12
$\begingroup$

While listening to an FM radio station (frequencies between 88 and 108 MHz) in my car, I've noticed that as I pull up to a red light, there's an increase in the amount of noise. It's very noticeable when I'm listening to a station that's somewhat far away, so I'm at the edge of the region in which reasonably noise-free reception is possible.

As I'm approaching the red light, and stopping, the lights in the perpendicular direction are green. When those green lights turn yellow, the noise suddenly drops. Then when those lights have turned red and the lights in my direction turn green, the noise comes back. Then I go through and as I get farther from the green light, the noise fades.

It's not a single weird intersection; I've heard it in multiple locations. The noise increase and decrease is perfectly synchronized with the change of light color. I believed it was a coincidence for a while, but I've heard it too many times now. It's a real effect. Being near a green light causes radio noise. Why?

$\endgroup$
0
19
$\begingroup$

At those intersections there are sensors buried in the asphalt which detect the presence of cars waiting for the light to change. The sensor is a large circular coil of wire which is fed a high-frequency AC signal at which the coil is resonant. When a car is situated above the coil, the iron in its frame and engine block detune the resonant circuit, which is detected by the control system for the lights. The control system then changes the lights to allow the stopped car to proceed.

A car radio can pick up the sensor signal when close to the intersection, and have its ability to receive the intended signal interfered with. Sometimes it's possible to actually hear the signal shift frequency as you roll to a stop over the coil.

A 4-way intersection will have from four to eight such sensors to determine whether or not cars are waiting for a light change, or as many as 16 if the intersecting roads have left-turn lanes which also must be regulated. When the light for a given path is red, the sensor on that path is energized and carries the signal so it can tell when a car has arrived and is waiting for a green. When that car gets a green, the sensor is switched to standby and the sensor for the opposing traffic (where the light is now red) is switched on to look for cars coming from that direction.

This means the strength of the sensor signals received by your car radio will shift about in concert with the light changes as the street sensors are turned on and off.

Motorcycles with aluminum engine blocks and lightweight frames are often unable to trip the sensors. An electromagnet coil and drive circuit can be installed on such a bike, with a pushbutton on the handlebars to blip the coil. The street sensor then responds as if a car were there and cycles the lights for the motorcyclist.

$\endgroup$
3
  • $\begingroup$ I never knew the sensors in the road had an on/off cycle. It makes perfect sense now. The yellow light gives the least noise because during the yellow light, all the sensors in the intersection are turned off since the yellow light has a fixed duration. And when I get the green and go through, I get close enough to the sensors in the perpendicular street to get a little bit of noise (shouldn't be as much as when I'm stopped at the red light - I'll listen for that next time) $\endgroup$ – user218481 Dec 31 '18 at 5:52
  • 1
    $\begingroup$ Having read the answer from @OldColdDreamer it might help if Niels states which country he is writing about. $\endgroup$ – Farcher Dec 31 '18 at 10:10
  • $\begingroup$ I live in the western USA and was active as an engineer during a timespan that covered the invention of street sensors (starting with pneumatic hoses) that controlled incandescent bulbs and ending with detuned loops of various types that controlled LED's. However, I did not work in that specific field as @Old Cold Dreamer did. $\endgroup$ – niels nielsen Dec 31 '18 at 17:45
3
$\begingroup$

I am not convinced by @Niels' answer - at least as far as the UK is concerned. I worked as a software engineer for a UK traffic-signals company 1976 - 2008. The signals certainly had sensors which worked by detecting the de-tuning of the loops in the road; but the detecting circuits remained active all the time (and during the green, gave the signal-controller an idea of how much traffic was flowing).

Maybe the interference is not from the detection loops at all, but from the circuits which drive the LED lamps. I would expect the drivers for different-coloured LEDs to have different noise spectra.

$\endgroup$
3
  • $\begingroup$ I agree. The current required to get a given human-perceived brightness is different for each LED color. So the dc-dc convertors are putting out different amounts of hash. $\endgroup$ – Jon Custer Dec 31 '18 at 15:43
  • $\begingroup$ I'll try an experiment with a car radio and a large LED array I own which was scrapped from a streetlight system, and report back on my findings. Will edit my answer as required. -Niels $\endgroup$ – niels nielsen Dec 31 '18 at 17:54
  • $\begingroup$ I have a red LED signal light retired from service here in Oregon. It runs directly on 120VAC 60Hz. When operating, it does not produce any interference with radio reception on either the AM or the FM bands, even when the receiver is held directly against the lamp lens. $\endgroup$ – niels nielsen Dec 31 '18 at 19:08