GRC from the inside


This page shows the reassembly of a GRC secondary movement, along with notes about adjustment and operation. For general information about GRC systems, see the Standard Electric Time GRC synchronous secondary homepage. To view individual examples of GRC secondary clocks of all kinds, see the GRC examples page.


 
Below are all the disassembled parts of the movement to #2 May 1959 FMT-12 GRC (except the buzzer). Let’s put it back together!

#2 inside: all movement parts
#2 inside: 001#2 inside: 002#2 inside: 003#2 inside: 004#2 inside: 005#2 inside: 006#2 inside: 007#2 inside: 008#2 inside: 009#2 inside: 010#2 inside: 011#2 inside: 012#2 inside: 013#2 inside: 014#2 inside: 015#2 inside: 016#2 inside: 017#2 inside: 018
#2 inside: from top of movement

And there you have it—one complete GRC movement. Here are some general notes about what just happened:
» This movement has an Ingraham Z drive motor and a Synchron 640 reset motor. I believe that all parts in this movement are original to the clock, correct for a May 1959 assembly.
» The pictures are presented step-by-step in order of reassembly. They were not shot in that order, though; so the filenames are not in order (in case you want to download the pictures, you will have to rename them).
» Note that the reset ratchet arm setscrew was not tightened until the picture showing the #31 drill bit used in the alignment procedure.

#2 inside: dial and backplate 01
#2 inside: dial and backplate 02
#2 inside: hands 01#2 inside: hands 02

These are the rest of the parts involved in the clock (except glass and bezel). Note that the back plate is supposed to be shiny on both sides, but this particular one is weather damaged.

correction scheme

The original GR system only had the ability to correct every hour (see bulletin 525A). Then, the GRC “anti-scatter” (Flexchron) 12-hour correction (or “Twelfth Hour Correction”) was introduced. The following is an abbreviated transcription of a chart from the 100-001 manual, detailing the contact closure times on a mechanical GRC master:

contact timing of mechanical GRC master
contact closes opens
01-sec h:mm.00 h:mm.01
05-sec h:mm.03 h:mm.08
35-sec h:mm.26 h:mm.01
01-hour h:59.12 h:60.12
16-min h:12.00 h:28.00
12-hour 6:00.00 4:45.00

Basically, there is an Hourly Correction Relay (“T”) that switches power from the drive motors to the reset motors, and a Twelfth Hour Correction Relay (“K”) that shorts power to the reset motors.
 
The Hourly Correction Relay activates when the following conditions are met: (1) the one-hour microswitch is closed, (2) the 35-second contact is closed, and (3) the Power Failure Relay is unlatched.
 
The Power Failure Relay (“N”) is wired to latch when the following conditions are met: (1) the 12-hour microswitch is closed, (2) the one-hour microswitch is closed, (3) the 35-second contact is closed, and (4) the one-second contact is closed; a condition that only exists during the last second of an hourly correction cycle occurring at other than the fifth hour. The Power Failure Relay will re-latch (if unlatched) each time the above conditions are met, and will unlatch when the 12-hour microswitch opens or upon a power failure.
 
The Twelfth Hour Correction Relay activates when the following conditions are met: (1) 12-hour microswitch is open, and (2) the 16-minute contact is closed; at which time, the Hourly Correction Relay remains deactivated, allowing power to be sent to both drive and reset motors.
 
Therefore, the output of an original mechanical GRC master looks something like this:


12-hour “anti-scatter” cycle (occurs twice per day, every day)
4:59.26 to 5:00.01 – odd hourly correction (strictly unnecessary)
5:12.00 to 5:28.00 – both drive and reset motors run
5:59.26 to 6:00.00 – normal hourly correction
 
hourly correction (occurs a single time only after a power outage)
(h):59.26 to (h+1):00.00 – normal hourly correction
 
program events (occur as punched)
h:mm.03 to h:mm.08 – 5-sec signal


 
For comparison, let’s have a look at a scheme commonly employed by newer electronic master clocks:


hourly correction (occurs every hour)
h:59.30 to h:59.59 – normal hourly correction
 
12-hour correction (occurs twice per day)
5:12.00 to 5:27.59 – both drive and reset motors run
5:59.30 to 5:59.59 – normal hourly correction
 
program events (occur as scheduled)
h:mm.00 to h:mm.05 – 5-sec signal


Note that the only timing that is important to-the-second is when the hourly correction switches the drive motors back on. The second hand on each secondary should be adjusted specifically to agree with whatever exact timing its master uses (each movement will have a slightly different gear-meshing delay at startup). All other timings have an allowable tolerance measured in minutes rather than seconds; keeping in-mind that the second hand advances at twice the normal rate during hourly correction, and an average movement takes about fourteen minutes to advance eleven hours (autumn DST).
 
The hourly correction occurs every hour to make up for the short-term differences between the mains/grid frequency and the independent non-synchronous timebase in the master—an error capable of being as much as ten seconds fast or slow in twelve hours. Obviously, this difference would not exist with a master that has a synchronous timebase, as the entire system will track the grid error together.
 
The LTR8-128 (and apparently older examples of smaller Lathem models also) starts the 12-hour correction three minutes late (from 5:15 to 5:30). In my tests, most movements can actually correct from just after the lockout pin drop-off to the beginning of the 5:00 hour in about 14-1/2 minutes. Those three minutes could possibly buy an extra couple decades of reasonably acceptable operation in any given clock (without adjustment). I’m sure it was an accident on the part of the Lathem engineers—but it happened to be one of those good accidents! (Too bad they “fixed” it in later versions…)

troubleshooting

Below is a simple troubleshooting chart that I have come up with, which may be taken as an introduction to the adjustment section. This chart assumes the following circumstances: (1) the presence of a clock fitted with a GRC secondary movement that has been assembled with respect to the manual (2) in which both motors work and (3) can be connected to a test supply, and (4) the clock malfunctions while connected to an appropriately formatted and time-switched source (see correction scheme section above).

problem possible solutions
During normal running, the second hand -and/or- minute hand fails to move, either at all or intermittently. » Adjust bronze ratchet spring clips.
During hourly correction, the hands fail to ratchet up to the zero position, either at all or intermittently. » Make sure nylon ratchet pawl is not stuck in an upward position (often caused by carrying clock upside-down or horizontally).
» Clean stray oil from nylon ratchet pawl (never allow oil near the nylon pawl—even an undetectable amount can impede its operation).
During correction, the hands fail to ratchet forward as expected, because the ratchet gearing moves back-and-forth as one piece rather than positively responding to the action of the pawl. » Adjust bronze ratchet spring clips and friction springs (bronze washers) if they are too loose.
» Clean oil off of the hand shafts and sleeves (internal, requires full disassembly; only possible if oil has been added to those parts).
After 12-hour correction, the clock becomes one hour fast, intermittently or otherwise. (Clock functioned properly for some time after last assembly, but begins exhibiting problem in service.) » Check for wear of correction lockout mechanism (bracket and pin); reform as needed to bring into tolerance (so that nylon roller is lifted clean away from elliptical cam by 5:10).
Correction functions, but hands end up in the wrong place. (Clock has been rebuilt or is of unknown condition.) » Check adjustment of reset ratchet pawl with #31 drill bit (or against adjustment pillar, in newer movement); pawl arm should only very lightly touch the reference rod when the roller is at the bottom of the cam.
» Check position of hands; reposition as needed.
» Check for dial slippage; tape-down dial to back plate as needed.
Movement functions, but hands drag on dial or crystal or each other. (Clock has been rebuilt or is of unknown condition.) » Check that hands are not bent.
» Check that case back plate and dial plate are not bent.
» Check that seconds shaft is not bent (requires disassembly); replace or rework seconds shaft as needed.
» Check that bushing (hole) in rear plate for seconds shaft is not egged-out, and that the rear tip of seconds shaft is not ground-away; replace or rework rear plate or seconds shaft as needed.
» If movement has been rebuilt from various parts (made during different time periods, especially), ensure that the center hole in front plate truly lines-up with bushing (hole) in rear plate; some plates may be slightly (almost undetectably) incompatible.

((This chart may be updated as necessary.))

adjustment

The GRC secondary movement can be somewhat difficult to adjust, especially if you are as much of a perfectionist as I am. The first step, however, is to read the 100-001 GRC system manual (if you haven’t already). Below, I will offer a few summarized notes about adjusting this movement that may be useful to someone who is actually going to do this. (The following will probably not be interesting to casual viewers. Sorry.)

the bronze ratchet spring clips

gearing close-up

center shaft gearing (from #2)

The bronze ratchet springs are some of the hardest things in these movements to keep adjusted—mainly because they wear more than any other parts. The clips should be more-or-less in the form of the ones shown to the right, and should be flat on the tips. The tips should fit into the grooves with just enough force to positively root themselves. I always make sure the seconds spring clip has just slightly too little force (but still enough to hold itself), as it will wear out most quickly (and make horribly loud clicking noises) with the constant hourly corrections of an electronic master.
 
Note that each click of the spring clip represents one-quarter tick-mark on the dial. Therefore, the correction of the hands has a one-quarter minute/second resolution. If a clock fails to release the hands in a consistent position (give or take one-quarter mark—because the spring clips move smoothly and will stop at a slightly different place each time), it is likely that either the ratchet spring clips or friction springs (bronze spring washers) are not adjusted properly and/or do not have the appropriate amount of tension.
 
Also, when somebody who does not understand what they are doing tries to “fix” a GRC movement, it is common for the ratchet spring clips to be accidentally maladjusted—e.g. causing the second hand (very common) and/or minute hand (and motion works; less common) to stop moving or only move intermittently.

the 12-hour correction assembly

The 12-hour “anti-scatter” correction assembly locks out the hourly correction during the 5:00 hour. As shown in the pictures and the video below, the lockout pin should not contact the bracket enough to interfere with the 5:00 hourly correction. The lockout lever should completely lift the nylon roller away from the elliptical reset cam by 5:10. The lockout lever should fall off the lockout pin sometime between 5:30 and 5:35.

#2 inside: 12-hour correction 001#2 inside: 12-hour correction 002#2 inside: 12-hour correction 003#2 inside: 12-hour correction 004

Note that the lockout lever rubs against the lockout pin twice per day for the entire life of the movement (whether any correction is taking place or not). These parts do eventually wear on the contacting edges. Luckily, there is plenty of room for modification to make up for this wear.
 
For severely-worn movements (that completely fail to perform 12-hour correction), the bend in the connecting bracket can simply be increased, causing the lockout pin to lift the nylon roller farther way from the elliptical cam.
 
For only slightly-worn movements, error can often be fixed by simply loosening the reset motor mounting screws and sliding the motor all the way down and to the right, causing the elliptical cam to be farther away from the nylon roller (don’t forget to tighten the motor mounting and readjust the ratchet pawl arm with the #31 drill bit afterward). I assemble mine with the reset motor in that position as a matter of course, as most of them are worn as such.
 
Additionally, I will admit that I avoid touching the setscrew as though I might catch a plague from it. If the movement worked to start with, it should be close enough. I simply insert a small screwdriver on top of the manual set gear in order to push the teeth apart, and move the hour gear by the set gear one click at a time until things come out right. You have to have a light touch to do things wrong like this without scratching/bending things up, though…

hourly correction: minute adjustment

This video shows a more-or-less proper hourly correction cycle:

 
This is my least favorite adjustment. It irritates me when the hands are positioned more than one-quarter tick-mark away from where they are supposed to be (which is the theoretical maximum accuracy). After reassembling a clock, I usually run at least six hourly corrections to ensure that the hands end up in the right places, before reattaching the glass. Even with a seemingly properly-adjusted movement, here are some things to watch out for:
 
• First, ensure that the reset ratchet pawl arm is adjusted properly. This is the adjustment outlined in the 100-001 manual using the #31 drill bit (for movements made before February 1968). The exact position of this arm will determine exactly where the hands will stop during the hourly correction cycle. If the shaft is marred up by the setscrew such that this is not possible, simply rotate the shaft a bit with pliers to expose a fresh section (this shaft is only held into the pawl arm by friction).
 
• Ensure that the dial is attached firmly. Some can rotate relatively freely. This can be fixed by placing a folded-over piece of tape between the dial and the back plate.
 
• Some movements will position the hands about two clicks (half a minute/second) off if the hands zero out in the wrong order. I always make sure the minutes zero out before the seconds in all tests. Additionally, it sometimes helps (possibly my mind, only) to run the drive motor just long enough for the minute hand to start moving before performing the test correction, to fully simulate a real-world on-time correction.
 
• Some movements (especially older) will drop the minute hand one click (one-quarter minute) slow about half the time, even though everything seems to be adjusted properly. I have accepted this as just a wear-related nuisance, and always run enough test corrections to ensure that the minute hand never indicates a quarter minute fast (as I believe that a quarter minute slow is easier to read than a quarter minute fast, in any event). This may, rather, be due to eccentricities in the bronze ratchet spring clips.
 
• If a clock simply stops correcting, but the motor still works, the nylon ratchet pawl is probably just jammed. If someone has gotten oil on this part, then it needs cleaned out—the nylon piece is very light and relies on gravity to draw it down. Never get oil anywhere near the nylon! Otherwise, these sometimes just stick while moving the clock if it has been turned upside-down—if you have just reinstalled a clock and this happens, simply smack the side a few times.
 
Additionally, if a well-meaning repairer has reassembled the movement with copious amounts of organic oil—especially on the shafts and in the sleeves—then this needs to be cleaned out. (I admit to having done this, long ago.) Gunky sleeves will mess with the correction cycle (actually, if you watch the hourly correction example video really closely, you will notice a sludgy feeling about that movement). I am convinced that only the 1rpm gears (in the back) could benefit from oil, and that everything else should be run clean.

hourly correction: seconds adjustment

Since I use electronic master clocks, I have the luxury of getting really exact (picky) about some adjustments. In the case of seconds, I run a correction cycle with an actual master, then observe a signal event within the first five minutes of the hour. I always try to get the second hand to “touch the dot” when the event starts and stops.

seconds testing

Most of my Ingraham Z movements are adjusted to drop the second hand within a quarter second before zero. It is not odd for looser Synchron movements to have to drop the second hand up to three seconds past zero to remain accurate, with most ending up about one second past. (This is in reference to the Lathem electronic masters that switch the drive motors on at h:59.59.)
 
Of course, each mechanical master would be adjusted slightly differently (and wouldn’t be nearly as exact), so this is almost certainly a higher degree of accuracy than these movements were originally designed to maintain.
 
And, of course, don’t forget these:
• adjust the hour hand properly (especially if you have adjusted the 12-hour correction)
• spin the hands around (using the correction ratchet gears) to ensure the hands do not drag on the dial, on the glass, or on each other (make sure the main movement shafts are not bent)—on 16-inch clocks: the back plate is often slightly warped due to the weight of the movement (may require gentle un-warping) and Lucite crystals are not as deep as glass ones, so this adjustment is all the more important on those units
• don’t forget to clean everything (and Novus plastic polish will do wonders on Lucite crystals)!

Advertisements