Fortunately, in Washington State, that's no longer true -- they're required by law to adjust sensors to detect bicycles. If they want to avoid false positives from adjacent lanes, they need to upgrade their hardware. An extra signal cycle that isn't needed isn't nearly as dangerous as forcing cyclists to make a left turn across oncoming traffic.

I get what you are saying but I think you missed my point. All of the lights are set so that they don't detect trucks on the crossing street to avoid too many light changes with on-demand lights. I doubt that you would find a single light, even in Washington, that is set sensitive enough to detect a bicycle that isn't lined up along the long leg of the induction loop or, at least, very close to the long leg of the sensor. New equipment would be great... although replacing induction loops with motion sensors hasn't been all that helpful in my experience... but we are living in the Age of Sequestration

The sensor is not necessarily at the intersection mouth, they are ideally AHEAD of it and sudden proximity could possibly trigger cycling if no flow is detected on the other road.

Some intersections have CAMERAS, no?

We'll know for sure when 911 dispatch is moved out of city hall and consoliated with county operations. Urban myth or not I'm not trusting my life to a sensor vs. common sense and I wouldn't run a light in the first place.

I always operate on the notion that nobody actuallky sees you and thinking, acting like they do is a deadly mistake. You never rely on others for safety, you work to maintain your own.

You're right about figure-8 loops (technically called quadrupole loops), but not about circular loops. Although it is true that there are "double wires" in the slot leading from the loop to the box, the current in those wires is in opposite directions, resulting in no magnetic field over them. For round and square loops both (which are dipole loops), you need to have your wheels directly over the wire in the loop itself.

The only loop configuration that can detect vertical conducting objects (such as metal bicycle wheels) anywhere within the loop is a diagonal quadropole, where the center wire is on a diagonal. This can be as simple as a round loop with a diagonal cut with the wire wound in a figure-8 or as complex as what Caltrans calls a Type D loop, which is more or less a square loop with two diagonal cuts.

A good starting point for getting faulty sensors adjusted:
https://www.phred.org/~josh/bike/SignalDetection.html
(see the Presenting Your Request section)


Thank you for your guide, BTW. I followed it, actually had some reps who gave enough of a crap to fix stuff, and OH MY GOD does it make commuting life easier.

I don't agree. A quadrupole loop has the current on the center leg traveling in opposite directions for each of 2 wires that make up that center leg because of the way that the loop is configured. The 2 wires actually increase the sensitivity as opposed to a dipole loop. The magnetic fields are additive and don't cancel each other out. The wires leading from the detection box to a circular loop would work the same way.

Yes, the magnetic field is additive, but it is additive for current in two wires flowing in the same direction (and subtractive for current flowing in opposite direction).

Now apply that principle to quadrupole loops. This is a detail from Caltrans Standard Plan ES-5B showing the winding detail for what they call a Type Q loop:



Loops run on radio frequency AC current, but assume for the moment that the current is flowing in the direction of the S and F arrows. If you follow the the S arrow, you will see that it points up in the left wire, down in the center wire, and up again in the right wire. Thus the current is flowing up in the left and right wires and down in the center wires. The total current in the center wires is double that in the upper and lower wires, meaning that the magnetic field is double also. That is why the center wires are the "sweet spot" for quadrupole loops.

Also, the S and F arrows in the lead-in wires are pointing in opposite directions, making the net current in the lead-in zero.

Ah. I see the problem. I've never seen the wiring diagram for the loop. I have seen exposed wires at intersections when the pavement has broken out and I've seen only two wires. The diagram you attached shows me my mistake. The wiring reverses the flow in the loop.

Bob, so are the Quad loops the newest designs?

I started this thread with a picture of confusion of a location where there are round loops and Quad loops... and in my local San Diego neighborhood the overlay is even worse... I could never figure out which loop I should be trying to trigger.

Sometimes it can be several minutes from sensing vehicle to a light signal change. That uncertainty can make such a wait difficult.

The biggest improvement that could be made to sensing systems is a method of immediate feedback that the vehicle has been sensed.

I've seen pedestrian cross request buttons with an LED over them to show they have been pushed. That location for an LED would be a start for thru lanes, but not sure of a good LED location for indicating left turn lane sensors.

Not for bicycles, no. There is new language in the California Manual on Uniform Traffic Control Devices that, for all new detector installations, bicycles that are anywhere in the "limit line detection zone", which is a 6'x6' square with its front edge at the limit line, need to be detected.

The only loop configurations that can detect bicycles are diagonal quadrupoles, such as a Caltrans Type D (which was invented by 3M in the 1980's) or a quadracircle (which was invented in Palo Alto in about 1990).



You need to look to see if you can tell which loop is the newest. Look closely at the loop sealant covering the slots. You may be able to see a continuous strip of loop sealant in one direction or the other. I don't see the photo attached to your original post any more, but when I looked at it a few days ago I couldn't see enough detail to tell which strip of loop sealant was continuous. You may be able to tell in person.

By the way, you may want to look at my FAQ entry at traffic loops for the rec.bicycles archive.

That is largely the problem... as the street wears and gets dirty, it is difficult to tell which loop is newer or active.

The photo in the original post is just representative of what I see from time to time... and this is on regular residential streets. Newer streets with BL do have loops in the BL. But I live in a 60 year old neighborhood.

Fortunately for me the traffic in this area often opens up enough for me to run the light, or a car comes along and triggers it... otherwise this would be a very very annoying problem.

Unfortunately genec problem sensor doesn't appear to be a quadrapole sensor. With a square dipole sensor, I can trip them by riding directly over one of the longitudinal wires. Could riding or, rather, rolling your wheel around part of the circumference of these loops trip the signal in the same way?

I agree that you need to determine which loop seems the newest but ...

as genec points out, this is not always possible.

I'd say at this point, genec, that you'll just have to experiment. If you can find a sensor that hasn't been cocked up, you can determine how to trip the light. Since the whole idea of the inductive loop is the same no matter how the loop is laid down and given that the box type loops work best for me when I travel as far as possible on the long leg, I'd try rolling the wheel around about half the circle. It may not work but then again... Then apply that technique to the above sensor and determine which loop is live.

In California, the requirement for limit line detection zones kicked in on September 10, 2009, and is now incorporated in the CA Manual on Uniform Traffic Control Devices. After that date, if any substantial changes were made to the detection at a traffic actuated signal, then limit line detection zones are required. Perhaps informing your city staff of that would be helpful.

Here are the relevant definitions in in the CA MUTCD:

The implementation details are in Section 4D.105(CA) of the CA MUTCD.

The implementation details are in Section 4D.105(CA) of the CA MUTCD.[/QUOTE]

Hey I'll try the "hot zone" described above... although that is just about where I would be anyway.

Yeah they have a nice street problem reporting system here, and I have seen it work effectively... I actually got a video system installed at one light near my neighborhood... and it is a dream to use. (talked to the engineers installing it and it was cheaper to do this than to rework the streets...)