This 10-Volt DC 1946 glass-encased relay made the lights alternate “40 to 45 flashes per minute”.
In the trackside signal cabinet it sat on springs and has two electromagnet coils which alternately energize and attract a pivoting armature.
A thick metal bar athe top of the coils spans from the LEFT coil’s metal core to a MIDDLE metal “core” to the RIGHT coil’s metal core. The MIDDLE “core” has metal washers stacked on it. They affecthe flash rate.
[||] || [||]
When the LEFT electromagnet energizes, it pulls the LEFTside of the armature up, the LEFT relay contacts close and a separate contact inside the relay basends currento the RIGHT electromagnet.
When the RIGHT electromagnet energizes enough, it pulls the RIGHT side of the pivoting armature up, the RIGHT relay contacts close and the separate contact inside the basends power to the LEFT electromagnet.
There is no timing circuit. Why does it not flip-flop rapidly?
Does it take time for an electromagnet to develop full magnetic flux?
It probably takes some time, at least a few milliseconds, but it’s likely the inertia of the moving parts that affects the flash rate the most though. I’m not familiar with this device, but if there are springs attached, that would affect the travel rate of the armature, as would any mechanical friction built into the device.
“Does it take time for an electromagnet to develop full magnetic flux?”
Yes. Any time the strength of a magnetic field changes, it generates a voltage in the coil that attempts to cause a current that cancels that change. See Lentz’s Law. That means that when you connect the coil to the DC voltage, the building up of the magnetic field generates a back voltage in the coil that causes a delay in the magnetization and the same thing happens when you remove the voltage, it generates a voltage that attempts to keep the original current and magnetization going.
I was looking for that.
Did not realize that it could take so long.
The coil cores and center “core” are magnetically coupled by a thick metal bar across the top.
. . . . . . . . ____
. coils -> [||] || [||]
middle core ^
Now must figure out how the metal washers stacked on the center pole, presumably adding magnetic mass, affect flash rate.
I just don’t see how the LR time constant could be more than one second with such small coils.
Are you sure there isn’t some bi-metallic thermal element involved?
The relay is a balance of mechanical and electrical properties that dictates its rate. The inductance can be adjusted by the copper washers to TRIM the rate. They were installed as a means to slightly adjust the rate to account for wear over time. Incrementally removing them will increase the rate slightly.
You aren’t intending to use this antique device to drive emergency lights are you?
To: circuitsmith:
The two coils are BIG.
I have notaken any coils apart buthe insulated wire is probably 1/8th inch thick.
To: TwinTurbo:
Haha - too big and heavy and too slow a flash rate.
(I already made a circuit which flashes the RED LED lights rapidly during the one-second “OFF” phase and ON steadily during the one second “ON” phase.)
I am making the relay into a desk lamp/nightlight.
From a cheap Walmart LED nighlight bulb I extracted the LED circuit.
Inside the relay glass bowl it willuminate the interior of the relay.
A switch will turn on 6 volts to the relay which will operate a Lionel Railroad Crossing signal sitting below the lamp on top of the relay terminals.
The 40 clicks per minute may be soothing like a clock pedulum.
Sounds like an interesting project. 40 clicks/min. wouldn’t be soothing to me though… I would be ready to hurl it out the window after about 5 minutes of trying to sleep. But I don’t like pendulum clocks much either. I don’t think the flasher will last long if you run it continually like you plan. And you’ll need a pretty hefty 6V power supply to drive it. Put an ammeter on it and see how much current it draws when in operation…
Should look pretty cool cleaned up with the interior lit up. As you know, it’s designed to run @ 10 Vdc so it may not function properly as low as 6. Unless you meant you’ll run the relay at 10 but use it to switch the 6 to the toy crossing light. Sounds like fun project…
To: Boblivion:
I’d operate the relay only until it became annoying. (I LOVE the TICKING of MANY CLOCKS in a clock shop. So pleasing and soothing when all at differing rates.)
(I quieted our cuckoo clock tick somewhat by placing electrical tape on the stem which moves the pedulum. Also quieted the two whistles and made the hammer strike the gong more gently.)
The flasherelay is designed toperate continuously for days.
To: TT:
It still operates on 6 volts and the click is quieter!
The (o)T(o) will be less bright and annoying on 6 volts.
The lamp has its own switch. I must find a place in the bakelite top for the relay switch and will need to run two cords to it - one for 120 VAC and 6 VDC. (Maybe I can use a cord with ground and have the DC share the neutral.)
How do YOU know so much abouthese!
You are quite correct about the washers.
I still do not understand how their magnetic mass on the middle pole affects the period off the coils.
“Put an ammeter on it and see how much current it draws when in operation”
Better yet a shunt resistor and an O’scope to see the current profile with time.
“I still do not understand how their magnetic mass on the middle pole affects the period off the coils”
I think the flux in the middle pole has to reverse when the armature flips over, so it’s an important place magnetically.
After more thought I think the coercivity and hysteresis of the core material come into play and those properties can work over periods of several seconds.
Thinking of this device as two opposing solenoids, I’ll offer this speculation:
The center, smaller coil is wired (or wound) to to oppose the magnetic field of the outer coils. When the left coil is energized it attracts the bar but has to overcome the magnetic field induced in the bars from the center coil. The strength of the center magnetic field is adjusted by the copper (non-ferrous) washers. As the moveable bar is attracted to the left coil, both the upper and lower bars become magnetized and remain magnetized after the left coil is turned off. In transformers called magnetic hysteresis. When the right coil is energized it must first ‘demagnetize’ both bars before it will start to attract the moveable bar.
How do I know? I’ve been around long enough that I had to design circuits using timer relays and various older technologies.
circuitsmith has it right- the toggling bar is more or less a conduit for magnetic flux between the two coils. The washers add to the inductance of that circuit and so affect the fundamental rate of switching.
RR enthusiasts are some of the most fanatical people you will ever meet. For grins I just did a quick search and looky here- I found a web site that does a decent job explaining how they work! http://matt.zont.org/signals/crossings/xngworks/xngworks.html
I do not understand why the washers have influence.
Are they stacked on a pole that is not coupled athe bottom to the coil cores?
(like an E flipped 90-degrees right)
Or does the armature complete a flux “circuit” from the center pole to the bottom of a coil core?
When the car battery charger clips allowed me to contacthe conductors, I felthe reversEMF jolt. I must place some diodes to shunt it from damaging the cheap 6-volt telephone charger transformer.
Also, maybe I should NOT share the lamp neutral since the computer, modem, speaker amplifier, cordless telephone and radio are plugged into the same power strip. Would NE-2 bulbs suffice?
I believe those copper washers serve the same purpose as the copper shunts in a shaded pole induction motor do. The electric current generated in the copper washer by the collapsing magnetic field tends to prevent the magnetic field from collapsing right away. Also, by generating current in the copper washers, it keeps the winding from developing a super high voltage which causes severe arcing of the points.
Large elecromagnets have a lot of inductance which can make switching them off very problematic. On large scrap metal lifting magnets, usually 240 volt DC and around 80 amps, it takes several seconds for the current to come up to full amps and to switch the magnet off, the induced current has to be shunted to resistors and then briefly reverse connected to the generator to drop the load in a reasonable time. If you just try to open the circuit by disconnecting a wire, you will get an arc that resembles the arc of a Jacobs Ladder in an old Frankenstein movie. I know this from experience.
Thanks, B.L.E. Just seems thathe washers are too far out of the flux circuit to have an effect.
Buthe alternating flasherelay is a clever excellent design whichas worked very well for many decades.
I will remove the top bars and wind copper wire around the middle pole to see if I can further slow the rate. Or musthey be independent rings?
A copper coil would do the same thing if it was short circuited, I suppose you could even short the coil through a rheostat and adjust the time constant.
It is very important to remember that magnetic fields are always a loop. The magnetic flux that comes out of the north pole must somehow find its way back to the south pole. In other words, you have a magnetic circuit and that center piece is the path that the magnetic flux takes to the other pole of the electromagnet.
