Bill,

One wasn't. It did latch. Now to get the simulation to latch, I added a 100 ohm resistor in series with the switch....between the +V and the diode connection. This made the diode the more favorable current path.

Simulation Video at http://www.tech-service.net/joe/latched.swf.asp

 

Joe,
I don't know what's going wrong on your end, but I just breadboarded it myself using a 1N4004 diode and a Teledyne 172-5 DPDT relay; using just a single COM and N.O. terminal and an ATX-form factor PS I converted to a bench supply that outputs 5.11v.

It worked as a latching relay just as I'd predicted.

It's not the first time I'd posted a circuit that had a diode latch, either; here's one from a year or so ago:

Perhaps you have the diode in backwards, or you're trying to us the NO and NC relay terminals rather than the COM and NO connections? Bad spot on your breadboard? Dirty relay contacts?

 

OK, no fair editing/deleting your post while I'm breadboarding!

You wouldn't believe how long it took me to track down a stupid 1N4004 diode after my wife "straightened a few things up"... I'm finding transistors mixed in with diodes mixed in with IC's. They're all black with silver legs, right? Must be the same things...

 

Sorry about that Sgt....

I revisited my ckt......for a quick check.