0

I want to use the ultrasonic sensor under a metal grid plate like below (just 1 of those plates instead of 3). The sender/receiver will be under (partly inside) two adjecent holes (hopefully it fits) and point upwards. The reason is to detect if someone (a person) steps on the plate.

The height of the plate is about 1,5 cm and the height of the receiver/transmitter is about 0,5 cm (if I leave some space for the enclosure), maybe even less space.

In the diagram below I showed hopefully clear enough what I want. Diagram

What I'm worried about is that the receiver picks up the distance to the plate (the red lines), so it will always received just a centimeter or so.

enter image description here

Michel Keijzers
  • 13,867
  • 18
  • 69
  • 139
  • Maybe a better diagram would help because I'm not getting what the setup is or what you are trying to achieve. – Andy aka Feb 16 '18 at 11:36
  • 1
    That looks more like a grid. Aim your sensor through the holes (if it will fit.) The sensor (or the driver software) should have a minimum distance. Below that distance (time, really) it should ignore echos. – JRE Feb 16 '18 at 11:46
  • @JRE If it would ignore it and still receives a real distance to a possible person on it, it would be perfect. – Michel Keijzers Feb 16 '18 at 11:48
  • 1
    That's my understanding of how those sensors work. [The datasheet](http://www.micropik.com/PDF/HCSR04.pdf) seems to agree. – JRE Feb 16 '18 at 12:06
  • 1
    If you then stand directly on it, though, you would be too close. You have to stand a bit off to the side and let it bounce off yer butt. :) – JRE Feb 16 '18 at 12:07
  • @JRE I also thought about this ... however, I assume that when somebody will stay on the plate, there is a very high chance the first positive sensor reading is not directly from a foot but from above, and when a person steps on it, I want to keep it e.g. 10 seconds on. And if that does not work, I add multiple sensors on various places to detect better. – Michel Keijzers Feb 16 '18 at 12:10

2 Answers2

4

This will not work if you stand on it. Ultrasonic sensors have a minimum distance from which they can accept reflections, because the transducers themselves ring, so the circuit has to blank any reflections received while ringing is expected. Even with distinct transmitters and receivers, there's enough electrical and/or acoustical coupling between them that they have to implement blanking.

If you did achieve a minimum distance, you'd still have the problem that the common inexpensive sensors are completely unprotected from moisture, making pointing them up (even indoors) dubious. And no, you can't put a plastic film over them. Weatherproof sensors however are available.

Likely you will have more luck solving this problem with some other technology - light sensors (either passive or emitter/reflector), force sensors, perhaps something capacitive, etc. Or if you want to use ultrasonic, aim it sideways or down from above.

Chris Stratton
  • 33,282
  • 3
  • 43
  • 89
  • Thanks for this insight ... I haven't thought about the moisture problem which indeed can be problematic. Lights might be tricky, I don't like the grid plate movable (unless no other way), force sensors are quite expensive, and ultrasonic from sideways or from above is not realistic since it would 'stick out' the box itself. Accepted because of realizing it will not work (or at least has disadvantages). – Michel Keijzers Feb 16 '18 at 22:42
  • I wonder if a IR PID sensor (HC-SR501 would work) ... detects humans (only), and if I put it in some kind of tube it will have a reasonable low angle. – Michel Keijzers Feb 17 '18 at 16:14
  • 1
    It's PIR, not PID, and it detects *change* not *presence*. For presence you probably want a reflectance sensor or an ambient one which can be shadowed. – Chris Stratton Feb 17 '18 at 17:18
  • I assume that the person will move a bit, and I can enable it for longer time after a trigger, so I think it works probably even better. I will also check reflectance/ambient sensors. – Michel Keijzers Feb 17 '18 at 17:32
3

This sounds like you're trying to build a boogie dance floor and sense when a person is on each tile. As others have said, the ultrasonic solution is unlikely to work well for you.

enter image description here

Options:

  • A force sensor in each corner or the middle or some combination of corners.
  • Strips of safety or intruder mat around the edges supporting the frame.
  • Rest the corners on piezo transducers. Sense the voltage induced by a boogie stamp on the grid.
  • Micro-switches in each corner with the grid sprung off them by a millimetre or two.

enter image description here

Figure 1. The momentary action version of these switches, so beloved by guitarists the world over for more than half a century are mechanically robust and feature a hard mechanical stop which protects the switch mechanism below. For extra security a second nut could be added to the top side of the mounting to spread the load across more threads.

The last one, whilst the crudest, might in conjunction with a spring be the cheapest and simplest to interface with. I've recently purchased five to convert a keyboard into a footswitch presentation controller. See Soldering on keyboard controller circuit board. One thing to watch is that the mechanical switch requires the grids to move vertically by a couple of millimeters. This could present a minor trip hazard when moving from a depressed grid to an "up" grid.

Transistor
  • 168,990
  • 12
  • 186
  • 385
  • It's not for dancing, at least not in principle. It's for a musician (guitarist/singer) to stand on it and be more in focus (higher and lighted - from below). About first: I checked force sensors but these are quite expensive, and need also the plate to be movable (or the sensor on top of the plate). About second: could be a good idea, need to check into this. – Michel Keijzers Feb 16 '18 at 16:45
  • About third: it might be possible the singer/guitarist does not move that much, not enough for the piezo transducers, but maybe it does. Also it will be in a setting where there might be vibrations anyway (loud music, other people dancing elsewhere). Micro switches need the plate to be movable (although slightly). I'm afraid it will wear out over time. – Michel Keijzers Feb 16 '18 at 16:45
  • Thanks for the edit ... it has somewhat the same requirements as such dance floor, except there is only one square, and I want to have it on at least 10 seconds (or even more) once it is triggered (to prevent a 'flashing' effect). A few mm moving should work (although the one who makes the box itself rather has the plate completely fixed, also for safety when stepping up/down and rigidness of the box). – Michel Keijzers Feb 16 '18 at 22:45
  • I wonder if a IR PID sensor (HC-SR501 would work) ... detects humans (only), and if I put it in some kind of tube it will have a reasonable low angle. – Michel Keijzers Feb 17 '18 at 16:14
  • 1
    Be aware that some of PIR sensors give an output on *change* of level of received IR. If your guitarist stands still enough the output will turn off. – Transistor Feb 17 '18 at 16:29
  • I want to add a microprocessor anyway and when I receive a signal, it can stay on for e.g. 10-20 seconds. I assume when he moves a bit it will continue (otherwise I make the time 3 minutes). It does not matter so much if the light is on for a bit longer (light shining up is not invisible if no one stands on it anymore). – Michel Keijzers Feb 17 '18 at 16:50