Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Automatic Inductive loop detector tester

Status
Not open for further replies.

Andymac

New Member
I am not a experienced designer for circuits and was hoping someone in here could assist me with a project i wanted to do. Background:- In the field of traffic signals testing and commissioning we have to check the correct inductive loop is connected to the correct input in the traffic signal controller. to do this it take two people, one to pass a metal object over the loop position and another to look at the LED's on the detector pack and see if they can identify the correct channel of input being triggered by the metal object. There are problems on large sites with communications between the two engineers and the triggering of other inductive loops by traffic and hence possible false identification. To overcome this i proposed to build a circuit that would produce a variable pulse pattern either mark to space variable or applitude or patterns of pulses to make the identification easier to differentiate from the background detections. I also wanted to mount this in a traffic cone so that we could place the traffic cone over a corner of a loop and the cars would avoid it. this would save have two engineers on site. of course this unit would have to be battery fed, not to bulky, able to transfer its pulse via a loop on the base of the cone to the loop in the ground ( note most inductive loops in the ground are arouns 15cm depth and around 3 to 5 turns). any help would be greatly appreciated.
 
??Do you want 5 to 10 different traffic cones, each a different color or number?? Place a cone near each loop.
OR
??Do you want only one cone and you place it near one loop and check if that loop is responding??

I think you want a very 'hard' product that can with stand a car.
OR
You want a very low cost product that gets replaced.

I have a inductive loop to detect cars in my drive way. I don't think a small loop in a cone will talk to it. I could give it a try.
 
??Do you want 5 to 10 different traffic cones, each a different color or number?? Place a cone near each loop.
OR
??Do you want only one cone and you place it near one loop and check if that loop is responding??

I think you want a very 'hard' product that can with stand a car.
OR
You want a very low cost product that gets replaced.

I have a inductive loop to detect cars in my drive way. I don't think a small loop in a cone will talk to it. I could give it a try.

I think you have missunderstood me. the inductive loops that i am testing are already in the ground and operating back to a controller which has a input outputbaord. what i need is a transportable device that i can sit on the corner of the loop in the ground, set up a particular pulse pattern and then return to the cotroller and look at the inductive detctor pack that is attached to the loop i am testing and look for the unique pattern of pulses and so identify that this particular loop is correctly connected to the input it was design for. so it is not a loop detection unit i am after, its a pulse generator that can transfer its pulse to a underground inductive loop so i can find it in the controller. the fitting this to a traffic cone was an idea to have a familiar object that drivers would recognise and avoid.

cheers
 
You may be able to do this by applying the desired signal to a large magnetic rod or C core (C might be best since you could place the open edges of the C on the ground directly over wire for best magnetic coupling), wound with turns of wire, and placed over the buried wire (core parallel to the wire). You would likely have to experiment with the core size (the longer the better), the number of core turns, and the signal voltage, to determine the best design. The idea is to have the wire loop act as a secondary in a transformer where the core is the primary.

Do you have a traffic loop you can experiment with? If not, you may have to make one, but that doesn't sound too difficult. It wouldn't have to be buried, just the same size and number of turns.
 
Last edited:
OK If you are not measuring the inductance of the loop and looking for changes but looking for patterns from cone then that is doable.

Coils have a amp turn thing happening. To transmit loudly with little power I would build a coil with many turns. Probably start out using a small spool of magnet wire. Drive the coil with a 2 watt audio amp. Use a 1khz oscillator.

On the receive side a simple audio amp and head phone or speaker. I think you will be able to hear the audio tone from the cone. In the receive side we may need a simple filter that will pass any frequency we want like 1khz to 10khz but blocks 50/60hz hum.
 
Thank You guys for the suggestions very helpful. Can anyone suggest a circuit for the pulse generator side? I am looking for something simple but robust that can deliver either a signal that can be changed for amplitude and mar to space ratios easily so i can generate an identifiable signal to follow or that generates a pattern of signal (3 long pulse and followed by 2 short as an example) or both. It doesnt matter what the signal is as long as i can easily follow it. Oh and as its a portable "in the field" piece of kit battery operated and as low power as i can get away with. as suggested above i would be driving the signal into a coil and trying to pass this signal into the buried loop like a transformer primary and secondary winding. ANy help with this is very much appreciated as I do not have the design experience to develope the circuit but do have the skills to build it. total respect to those of you who do have the vision and expertise to come up with this black magic.

Regards Andy
 
I'm a little confused - are you watching an existing indicator blink in the traffic light control box to find the coil? And by watching for one of the indicators to blink out a special pattern, it would show you which one of the buried coils has this cone gizmo sitting on top of it?
 
Last edited:
yes that is more or less what we do. The inductive loops we wish to verify are normally cut into the road at a depth of around 10 - 15cm. the loops themselves are around 3 to 5 turns in that slot cut into the road surface. these are connected via a cable (via a joint to the loop cable) to the traffic signal controllers I O via detector packs which turn the changes in inductance cause by a vehical driving over the loop into a voltfree open and close contact input which is connected to the controller IO. what i wanted to do is to allow a single person to test these loops with confidence back through the detector pack into the controller IO without the need for additional people to "wave" a metal object over the loop to simulate a car. If i could put say a traffic cone over the edge of a loop with a elctronic circuit that produces a pulse pattern i could look for that initially on the detector pack to verify its connected to the correct pack and then using a laptop check the input right through to the input within the configuration software hence fully verifying the detction into the software through the hardware. my problem is that i need a circuit to develop the pulses and a means of transfering the signal to the buried loop using a portable battery operated circuit and a portable loop attached to the base of the traffic cone to act as a 'primary' widing to the inductive loop 'secondary'. I hope that clears up what i am trying to achieve for you. If you have a better idea than using a winding to impart the signal across please let me know.

regards,
Andy
 
I would start by using Carl's suggestion based on inductive coupling. My only understanding of inductive traffic sensors is the inductance decreases when a vehicle (metal mass) is over them. Not knowing the actual real world numbers makes this difficult but I would start by pulsing a coil and seeing what happens as to any inductive coupling. I believe you mentioned an active channel is shown by a LED. I guess starting crude and looking for any results, then working from there. Interesting project and a really good idea to work on.

Ron
 
Using the car detector I have; It can not detect pulses injected into the coil by a second coil unless they are very strong. The detector I use, makes a tuned oscillator out of the coil in the road. When a car enters the coil the inductance changes and thus the frequency of osculation changes. Add energy from a second coil does not effect the frequency. Actually shorting out the coil in the cone has some effect and might work!

It might work backwards! Make a cone that detects the presents of a road coil that is energized and blinks. At the control point unplug coils until you find the one that stops the cone from blinking.
 
Carl's got the orientation wrong - Andymac's coil is oriented with North-South straight up and down, he needs his cone's coils wrapped around (or inside) the base so the poles are oriented vertically, same as the ground coil. It does not need or want a ferrous core.

I've had experience spoofing a metal detector, and simply blasting it with a signal doesn't work (like Ron's saying. And by "strong" I think he means "about to burn it out"). If you want to try this anyway, it's actually pretty simple: wind up about 8 ohms of 28 AWG around the base of a cone (about 30 turns, or 100'). Connect that coil in place of one of the speakers on an amplified MP3 player (using a good big amp). Play a tones (maybe 1-20khz) through the coil on the cone positioned over the ground sensor.

What I found was the detector's signal needed to be read, then re-broadcast out of phase. This opposing field damps the amplitude of the metal detector's signal. You would think the mode of detection would be a frequency shift from a change in inductance (with the sense coil being the reactive component in a free-running oscillator) and that imposing a strong harmonic (like with the mp3 player) would trip it - but what happens is the imposed signal just increases the overall energy in the detector's tank circuit.

It was a decrease in amplitude that was required, and that needed some finesse. For this you would first need to detect the signal from the ground loop with the coil in the base of the cone, and sync to its phase and frequency. Then drive the cone's coil 180° out of phase, pausing every few thousand cycles to re-sync.

Andy - this is getting a little out of the hobby electronics zone here, have you considered hiring an engineer? It sounds like a good idea, you could probably sell these things.
 
Last edited:
I did not mention the "drive out of phase" because of the complexity of that.

The idea of "shorting out the coil" may work. I think I could make a rubber mat with 100 turns in it. The cars can drive over the mat. The coil would be "shorted" then "opened up" every two seconds.
 
Thanks for the help Duffy. The idea is sound as it free up man power and should make identification of the loop almost idiot proof. I take your poit that it could be developed commercially using a hired in engineer, but at this point in time there is no development money.... just me trying to find solutions to help my team. Looks like i will have to try some trial and error see what happens. Do you have a circuit diagram for a frequency generator that i can build with adjustment so i can alter mark 2 space and frequency, and amplitude or pahse shift?
I suppose I didnt expect the complexity of trying to do something that on the surface of things seems straight forward. thank you all for your help so far.

Andy
 
It can not detect pulses injected into the coil by a second coil unless they are very strong.
That is going to be the most difficult hurdle to overcome. Presumably the area of the buried coil is the area covered by several vehicles? So any cone-sized coil will have an extremely low coupling coefficient with the buried coil.
I like Ronsimpson's idea of working this backwards. Disconnect a coil from the controller and inject into it a (legal) RF signal. Use a portable radio to pick this up at the roadside. There's a beneficial side-effect: the guy with the radio gets some healthy exercise going from one loop to another. :D
 
Carl's got the orientation wrong - Andymac's coil is oriented with North-South straight up and down, he needs his cone's coils wrapped around (or inside) the base so the poles are oriented vertically, same as the ground coil. It does not need or want a ferrous core.
.........................
My scheme is to induce voltage in the wire per Faraday's Induction Law as if it were a straight piece of wire, not a coil, since the coil is so large that is would appear as a wire to the cone coil sitting on the wire. And a "C" magnetic core will increase the magnetic flux coupling to the wire.

But you are correct that the orientation is incorrect. The core should be relatively small and placed such that the C is perpendicular to the ground coil wires with the ground wire centered in the open end of the C so the flux is cutting the wire, not parallel to the wire as I originally stated.
 
Thank you everyone for your help, sorry its delayed i was off sick the last few days and not near my computer, think the computer gave me the virus in the first place! going to try some experiments with some of the ideas you have given me and see where i get. may be back when i get stuck guys thanks again.
Andy
 
That is going to be the most difficult hurdle to overcome. Presumably the area of the buried coil is the area covered by several vehicles? So any cone-sized coil will have an extremely low coupling coefficient with the buried coil.
I like Ronsimpson's idea of working this backwards. Disconnect a coil from the controller and inject into it a (legal) RF signal. Use a portable radio to pick this up at the roadside. There's a beneficial side-effect: the guy with the radio gets some healthy exercise going from one loop to another. :D

I made a RF wire finder. It works with a AM radio. I locate wires 3 feet down. Last summer I was trying to find a break in a power cable underground. It can not find the brake because the RF jumps over the brake. I also found that the RF will jump form one wire to another through the capacitance between wires. I think the wires in the stop light control box are in a bundle and will have high capacitance. Thus all the coils may have some RF on them. AM radios have a automatic gain control in them thus it is hard to know how strong the RF is with out a signal level meter. I was thinking of using a audio tone in the coil, or maybe 100khz, so the cross talk would not be as bad.
 
I actually work in this field as well and perhaps a different approach is in order. Rather than generating the signals from the loop, why not generate the signal, with a toner, from the controller side and use the handheld detector that comes with the toner to pick out the corresponding loop in the road. Just in case you need to go with the original idea, a small DC powered hobby motor with a cheap magnet attached to a rod on the output shaft would generate a frequency equal to the angular velocity of the magnet (i.e. the RPM). By the way, you will pick up the tone from the toner on all loops within the same conduit but the signal will be noticeably strongest on the test loop.
 
Last edited:
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top