SET AR impulse sec.

The Standard Electric Time Company introduced their Automatic Reset (AR) system in 1932, which is a 24v parallel minute-impulse clock system that includes separate reset coils that cause, through gravity and magnetism, the secondary movements to snap to the nearest 59th minute when a 24v signal is applied to the third wire.
In late-1948, SET introduced the AR2 system, which accomplished the same type of correction using secondary movements that contain only one coil supplied with a 48v reset signal over the same pair of wires as the 24v minute pulses. The single-voltage three-wire original AR type was thereafter labeled AR3 to differentiate it from the new dual-voltage two-wire AR2 type. (I tend to use “AR” rather as an umbrella term to describe all types as a group, and shall do so within this site.) The type AR2 secondaries also reset to the 59th minute.
The later AR2A type is equivalent to AR2, but resets to the 60th minute. (I would be interested to know the introduction date, and the end-of-production dates of the other types.)
All three types are functionally equivalent and may be driven by the same master (with the right power supply). A simple gentle twist is all that is necessary to change the minute to which an AR secondary movement resets.

identification and operation

AR3 secondary movementAR2 secondary movement

Above is a comparison of an AR3 (left) and an AR2 (right) secondary movement of the no.1 size, both made in 1956.
An original AR3 master maintains the following output timings:
h:mm.00 to h:mm.01, minute impulse (sixty times per hour);
h:59.13 to h:59.23, reset pulse (every hour);
h:mm.03 to h:mm.08, signal duration.
And an AR2 master:
h:mm.00 to h:mm.01, minute impulse;
h:58.50 to h:59.02, reset pulse (replaces 59th minute pulse entirely);
h:mm.05 to h:mm.10, signal duration.
((When I find copies of the actual manuals, they will be posted here.))
Note that, given the right power supply and output relay wiring, an unmodified example of either secondary type (or both types simultaneously) can be driven by either type of master.


#14: FMT-12

#14 deco x deco
#14 front#14 back#14 movement

inspection date: June 1952
type: AR2
I just had to have one of the art-deco units (specifically to match that speaker). I have a feeling this one will be mounted in my library-room, if I ever have one of those.
Honestly, I am not really a fan of the art-deco dial (hence why I would hide it somewhere that I rarely spend much time!). That said, it seems to have a slightly higher visibility (and, therefore, usability) in a given lighting and distance than the newer international-modern style. Then the Johnson Tempo is even worse; and we all know about the industry-standard nowadays (see: Simplex, et al). It is odd how the usability of the designs of institutional clocks has been notably decreasing over time, almost like a part of some type of conspiracy…
(This clock gave its bezel clips and speed-nuts to #2; it is supposed to have those. All other parts are original. Note also that this clock is the only one assigned a number out of acquisition order; I think it actually should be just before #8—I have definitely had it for almost as long as my first GRCs.)

#16: FMT-12

#16 front
#16 back#16 movement#16 movement#16 movement#16 movement#16 movement#16 movementFMT backbox

inspection date: April 11, 1956
type: AR2
I bought this clock specifically for the backbox that was included. This set came from a school building somewhere in Ohio that no longer exists, supposedly; and was delivered with(out) three missing screws (one for the hanging brackets and the two on the connection block). The back edge is dripped with paint of that one certain shade of pleasant minty-green that also appears on school clocks from entirely different geographic locations, somehow or other.
This clock has the original rough-printed modern dial. (Shown with all original parts—except the terminal screws. The coiling on the light-colored wire is probably not original in any picture; the wire coiling on other example clocks is rather more likely original.)

#12: FMT-16

#12 front
#12 back#12 movement#12 movement#12 movement#12 movement#12 movement

inspection date: August 16, 1956
type: AR3
crystal: Lucite (acrylic)
Now reunited with all of its original parts (as shown), this clock spent eight years and nine months (2005-04-03 to 2013-12-25)—over one-sixth of its life so far—as a homemade GRC (the slight shiny shadow that should go away, and the permanent scuffs around the outline of the dustcover, can still be seen in the pictures). See #12a/b for the previous GRC version.
This clock includes the 16-inch version of the original rough-print modern dial. Note that, traditionally, a 15-inch dial was the largest no.1 size, with 18-inch being the smallest dial diameter to use a no.2 movement. I have never seen a 16-inch dial with any older version of dial or hands—I suspect the modern “necktie” hands are lighter than any older style, specifically allowing this dial size, making this clock an example of the first ever style of 16-inch SET secondary dial (this is original research/speculation on my part and should not be quoted; corrections requested).


((Something similar to a “from the inside” section is planned for this page.))

for now, notes about adjustment: changing the reset minute

Take your AR secondary clock in-hand, fully assembled with dustcover removed, in freshly-reset position. There is a pin that goes through the minute shaft, maintaining pressure on the spring-washer-like piece that holds the main gearing together; see that pin? Good. Now place the tips of a pair of needle-nose pliers parallel to the pin, lightly gripping the shaft. While holding the center ratchet gears firmly with your fingers, gently twist the shaft. (This is easier with AR2; may be better to set to a different minute before doing this on an AR3 movement.) Only the minute hand (the center shaft) and hour hand (motion works) will move—if anything else moves, you’re doing it wrong.
These movements were designed to slip in this way for some reason or other, possibly a holdover from the pre-reset days. It should be loose enough that you can move the hands (shaft) without messing up the alignment of the ratchet gears or moving the reset weight or bending the pin. That said, pay close attention to the alignment of the two main ratchet gears—they are not attached to each other. If the movement now misses or skips or otherwise doesn’t work, then you probably moved one of these gears ever-so-slightly. This is a good time to make sure the ratchet pawls still fall into the gear teeth properly.
Advance manually a few times and fine-adjust as necessary to ensure the minute hand lines up with the chapter-marks. Finally, check to make sure you adjusted it to reset to the right minute. (Then start over, if necessary!)
To check the reset action without applying power: look at the movement squarely from the rear and find the small vertically-elongated hole directly above the bottom-center of the back plate. You should see a small metal tip pointing upward, made of the same material as the hexagonal reset weight. The reset signal will cause the bottom tip of the weight to line up with this fixed tip, as viewed squarely through the hole. There will only be one minute that places both tips in view through the hole—the minute to which the weight will fall-and-snap in response to the reset signal, that is. If the tips don’t line up exactly when the movement is at-rest on the reset minute, then someone has messed with the position of the reset weight or otherwise assembled the movement incorrectly.


The AR movements are way more simple than the GR(C) movements, and the operation is totally less finicky/tinkery/messy. And they seem to be more durable. If I had it to do over, I probably would mainly have an AR system right now, with a couple GRCs on the shelf just for fun. At the very least, I could have saved countless hours spent adjusting movements—whereas a GRC can take hours to get the minute hand to line up properly (or, that long for me to settle on exactly how far off it might decide to be sometimes and whether or not I can live with it), an AR takes less than thirty seconds to line up. And it’s always exactly on the mark!