1

I've seen devices that detect a human chain. For example, they detect when 1 or more people (possibly several) hold hands to close a circuit between two distant metal objects. I've tried to find instructions on how to build a circuit like this, but without success. The one's I've found so far are $300, which is too expensive for my application because I may need a few of them.

Can anyone point me to a circuit diagram on how to build one of these? It could either be a "stand alone" circuit with a simple 0v/5v output or it could be implemented using the GPIO's or ADC's of something like an ESP32.

UPDATE: I found this example using a Darlington resistor. I'm a bit suspicious of it, but given that I'll have a micro controller that can vary the input threshold that could allow me to "tune" it. Any option on the pros/cons of using a Darlington like this instead of an Op Amp like MattMan suggested

Mark Stevens
  • 57
  • 7
  • I assume it is just an ohm (resistance, current) meter with a threshold. – Oldfart Jun 20 '19 at 05:27
  • This sounds like a "room escape" game prop to me. And given that you may need a few of them, it tends to add stronger suspicions. The price makes it even stronger, still. (Something like [this](https://www.frightprops.com/human-circuit-puzzle.html).) Is that what you are asking about? – jonk Jun 20 '19 at 05:31
  • 2
    I'm thinking that this will be hard to get to work reliably with DC. Each person has several megs of resistance, more with dry hands. I would use AC, maybe a few kHz. Output on one metal pad, sense it on the other pad. Line frequency (50 or 60 Hz depending on where you live) pickup is everywhere, you will need to ignore it. IT MUST BE SAFE, as benign as possible. Battery powered and 1M resistors in series with each pad. It will likely take some experimenting to get this to work reliably. – Mattman944 Jun 20 '19 at 07:55
  • 1
    @Mattman944 I would think the higher the AC freq the better? – hekete Jun 20 '19 at 11:05
  • Jonk, you are correct. It is for an escape room and if I needed just one I'd buy it from Fright Props. Needing to have a few of them makes that a bit painful. Mattman944 that's a good suggestion on AC rather than DC. – Mark Stevens Jun 20 '19 at 15:42
  • 1
    Your darlington post isn't visible. – Mattman944 Jun 21 '19 at 17:51
  • somewhat related: [A review on Body Area Networks (BAN)](https://www.hindawi.com/journals/wcmc/2017/5842310/) – Nick Alexeev Jun 22 '19 at 00:25

1 Answers1

2

Here is a DC circuit that I built and seems to work. So, oldfart was right. I tested it with two 47M in series plus my own hand to hand resistance. My fingers work really well, other parts of my hand/arm do not, dry skin has a very high resistance. You should test the circuit with a lot of different people.

The cap must be very low leakage, ceramic or film, do not use an electrolytic. The op-amp must have extremely high input impedance, anything with a BJT input isn't going to work. The worst case offset voltage will change the threshold significantly, but it should still work. You could try to find an op-amp with lower offset voltage.

I recommend running on batteries.

The circuit takes a long time to settle out at power-on, like one minute. So, if you are driving a solenoid, you may want to add a lock-out on power-on

Sorry, Circuit Lab keeps crashing on me. This is the best I can do right now.

enter image description here

I still believe that an AC circuit might have advantages with dry skin. But, I couldn't get anything to work with a solderless breadboard. I had 60 Hz pickup, the circuit wants to oscillate, etc.

Mattman944
  • 13,638
  • 1
  • 19
  • 43
  • Thanks for the reply Mattman! I'll order some parts to give it a try. Did the LMC6482 op amp work well for you? Is there an op-amp you recomend for the comparator? As I'm looking at the circuit, I'm wondering about two things. First, how to get it to settle down faster at power-on. Second, how could I "tune" it for the situation? Maybe a variable resistor on R3? I'm also wondering how much noise on the 5v power supply would affect it. Possibly have the 5v inputs tied together to a capacitor to smooth out any incoming noise? Not quite sure what size to go with there... – Mark Stevens Jun 21 '19 at 16:31
  • The LMC6482 is what I used, it is the only ultra-high impedance op-amp that I could find quickly. It is a dual, so it could also be used as the comparator. Many true comparators have an open-collector output that makes it easier to drive whatever is next. If you are going to adjust anything, R5 is a better choice. Put a small cap across R6 if you do this, pots are noisy. Always use de-coupling caps on the power. 10 uF electrolytic and 0.1 uF ceramic would be good. – Mattman944 Jun 21 '19 at 17:15
  • C1 is somehow charged slightly on power-on. It takes a log time to discharge completely. Changing the value of R3 or C1 isn't recommended or you could lose sensitivity or be too sensitive to noise glitches. Adding a circuit to discharge C1 could be done, but this part of the circuit is very sensitive, so you could easily disrupt it. – Mattman944 Jun 21 '19 at 17:18
  • Also note: If you are using an analog input on a microcontroller, you don't need the comparator at all. – Mattman944 Jun 21 '19 at 18:30