Here's a brute-force approach in action:
The normally-closed spring-loaded lever switch on the left is meant to simulate the pedal switch. The circuit is powered by 5V DC. The circuit drives a 5V DC relay module (about $1 on eBay, free shipping from Asia). The COM and N.C. relay contacts would be used to drive the secondary circuit; the relay fires for about 0.2 sec when the pedal is pressed and again when released. You can hear the relay firing or watch the red LED on the relay module turn on and off.
This uses 3 garden-variety digital logic IC chips (maybe 25-50 cents each) and a dozen or so resistors/capacitors...so maybe $3 or so in parts including the 5V relay module. The "trick" if anything is dealing with what I call the re-trigger holdoff requirement. When pedal is pressed, the relay fires for 0.2 sec (settable by changing a resistor value). But if the pedal is instantly released, you must wait some interval (settable by changing a resistor value) before firing the relay again. This allows the secondary circuit to "digest" the first relay action so the subsequent relay action is recognized as a separate command. This trigger holdoff is demonstrated in the video where the lever is quickly pressed/released. The circuit queues up the release such that the second relay firing is delayed. A related requirement is to inhibit multiple triggers which is shown in the video when quickly pressing/releasing the switch...yet only one pair of relay firings goes thru.
Going thru the circuit logic is tedious so I'll leave it at that and in any case it appears this is not an O-gauge train application (?). There are some i's to dot and t's to cross if someone were to actually attempt to build this or commit to a PCB design from OshPark or wherever.
This could be cobbled together using eBay $3 timer/relay modules and a handful of loose components (diodes, resistors, capacitors) but would be fairly bulky and run over $10.
Bottom line. If an Arduino or similar is a viable option, that's what I recommend with this one!
