Dual Coil Latching Relays

[Joseph Sroka-10.2.8]

In another posting I asked for ideas on how to add red and green LEDs to
the switch tracks of an N Gauge model electric train. See below for my
earlier post.

It seems like the way to go would be with DPDT Dual Coil Latching Relays.

I would like to obtain a few (< 10) DPDT Dual Coil Latching Relays that will
work with both AC and DC impulses to the coils.


Digikey, current catalog page 1229, has

24VDC 3,200 7.5 2 2 16.8/16.8/48 255-1068-ND

a DPDT dual coil latching relay.

A couple of questions:

1. The 16.8/16.8/48 is in the Pickup/Dropout/Max. Voltage column of the
catalog page. Why isn't the middle entry a "-" as it is for some of the
other dual coil latching relays? Does this mean that the coil must have
a continuously applied voltage of at least 16.8 volts? This doesn't seem
to make it a *latching* relay.

2. All of the relays on page 1229 state that the coil voltage is DC.
Could some of these also be used with AC, if the AC is just a momentary
pulse, say for a relay with a 16.8/‹/48 entry in the Pickup/Dropout/Max.
Voltage column?

3. Does anyone know where I might obtain an AC/DC DPDT relay that holds
the setting with no voltage other than a pulse to the coil?

A self-latching relay seems like the simplest solution, except for the
questions that I have (above).

My previous post:

***********************
I have an N-gauge train with some remote-controlled switch tracks,
"snap-switch" type.

This is an older style train, not the DCC type.

I would like to add red and green LEDs to the switch tracks. The switch
tracks are operated from the "accessories" terminals of a small power
supply which are nominally 17 V AC, (I measured it open-circuit at 17.7 V
AC).

The switch tracks are operated by pressing momentarily on a momentary SPDT
switch.

I'd like a simple way to hook up the LEDs either at the remote location or
at the SPDT switch. Since power to the switch track is applied only
momentarily, I suppose that some sort of latching mechanism is needed.

The 17 V AC is part of the complication, it seems.

For reference, the switch tracks are Atlas #'s 2580 and 2581; the power
supply is Model Rectifier Corp. Railine 370N.

The switch tracks are apparently solenoid powered.

The power for the tracks only is 15 V DC. I wouldn't want to power the
LEDs from that terminal pair because it turns off when the train is
stopped, and includes the speed control.

Anybody here ever do this stuff with this type of remote turnout?

My guesses are some sort of relay or solid state relay, but I have no idea
what the specific part number or circuit would be.


*****************

Thanks.

--- Joe

--
------------------------------------
Delete the second "o" to email me.

 

[Dave]

Hi Joe,
I used some latching relay's on my layout and found that they required
more current than the Model RR power source was capable of supplying.
To overcome this problem I found it necessary to use a capacitor
discharge circuit.
This circuit is also used on some RR's to drive the switch solenoid,
run AC into bridge rectifier through a resistor to a suitable
capacitor. The voltage across the capacitor is then used to pulse the
latching relay and the resistor will limit the current from lowering
the RR power supply voltage.
Dave

 

Joseph Sroka-10.2.8 wrote:

2. All of the relays on page 1229 state that the coil voltage is DC.

Could some of these also be used with AC, if the AC is just a
momentary
pulse, say for a relay with a 16.8/‹/48 entry in the
Pickup/Dropout/Max.
Voltage column?

Put your ac thru a diode and you have chopped dc. This should work, but
you'll need to apply a higher v than rated, maybe 2x.

Larger relays are as happy on ac as dc, though they need 2x the v at
ac. Small relays however are too fast responding to ignore the 50Hz,
but can be run on chopped ac if latching.

3. Does anyone know where I might obtain an AC/DC DPDT relay that
holds
the setting with no voltage other than a pulse to the coil?

bistable relay

A self-latching relay seems like the simplest solution, except for
the
questions that I have (above).

I doubt it, logic is usually much easier implemented electronically
than electromechanically.


NT