Beat error, the millisecond difference between the balance wheel's tick and tock, is a common but often misunderstood metric in watchmaking; according to Antoine Simon, beat error under 1ms has no measurable effect on rate or amplitude, so it should only be addressed when genuinely high (2ms+) and only after ensuring the movement's overall health, amplitude, and rate are properly optimized, as high beat error is typically a symptom of underlying mechanical issues like loose jewels or low amplitude rather than the problem itself.
Deep Dive
Prerequisite Knowledge
- No data available.
Where to go next
- No data available.
Deep Dive
When To Worry About Beat Error & When Not To
Added:You put a movement on the time graffer and three numbers jump out at you.
Rate, amplitude, and beat error.
Now, beat error is the one almost every new watchmaker goes after first because it's sitting right there at the top of the screen and it's usually pretty easy to adjust.
But, it's the number that matters the least.
Now, a healthy watch, a beat error that's a little high doesn't hurt anything. And chasing it towards zero is one of the best ways I know to waste a bunch of time.
My name is Alex and this is watch repair tutorials.
Now, for those of you that don't know, beat error is the difference in milliseconds between the tick and the tock. Now, those are the two rotations of the balance wheel which are timed in milliseconds. And when the escapement is timed right, the impulse jewel comes to rest on the line of center when the movement is at rest.
Now, the line of center is just an imaginary line that runs from the balance wheel pivot through the pallet fork pivot. Now, when that rest position is off a little, one side of the balance wheel's rotation takes slightly longer than the other and the time graffer shows that difference as a beat error. So, the number is really just telling you one thing. If the balance is resting where it should relative to the fork.
Now, here's the part nobody tells beginners.
It doesn't have to be zero and it doesn't even have to be close to zero.
Now, the late Antoine Simon, and the former director of WOSTEP, actually put a number on it. He said, "A beat error under 1 millisecond has no measurable effect on the rate or the amplitude of the watch. None."
>> [sighs] >> So, if you've got a watch sitting at half a millisecond or even 3/4, and you're in there twisting the stud carrier trying to get it to read zero, you're chasing a number that actually changes nothing about how the watch runs on your wrist. So, if it doesn't affect the rate, you might wonder why beat error is a spec or specification at all.
>> [sighs and gasps] >> Well, there's really only one practical thing a slightly high beat error costs you, and here's what it is.
Take a watch that's fully run down.
There's no power in the mainspring at all.
Now, you start winding it.
A watch with near-perfect beat error will usually kick right off on its own and start ticking as that power starts coming into the escapement.
Now, a watch with a higher beat error might just sit there for a second until there's enough torque to get it to start moving.
Basically, it just comes down to the balance doesn't self-start as easily when the beat error is slightly elevated. And for some, that might sound like a problem, but in reality, all you do is rock the movement back and forth a couple times, and it just starts right up. That's the whole downside to a slightly elevated beat error.
It's not a running fault in the movement. It's not hurting your timekeeping.
It just might need a little nudge to get it going when it's starting from a dead stop. So, if you have a low beat error, just leave it. A little high? Leave it.
Just start it with a little shake.
Now, beat error should really only get your attention in one situation. And that's when it's high. And I don't mean 1 ms high. I mean 2, 3, 4, and up.
But even then, the answer is not to go chase the beat error number first.
Think about what a big beat error is telling you.
The balance is resting well off center from the fork.
And on most watches, that's going to come across your bench.
When you see a number like that, that's usually not the problem.
You'll usually find the amplitude's low, or the rate's wandering all over the place. The traces on the timegrapher screen look like a snowstorm.
Something mechanical is wrong. A high beat error is really just the symptom, not the disease.
So, when you see that high number, go find the real problem. Low amplitude usually means you got dried oil, you got a dirty jewel, you got a worn pivot, you got a hairspring rubbing, you got something that's eating up the torque coming from the mainspring.
Fix that first.
Clean it, service it, get the amplitude back up where it belongs for that particular caliber.
A lot of the time, once you've sorted the actual fault and the watch is running strong again, that beat error settles down on its own, or it drops into that range where it really doesn't matter anyway.
Trying to adjust the beat error on a watch with low amplitude or with a lousy rate is kind of pointless.
You're doing fine-tuning on a movement that isn't even running right yet.
Get it healthy first, then if there's anything left worth doing, deal with it then. Now, there are some cases when a high beat error really is the problem. Now, you know, there's a bunch of mechanical faults that can affect the timing of the balance wheel's oscillation.
And a couple that just come to mind are loose jewels. It's not uncommon to see a loose impulse jewel in a vintage watch where the shellac has loosened up or it's loosened up during your cleaning process.
And what happens now is as the impulse jewel goes back and forth through the pallet fork, it's actually moving a little bit and that's changing the timing of each swing of the balance wheel. Therefore, you get a beat error.
And usually a pretty big one. You know, a loose pallet stone can also have a very similar effect as well.
So, those are the kind of things you want to check first.
Another is when the amplitude is really high, right?
The amplitude reading on your time grapher is an average of both rotations of the balance wheel, right?
And when you have a big or high beat error, one side swings farther than the other.
So, if that side swings far enough, the impulse pin can slam into the outside of the pallet fork, and that's what we call overbanking.
And once it starts, you'll see the rate jump around like crazy, and you'll probably think you have some sort of power transmission problem, or you got something rubbing in the balance, you know, which when the real cause is that high beat error, letting one side overswing.
So, on a movement that's already running at a high amplitude, a high beat error is worth taking care of. Not because the number itself matters, but because of where that extra swing can actually end up, maybe causing some damage. Now, one of the other things you got to consider is you got to be realistic with your comfort level when you're adjusting beat error.
Correcting beat error means adjusting the balance wheel's position.
Now, you know, if you're working on a modern movement, you just nudge the stud carrier a little bit, which is pretty straightforward.
But, as you might know, on vintage watches, the stud is fixed, and you to fix beat error, you have to rotate the collet on the balance staff.
Now, if you do it wrong, you can actually do more harm than good. I just talked to someone the other day who had a watch running with a beautiful amplitude.
And he went in to fix the beat error, and he tanked the whole thing because he bent the hairspring, and now it wasn't on center, and it was rubbing basically on the bridge.
So, he traded a healthy watch for a broken one to fix a number that didn't really matter.
So, when you're new, here's how I would look at it. First, you got to keep in mind you're not a professional watchmaker, and you're not working under standards set by some watch brand or movement manufacturer.
And you're probably not working on a chronometer, which, you know, they operate at much higher standards.
So, the first thing is worry about the overall movement's health.
Things like good amplitude, you know, clean trace lines, uh train that's running free, and a balance that's running free with nothing rubbing, right? No hairspring rubbing, that's a big problem. Second, look at the rate.
Get it running as well as you can for the movements you're working on.
And the third, which is way down at the bottom of priorities, is the beat error.
And then, only worry about it if it's genuine genuinely high, and only after everything else has been handled. If you were to flip that order around and start with beat error, I mean, you could burn a whole evening just chasing a number that was never the problem in the first place.
And sometimes, you'll make the watch worse when you're doing it. So, next time you drop a movement on the timegrapher, and that beat error number catches your eye, take a breath.
Ask if it's actually high.
Ask if the watch is even running well enough to worry about it. Most of the time the answer is leave it alone and go fix something else that really matters.
And because there's so much more to learn, I hope to see you in the next video.
Related Videos

Audi RS5 4.2 Tuning JDEngineering
JDEngineering
1K views•2013-11-14

DALI + KNX: 500 Lights Offline! BCU Code Lock & Short Address Fix!
EngineerIsmailTech
560 views•2026-04-13

Explaining Quality Control of Concrete
maherbader
4K views•2019-05-25

World Mining Production Peaks - Lead Antimony Arsenic Titanium & more
LucarioandDialga
1K views•2019-04-19

Tech Titans: LFP vs Sodium-Ion Battery | Which is the most effective energy storage solution?
Enfsolar
522 views•2025-11-06

Doing the Math: Analysis of Forces in a Truss Bridge
TeachEngineering
1K views•2025-06-06

Inside Midnight Fighter Jet Refuelling Secrets of Stealth Missions | WION Podcast
WION
3K views•2025-09-20

Flash Point, Fire Point & Auto Ignition Temperature
HSELessons
38K views•2019-08-28
Trending

Dokibird Being ABSOLUTELY Unhinged
Dokibird
44K views•2026-07-25

Agrees To Prove God… Then Starts Threatening Me
JonCohenOfficial
17K views•2026-07-25

One Must Imagine Sisyphus Important
vlogbrothers
55K views•2026-07-25

Egos ruining Open Source projects, distros vs software conflicts - Linux Weekly News
TheLinuxEXP
25K views•2026-07-25