Shop OBEX P1 Docs P2 Docs Learn Events
Ping))) now has a wider detection range? — Parallax Forums

Ping))) now has a wider detection range?

LucMLucM Posts: 5
edited 2012-08-23 08:44 in Accessories
Greetings!

We've been using Ping))) sensors to implement a passenger-counting system in city buses. We use a pair of sensors set up next to (or near to) the main door of the bus to determine whether passengers are going in or out. We've had our problems with this system, as we have several types of buses to contend with, and "real life" can get somewhat chaotic on a day-to-day basis, but we're working to remedy these problems with adjustments to our BasicStamp programming.

However, recently we've encountered another, more serious problem. As some of the installed sensors became defective, we had to order some new ones. We discovered that the newer sensors have a much wider detection "cone" (roughly twice as wide as the sensors we initially installed) and this makes our twin-sensor passenger-counting system practically non-functional, because either sensors can now detect passing objects in a random manner. We need Ping))) sensors that have a narrow detection cone, which is what we had up until recently.

Is there any way to adjust the detection "cone" in Ping))) sensors?

Thanks,

Comments

  • W9GFOW9GFO Posts: 4,010
    edited 2012-08-10 11:24
    I don't know the answer but since it has been three hours and no responses I will offer this suggestion. A baffle around either the sender or receiver (or both) may restrict the cone of sensitivity. It would be a simple experiment to wrap a piece of tape around the transducer to test it out.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2012-08-10 12:20
    Rich, contrary to one's initial intuition, a baffle will not narrow the cone, only extend the point at which it begins to spread; and a narrowed aperture will actually make the cone wider.

    Luc, it's odd that the cone would change with the new units. Do the old and new units have different rev numbers? If so, contact Parallax tech support, give them the rev number info, and ask them what, if anything, has changed. It may also just be a matter of overall sensitivity, since the cone angle is a function only of the ultrasonic frequency and the active diameters of the transducers.

    -Phil
  • LucMLucM Posts: 5
    edited 2012-08-10 12:29
    Rich, contrary to one's initial intuition, a baffle will not narrow the cone, only extend the point at which it begins to spread; and a narrowed aperture will actually make the cone wider.

    Luc, it's odd that the cone would change with the new units. Do the old and new units have different rev numbers? If so, contact Parallax tech support, give them the rev number info, and ask them what, if anything, has changed. It may also just be a matter of overall sensitivity, since the cone angle is a function only of the ultrasonic frequency and the active diameters of the transducers.

    -Phil

    Thanks for the suggestions Phil, I'll check and compare the rev numbers next week (when I go back to the client's bus garage) before I contact Parallax for info. All I can say for now is that I can really see the difference when I approach the palm of my hand near the sides of the cone. The cone really is wider on the newer sensors. Way too wide for our operational specs.
  • W9GFOW9GFO Posts: 4,010
    edited 2012-08-10 13:45
    Rich, contrary to one's initial intuition, a baffle will not narrow the cone, only extend the point at which it begins to spread; and a narrowed aperture will actually make the cone wider.

    Sounds reasonable. Does the same apply for a baffle on the receiver?
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2012-08-10 14:06
    Yes, the same applies to the receiver.

    -Phil
  • PublisonPublison Posts: 12,366
    edited 2012-08-10 14:22
    While you wait for a reply from Parallax, check out the programmable PING to see if it also might fight into your program:

    http://gadgetgangster.com/find-a-project/56.html?projectnum=138

    Programmable range, sensitivity, and output power.

    I don't work for them, just thought it was a great product.

    Jim
  • Duane DegnDuane Degn Posts: 10,588
    edited 2012-08-10 19:17
    Maxbotix makes sensors with a variety of detection cones. Here's a page showing the beam patterns made by the different models.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2012-08-10 19:34
    I would not give up on the Ping))). My suspicion is that it's nothing more than a sensitivity issue. If the sensitivity has increased, echoes from the periphery, which didn't register before, will have a greater chance to make their presence known. One can cut down on the sensitivity by mechanical means (e.g. open-cell foam adhered to one or both transducers) to, at the very least, test this hypothesis.

    -Phil
  • John AbshierJohn Abshier Posts: 1,116
    edited 2012-08-11 07:10
    The Ping Reflector mounted vertically may help.

    http://www.gadgetgangster.com/find-a-project/56?projectnum=361

    John Abshier
  • LucMLucM Posts: 5
    edited 2012-08-14 07:28
    I haven't contacted Parallax yet, and I wanted to reply to some comments here before I do so.

    - Phil Pilgrim: All the units I have here (new ones and older ones) have the same rev number (namely 'REV B'). I can't spot anything different between the newer and older units just by looking. About your sensivity-reducing foam idea, you have to keep in mind that city buses are the most hostile environments for electronic equipment, not just because of vibration and wide temperature/humidity variances, but because of "casual" abuse by everyday bus users and sometimes even disgruntled bus drivers. In each bus, we had to encase the sensor pairs (and central BS boards) into custom metal casings, and even then, the sensors extremities are still vulnerable to poking fingers (mostly by little kids who can't control their destructive impulses). Foam add-ons over the sensors probably wouldn't last a week. People just love to scratch anything that's made out of thin foam, for some mysterious freudian reason. :)

    - Publison: The programmable sensor is interesting, but just looking at the 5-pin connector tells me that such a change in technology would be expensive to implement (R&D, new wires inside each bus, software protocol adjustments, maintenance of more than one type of sensor on the field, etc.), and my boss probably won't go for it unless I can provide proof that I can actually control the detection cone with it, in a precise manner. Company politics being what they are, I'd have trouble just getting the green light from my boss to order a single unit for testing purposes.

    - John Abshier: There's no room for such a large reflector inside the buses.

    Now with all this said, I'll be contacting Parallax shortly... and keeping my fingers crossed. ;)
  • LucMLucM Posts: 5
    edited 2012-08-20 07:01
    I opened a ticket (with Parallax support) about the detection cone problem detailed in this thread almost a week ago, and I haven't heard back from anyone at Parallax yet. Does it always take this long to get in touch with tech support?
  • Ed TEd T Posts: 50
    edited 2012-08-20 09:14
    Luc,
    Just to clarify on the 5 pin connector for the programmable sensor, in Ping mode only 3 of the 5 pins are used, with the same wiring and protocol as the Ping. So you could possibly do a direct swap depending on the mechanics around the connector. With that said, I think Phil is on the easiest track with decreasing the Ping sensitivity. I have found foam works well, as Phil suggests, so maybe you can protect the foam with some wire screen? Another possibility might be a divider between the two transducers (sort of a nose to the transducer eyes) that shields the side view of each transducer in opposite directions. Also are you looking at the distance that the pings are measuring or just that they had a detection? It seems that even with side detections the distance will tell you which sensor they are closer to.

    --Ed

    PS: Jim and John: Thanks for the reference....
  • SRLMSRLM Posts: 5,045
    edited 2012-08-20 18:44
    LucM wrote: »
    I opened a ticket (with Parallax support) about the detection cone problem detailed in this thread almost a week ago, and I haven't heard back from anyone at Parallax yet. Does it always take this long to get in touch with tech support?

    I would email them again...
  • LucMLucM Posts: 5
    edited 2012-08-21 07:02
    Ed T wrote: »
    Luc,
    Just to clarify on the 5 pin connector for the programmable sensor, in Ping mode only 3 of the 5 pins are used, with the same wiring and protocol as the Ping. So you could possibly do a direct swap depending on the mechanics around the connector. With that said, I think Phil is on the easiest track with decreasing the Ping sensitivity. I have found foam works well, as Phil suggests, so maybe you can protect the foam with some wire screen? Another possibility might be a divider between the two transducers (sort of a nose to the transducer eyes) that shields the side view of each transducer in opposite directions. Also are you looking at the distance that the pings are measuring or just that they had a detection? It seems that even with side detections the distance will tell you which sensor they are closer to.

    --Ed

    Hello Ed.

    I attached to this reply a picture of the front side of the metal casing we're using, just to give you a better idea of how we're using the sensors. This is an image I whipped up from scratch in less than a minute, so the distance between the sensors is not exact compared to real life. As you can see, putting some foam and wire screen over the holes is not exactly an easy thing to implement. Also, adding a "nose", as you suggest, is not a good solution here because passengers can bump or hit the nose and hurt themselves as they go in and out of the bus. Our clients will never accept that kind of modification.

    To answer your other question, we look at the distance value when detecting passengers. If the reading falls under a certain threshold, we record it as "activity". This "activity" in front of the two sensors is analyzed to determine if a passenger is going in or out of the bus.

    casing_picture.png
    361 x 215 - 3K
  • RobotWorkshopRobotWorkshop Posts: 2,307
    edited 2012-08-22 12:52
    You mention that you are putting these sensors in a metal case. Are the metal cans touching the metal case at any point? If there is any question about them touching another metal object then I would suggest wrapping a piece of electrical tape around the metal cans for each sensor to ensure it is electrically isolated. I have one project where I am using a sonar module attached to a metal panel and since the clearance is so tight had to do this to ensure there was no unintended electrical connections. When mounting it is a good idea to use nylon spacers and/or fiber washers to ensure that the mounting hardware doesn't short out or make any connections that shouldn't be there.

    It is easy to overlook issues that are created just from mounting sensors like this and other devices. I rarely see it mentioned if mounting points are isolated or tied to the ground on the device. Also notes/warning about clearances and keeping certain portions of a device electrically isolated. Without considering that one can often create connections that they didn't intend and perhaps even create a ground loop.

    Robert
  • Ed TEd T Posts: 50
    edited 2012-08-23 08:44
    Luc,
    What about a rectangular piece in the middle of your front side that blocks the sensors from seeing the middle? Or angling the sensors outward so they have less overlap?
    --Ed
Sign In or Register to comment.