you do if every time you tried to active a drinking fountain you felt a shock on your tongue. ;)

    Ken
     
  19. WellGrounded

    Member

    Jun 19, 2011
    32
    2
    Are you trying to detect or trap the rodent, because in most cases you want to trap the little guy?

    I would make a cage enclosure with the spout protruding into the cage. The activation would be without all the fancy, smancy electronics. All you you require is a walk in platform into the caged hinged at the floor entrance. The platform sits on top of a microswitch. The entry door has a drop down door hinged at the entrance ceiling that is is held open at the ceiling by a latch that is part of a solenoid assembly(not hard to custom make). The latch can be retracted by activation of the microswitch connected to a power source. When the open door is held in the "Open" position there is also a spring between the latched door and the ceiling of the cage. This will allow the door to shut much faster when when the solenoid is activated and retracts the holding latch. Some of these rodents are super fast!

    Make sure the width of the trap is only wide enough for the rodent to go through without having the ability to turn around quickly. The door itself can have a small rectangular cutout so that when it closes it raises up a latch that has an angled front piece upon closing but the latch locks the door closed after the full closure of the door. The locking latch can have a locking spring on it like a pawl on a come-along(ratchet pulley).

    Since there is no electricity flowing until the rodent activates the microswitch they can't sense any (micro)currents.

    If you wish you can also put a secondary microswitch on the door locking latch that holds it closed to disconnect power from the power source. This will matter on your power source, but most solenoids of this type only draw milliamperes, so it is a matter of how technical you want to get.

    The secondary microswich on the locking latch could also be your activation signal that lets you know you nabbed the little bugger.

    To calibrate the weight of activation simply place a tiny spring (if needed) between the walk in platform and the base of the cage or add weight to the platform. The longer the entry the more accurate the calibration.

    You can always use food as bait near the spout.

    Happy hunting.

    Danny
     
    burtS likes this.
  20. wayneh

    Expert

    Sep 9, 2010
    12,100
    3,034
    OK, so we're off topic, but why can a rat detect a current so much smaller than a human? Or, why have humans lost the ability?

    Back on topic, if that 741 circuit works, a high impedance op-amp will do a fine job. Just about any op-amp (or the LM339 comparator I diagrammed) will beat the 741.
     
Loading...