Not necessarily. If the entire section of track to each side of the grade crossing is electrically isolated from the rest of the track (and from each other, of course) as a track circuit up to, say, 1000 feet from the crossing, then any connection between the rails at any point will be read as a train entering that track circuit. Doesn’t matter if it’s 1000 feet away, or only 100 feet. See?
Sorry to be a pain about this David since your posts are always spot on and well intentioned. However, you are straight out calling me a lier here: I was not a child, neither was I drunk. I can recall the place where we did the deed. It may have been 50 feet; it may have been 200 feet. It was definitely not 1000 feet. As one other has stated, the fact that the detector is triggered by a train a quarter of a mile out does not mean one cannot trigger it right next to the signal.
We’re not talking about a single black box that sits at some predetermined point, we’re talking about a long conductive path, made up of rail, that runs from signal out to the boonies where the train should be detected.
I have to ask-in this day and age, could you be considered a terrorist suspect?
It’s a lot easier to look for the insulated fish plates joining the rails.
That’s where the action is. 
People have metioned that the detection needs to be quite a long distance from the crossing so the barriers come down in time for fast trains. I can understand that. How is it that almost as soon as the last wagon of the train clears the crossing, the barriers start rising, or if there are only warning lights, they shut off?
In fact, if it’s a matter of a simple current running through the rails, through the train or metal bar, then back to the detector, it may have been NECESSARY to be much closer than the train to use a homemade pseudotrain, as the real train’s resistance might be less due to greater volume of metal and greater area of metal in contact with the rail.
Again, just because the detector’s parameters REQUIRED that it detect trains at a thousand feet or more doesnt mean it was required NOT TO detect them closer.
Adding to samclem’s warning: Part I
If you have actually done this - and it worked - it is extremely good that you ran off as fast as possible. Here’s why: by bridging both the insulated rail joins with pennies, you are not closing the circuit that operates the crossing. Not directly anyway. What you are doing is making the relevant section of line twice as long by electrically connecting it to the adjacent section. This will not activate the crossing, unless the adjacent section is already occupied. By. A. Train. Once again, good that you ran off when you did. To the pranksters, this is even more dangerous than the crowbar.
As has been explained previously, that’s just where the action starts. The crowbar will achieve the same effect from that point right down to the crossing and a little beyond.
The crossing section ends shortly after the crossing (no reason for it not to, and plenty of reason for it to do so unless you want irate motorists). When the last wagon clears it, the section is clear, and the gates will open. Of course, the last wagon is the last wagon is the last wagon, so in this case the length of the train does not matter, as it does with an approaching train. The section break need only be a matter of feet away (unless it is a single track crossing).
This is a good point, though it depends on where the rails are joined to the signalling system - it might not necessarily be at the crossing itself. As a case in point, Australian (And probably US) railways use what are called Hi-Rail vehicles for maintenance and track gang transport. These tend to be road-registered Toyota 4WD vehicles (small SUVs). They are mechanically identical to the ones you or I could buy, with the exception that they have little flanged metal guide wheels which can be lowered into place for use on the tracks (they are quite humourous to see whizzing by). The steering is locked with a pin, and the rubber tyres still provide traction. When one of these is on the line, the signallers either have to book the entire line out of use to regular trains, or manually ‘block work’ the Toyota through the signals, section by section, with constant radio contact. The reason for this is that the little steel guide wheels are quite hit and miss with regards to their ability to close the signalling circuit, and can’t be relied upon to do so. So of course, weight and resistance plays an important role - not such a worry when you’ve got locomotives with thirty tons weight per axle, but a concern for the small fry.
Which brings me to Adding to samclem’s warning: Pt II
If you are silly enough to do the crowbar trick, and the crossing activates, don’t be too eager to slaps yourselves on the back and say, “Hey, it worked!” It might not have worked. Something else may have activated the crossing. Be prepared for the sickening sound of five-chime air horn approaching at 70mph. And when scramble for your lives, take the crowbar with you. They can derail a locomotive.