An electro-magnetic clock pendulum is a precision pendulum sustained not by a mechanical escapement but by short electrical impulses from a coil acting on an iron or magnet armature mounted on the rod. The Synchronome master clock and the Hipp toggle pendulums used across European telegraph offices both relied on this principle. By replacing the rubbing escapement with a non-contact magnetic kick delivered every 30 seconds, the pendulum stays nearly free, holding rates better than 0.1 second per day in temperature-stable rooms.
Electro-magnetic Clock Pendulum Interactive Calculator
Vary bob mass, swing amplitude, pendulum Q, impulse interval, and coil gap to see the required electromagnetic impulse energy.
Equation Used
The calculator estimates the mechanical energy that each coil pulse must add to replace pendulum losses. Stored energy comes from the bob mass, pendulum length for a 2 second full period, and peak swing angle. The impulse energy is then reduced by pendulum Q and by the number of half-swings between coil firings.
- Pendulum is a 1 second half-period regulator, so full period is 2 s.
- Length is calculated from L = g*(T/(2*pi))^2 with T = 2 s.
- Amplitude is the peak angular swing from vertical.
- n is the number of 1 second half-swings between impulses.
- Relative magnetic pull uses F/F0 = (0.5 mm/gap)^2.
How the Electro-magnetic Clock Pendulum Actually Works
The pendulum swings under gravity, and a coil mounted near the bob delivers a brief current pulse that nudges it just enough to replace the energy lost to air drag and pivot friction. That pulse is the entire job of the electromagnetic system — keep amplitude constant, then get out of the way. Everything else is the pendulum behaving as a free oscillator, which is why these clocks beat the rate stability of mechanical regulators by roughly an order of magnitude.
Two dominant impulse schemes exist. The Hipp toggle, invented by Matthäus Hipp in 1843, uses a tiny vane on the rod that drops onto a contact only when amplitude has decayed below a threshold — typically a 2° to 3° swing. The contact closes the impulse circuit, the coil fires, amplitude recovers, and the toggle lifts clear on the next swing. The pendulum is impulsed only when it actually needs energy, often once every 30 to 60 seconds. The Synchronome free-pendulum scheme, patented by Frank Hope-Jones and George Bennett Bowell in 1895, takes a different route — it impulses on a fixed cadence, every 30 seconds, with a gravity arm released by a synchronised slave mechanism. The pendulum never touches the count train. If timing is off — say the impulse arrives during the upswing instead of at bottom-dead-centre — the pulse adds phase error rather than energy, and the rate drifts. Coil air gap matters too. We hold it to 0.5 mm ±0.05 mm in modern rebuilds. Open it to 1 mm and impulse force drops by roughly a factor of four because magnetic pull falls off with the square of distance, and the toggle starts firing on every swing instead of every other minute. Failures are almost always electrical: pitted contacts on the toggle (silver-tungsten is the fix), weak impulse capacitor, or a coil whose insulation has broken down between layers and shorted half the turns.
Key Components
- Pendulum rod and bob: Invar or fused-quartz rod carrying a 6 to 7 kg cylindrical bob, tuned for a 1 second half-period (2 second full beat). Rod thermal coefficient must stay below 1.2 ppm/°C — Invar 36 hits this naturally. A 1°C room shift on a steel rod would drift the clock 1 second per day, which is the entire error budget of a good regulator.
- Impulse coil: Air-cored or soft-iron-cored solenoid mounted on the case, energised for 20 to 50 ms by a contact closure. Typical resistance 200 to 400 Ω at 24 V DC, delivering an impulse energy on the order of 5 mJ per kick. Air gap to the armature must be 0.5 mm ±0.05 mm — wider gaps need higher voltage and produce a softer, less repeatable kick.
- Armature: Soft iron slug (Hipp design) or permanent magnet (later Riefler-derived designs) attached to the rod near the bob. Mass kept under 20 g so it does not shift the pendulum's centre of oscillation. Permanent-magnet armatures give a cleaner, polarity-defined impulse but require the drive electronics to switch direction in step with the swing.
- Hipp toggle (or release detent): Pivoted vane that drops by gravity onto a contact only when bob amplitude falls below a set threshold. Contact pressure is just 2 to 5 grams — too high and the toggle drags on the rod, too low and contact resistance climbs and the impulse misfires. Silver-tungsten contacts last decades; brass-on-brass pits within a year.
- Impulse drive electronics: Originally a simple battery and capacitor; modern rebuilds use a 24 V regulated supply, a 470 µF reservoir cap, and a MOSFET to switch the coil. Pulse width is set so impulse delivery completes before the bob travels more than 5° of arc — longer pulses smear the impulse across the swing and degrade isochronism.
- Counting train and slave outputs: Mechanical or electrical counter that increments on each beat. Synchronome systems drive slave dials by sending a 1 second polarity-reversing pulse down a parallel telegraph line — one master could drive 50 slaves across a building.
Who Uses the Electro-magnetic Clock Pendulum
Electro-magnetic pendulums dominated precision timekeeping from roughly 1890 to 1940, and they still appear today in horological reference rigs, museum installations, and a small number of metrology backup systems. The reason is simple — a free pendulum impulsed electrically beats the rate stability of any mechanical escapement because the impulse is small, infrequent, and non-rubbing.
- Astronomical observatories: Royal Greenwich Observatory installed a Shortt-Synchronome free pendulum as primary time standard in 1924, holding rate to better than 1 ms per day until quartz took over in the late 1940s.
- Railway and telegraph networks: British, Swiss, and German railway master clocks from the 1880s onward used Hipp toggle pendulums to drive station slave dials over telegraph wires — a single Hipp master in Bern coordinated dozens of platform clocks.
- Metrology laboratories: PTB Braunschweig and the National Physical Laboratory ran Riefler electromagnetic regulators as transfer standards for caesium-comparison work into the 1960s.
- Public building master clock systems: Synchronome and Gent's of Leicester installed thousands of impulse-driven master-and-slave systems in hospitals, schools, and Parliament buildings — many still operate today after a contact refurbishment.
- Horological education and museums: The Science Museum in London, the Deutsches Uhrenmuseum in Furtwangen, and several university horology programmes run working Hipp and Synchronome pendulums as teaching pieces for impulse-timing demonstrations.
- Heritage clock restoration: Specialist firms like Smith of Derby restore turret-clock electromagnetic pendulum drives — typically a 1.5 second pendulum with a 30 second impulse cadence driving a remontoire that winds the going train.
The Formula Behind the Electro-magnetic Clock Pendulum
The practical question with an electromagnetic pendulum is always the same — how much energy must each impulse deliver to maintain a chosen amplitude against the measured Q of the pendulum? At the low end of the typical operating range (high Q, narrow amplitude) the impulse is tiny and infrequent, which is exactly why these clocks are so stable. At the high end (low Q from a draughty room or a worn pivot) the impulse must grow, the toggle fires more often, and isochronism suffers because larger amplitude excursions push the pendulum off its small-angle approximation. The sweet spot sits where impulse energy replaces exactly the cycle loss at a 2° amplitude, with the toggle releasing once every 30 to 60 seconds.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| Eimp | Energy per electromagnetic impulse | J (joules) | ft·lbf |
| Estored | Peak kinetic energy of the pendulum at bottom of swing (½ × m × v2) | J (joules) | ft·lbf |
| Q | Quality factor of the pendulum (energy stored divided by energy lost per radian) | dimensionless | dimensionless |
| n | Number of swings between impulses | dimensionless | dimensionless |
Worked Example: Electro-magnetic Clock Pendulum in an observatory free-pendulum reference
A timekeeping research group in Wellington is rebuilding a 1932 Shortt-Synchronome free pendulum as a demonstration reference for a university metrology course. The pendulum is Invar, 994 mm pendulum length for a 1 second half-period, bob mass 6.5 kg, peak amplitude 1.5°, measured Q of 12,000 in the vacuum tank at 50 mbar. They want to size the impulse coil energy delivered every 30 seconds (n = 15 full swings, since the Synchronome impulses every other 30 second cycle on the master beat).
Given
- L = 0.994 m
- m = 6.5 kg
- θ0 = 1.5 degrees
- Q = 12000 dimensionless
- n = 15 swings between impulses
Solution
Step 1 — compute the peak kinetic energy stored in the pendulum at bottom-of-swing. Convert amplitude to radians and use the small-angle energy expression Estored = m × g × L × (1 − cos θ0):
Step 2 — at nominal Q = 12,000 and n = 15 swings between impulses, compute the required impulse energy:
That is a tiny kick — well under a microjoule — which is exactly why the pendulum stays nearly free. A 250 Ω coil pulsed at 24 V for 25 ms dissipates 60 mJ in the coil itself, but only a fraction of a microjoule actually couples into the bob. The rest is wire heating and field collapse losses.
Step 3 — at the low end of the operating range, with a degraded Q of 4,000 (typical for an air-pressure rig with a worn suspension spring) and n unchanged:
The impulse must triple. At the high end of practical operating range, a vacuum-tank Q of 25,000 (close to the original Shortt achievement):
Below 0.5 µJ the impulse becomes so light that contact bounce on the release detent dominates the energy budget — you cannot reliably deliver less than the contact's own mechanical noise floor. That is the practical Q ceiling for a Hipp-style toggle, which is why Shortt went to a separate slave-released gravity arm to escape it.
Result
The nominal impulse energy required is about 0. 76 µJ delivered every 30 seconds. In practice that is a coil firing for 20 to 30 ms with a peak current of 80 to 120 mA into a 250 Ω solenoid at 24 V, with the bulk of the electrical energy lost as coil heat. At the low-Q end (Q = 4,000) the impulse must rise threefold to 2.27 µJ, and you will hear the toggle firing every other swing instead of every 15th — a clear audible diagnostic. At the high-Q end (Q = 25,000) the impulse drops to 0.36 µJ, which approaches the contact-bounce noise floor. If your measured impulse rate is much higher than predicted, check first for a cracked Invar rod weld (instant Q collapse), then for residual air leakage in the vacuum tank above 100 mbar, then for a misaligned armature creating eddy-current drag in the bob support spider.
When to Use a Electro-magnetic Clock Pendulum and When Not To
Electromagnetic pendulums sit between purely mechanical regulators and quartz/atomic references. The decision is rarely about absolute accuracy alone — it is about maintenance burden, environmental sensitivity, and whether you need the pendulum to drive slave dials or just keep its own time.
| Property | Electro-magnetic pendulum (Synchronome/Hipp) | Mechanical deadbeat regulator | Quartz master clock |
|---|---|---|---|
| Rate stability (typical) | 0.01 to 0.1 s/day | 0.5 to 2 s/day | 0.001 s/day |
| Pendulum Q achievable | 10,000 to 25,000 (vacuum) | 2,000 to 6,000 | N/A |
| Impulse cadence | Every 30 to 60 s | Every swing (1 to 2 s) | Continuous electronic |
| Maintenance interval | Contact dressing every 3 to 5 years | Full strip-and-clean every 5 to 10 years | Battery every 2 years |
| Initial cost (restoration grade) | £8,000 to £25,000 | £3,000 to £12,000 | £200 to £2,000 |
| Slave-dial drive capability | Native, up to 50+ slaves | Requires add-on contact unit | Native digital output |
| Sensitivity to room temperature | Low (Invar, sealed tank) | Moderate (depends on rod) | Very low |
| Typical lifespan with care | 80+ years | 100+ years | 20 to 30 years |
Frequently Asked Questions About Electro-magnetic Clock Pendulum
The toggle is supposed to drop and close the impulse contact only when amplitude has decayed below the set threshold, typically 2°. If it fires constantly, amplitude is not decaying — which sounds good but is a symptom, not a feature. The usual cause is the toggle's release vane sitting too low, so the rod brushes it on every swing regardless of amplitude. Lift the vane mounting by 0.2 to 0.3 mm and the cadence will return.
The other common cause is a weak impulse — the coil is delivering too little energy per kick to lift amplitude above the threshold, so the toggle drops again immediately. Measure peak coil current with a scope across a 1 Ω shunt; if it is below half the design value, the reservoir capacitor has dried out or the coil has a partial inter-turn short.
You can, and modern rebuilds often do, but it changes the drive requirements. A soft-iron armature is pulled by either polarity of coil current, so a simple unidirectional pulse works. A permanent magnet is polarity-sensitive — drive it the wrong way on a given swing and you brake the pendulum instead of impulsing it.
The fix is an H-bridge driver synchronised to the swing direction, usually triggered by an optical sensor on the rod. The payoff is a cleaner, more repeatable impulse because the force-versus-position curve of a magnet-in-coil is more linear than iron-in-coil, which improves isochronism at the µs level. Worth doing for a metrology rig, overkill for a master clock driving slave dials.
This is the classic isochronism trap. A stronger impulse raises amplitude, and a real pendulum is not perfectly isochronous — period increases slightly with amplitude per the circular-error term (about 0.7 ms/day per 0.1° amplitude increase at 1.5° nominal). Counter-intuitively, the clock should run slow with more amplitude, but if the impulse is delivered off bottom-dead-centre it adds a phase advance that overwhelms the circular-error correction.
Check impulse timing with a phototransistor at the rod's mid-swing position. The impulse should peak within ±10 ms of bottom-dead-centre. Late impulse delivery is the single most common rate-shift cause after a coil swap.
The distinction is whether the pendulum interacts with the count train. In a Hipp design, the same pendulum that keeps time also closes the impulse contact and drives the counter — it is doing three jobs. In a Synchronome free pendulum, a separate slave clock handles counting and slave-dial drive, and the master pendulum's only job is to keep the slave synchronised via a gravity-arm release every 30 seconds.
Pick Hipp if you want simplicity, a single mechanism, and 0.1 s/day stability is enough — railway master clocks, public buildings, teaching pieces. Pick Synchronome if you want sub-10 ms/day, are willing to build and align a slave, and ideally can run the master in a sealed low-pressure tank. Below about 5,000 pendulum Q the Synchronome architecture buys you almost nothing over a well-tuned Hipp, so do not over-engineer.
Halving Q means doubling the energy loss per cycle, and there are only a few places it can go. Air drag on a 6.5 kg bob at 1.5° amplitude in atmospheric pressure accounts for a Q of roughly 8,000 to 10,000 — if you are not running a partial vacuum, you will never see 12,000. Original Shortt clocks ran at 30 mbar precisely for this reason.
If you are already in vacuum, suspect the suspension spring next. A spring with a hairline crack or work-hardened flexure zone dissipates energy as internal friction and can drop Q by 30% with no visible damage. Replace the spring with a fresh phosphor-bronze strip ground to original thickness (typically 0.10 to 0.13 mm) and re-measure. Bob support spider eddy-current losses come third — if the spider is steel rather than non-magnetic brass, the impulse coil's stray field induces drag currents in it.
Yes, badly, if you do not snub it. Switching off a 250 Ω, 2 H coil at 100 mA generates a back-EMF of several hundred volts across the toggle contact, and that arc pits the silver-tungsten surfaces over thousands of operations. Pitting raises contact resistance, which weakens the next impulse, which lowers amplitude, which fires the toggle more often — a self-accelerating failure.
Fit a flyback diode (1N4007 is fine for these currents) across the coil, or a 100 nF / 100 Ω RC snubber across the contact. Original Synchronome installations used a small spark-quench capacitor for exactly this reason. After fitting, contact life extends from roughly 5 years to 30+ years on the same set of points.
References & Further Reading
- Wikipedia contributors. Electric clock. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
