The diode "snub" also suppresses the inductive kick caused by the relay's magnetic field collapsing across its coil after the circuit has opened; preventing arcing at whatever control contact had opened.

This is its primary purpose, although it will slow the collapse of the field, and thus the release time of the relay. It slows the field collapse by shunting the current generated by the collapse, back through the coil, tending to build the field back up. Slowing the rate of magnetic collapse lowers the peak voltage produced by the inductive kick by several orders of magnitude, thus preventing contact arcing; which can be a very destructive problem.

However, because the shunt path presented by the diode has very little resistance (just that of the coil itself), the action described above is fairly quick (about 300ms). Therefore, if you want a slower release with an "ordinary acting" relay, you would snub it with a resister instead of a diode. This prolongs the effect of the snub, depending upon its value. But to get a really long release delay, you would snub the magnetic circuit itself with a number of copper rings, which causes eddy currents during the magnetic collapse, producing the same effect. Nowadays, this type of snubbing is usually done in the relay factory.

Except that there are few if any new relay systems being installed. Relay use has been mostly relegated to only interface matching with outside sub-systems, since the advent of microprocessor based interlocking and signalling.

Incidently, microprocessor based signaling is an absolute prerequisite for PTC; thus we see massive changes going on all over the country, in preparation for PTC installation. The dubious part of this PTC mandate though, is that all such microprocessor systems communicate through the rails with neighboring locations (no pole lines at all), and are already capable of communicating speed commands (signal aspects) to trains directly (no radio needed). But Noooooo! Congress forced the use of an exorbitantly expensive, unproven, error prone technology, to be layered out top of this prerequisite; in place of the simple and already proven solution that will already be there anyway. But I digress!


When we wanted a latch function, such as directional sticks in APBS block signalling, or in its descendent varieties of CTC, or in approach and routing locking, etc; we would use the stick method described so far in this thread.

But in some applications, it was required to retain the last known state, even in the face of a power failure; or other reasons required the use of a non-electical latch. Mechanical latch type relays would then be used in non-safety critical uses. In safety critical applications, magnetic latch relays were available. These had a biasing permanent magnet in the coil's magnetic path, which was strong enough to hold the contacts up - if already up; but not strong enough to pull them up. Pulling the contacts up required help from the coil. Dropping them required the coil to be reverse energized, thus opposing the bias magnet and dropping the contacts.

Some CTC installations used the mag stick method at the headblock, while others used the simple stick contact method. WP had an interesting arrangement in certain locations, where the sticks at the headblocks were mag-stick relays, while at the intermediate signal location, ordinary sticks were applied. The mag sticks determined which headblock would listen and which headblock would transmit a tri-state polarized signal (plus, minus, off - (Yellow, Green,Red)) down a single HD line circuit to the intermediate signal and on to the next headblock. This arrangement allowed pole line signal wires to be reduced to only two!

In some locations, the line wires were also eliminated by using code rate relays to transmit a code directly down the rails. The mag stick then determined which end would transmit in the rails and which would listen. Nothing much left to do there, to get automatic train control (ATC, ATP, ACS.....) and even PTC, except to let the trains listen in too.


Those were the days!