A quartz watch that is two seconds fast after a fortnight is not broken. It is doing what a quartz watch does. The question people bring to the counter in Durrës is almost never whether the watch is perfect, it is whether they should be worried, and the answer depends entirely on which of four very different published numbers they happen to have read.
In short: the crystal is a component, measured under steady conditions. The watch is that crystal inside a case, on a warm wrist by day and a cold shelf at night. A few seconds a month is normal. A few seconds a day is not, and on a fresh battery it points at the movement rather than the cell.
Four Numbers That Only Look Like They Disagree
Every one of these is correct, and each measures something different. Read them as a ladder rather than as competing claims.
| What is measured | Figure | What it actually describes |
|---|---|---|
| The bare 32,768 Hz crystal | a few seconds per year | The oscillator alone, under normal steady conditions. The laboratory ideal, as set out in our history of the Quartz Crisis. |
| A finished watch in normal wear | a few seconds per month | That same crystal in a real case at a real temperature. This is life rather than the laboratory, and it is what a quartz watch does whatever it cost. |
| The published Japanese quartz specification | ±15 seconds per month | The guaranteed outer limit a manufacturer stands behind, not the result you should expect. Most watches sit well inside it. |
| A watch losing time on a fresh battery | 15 seconds per day | Roughly thirty times that outer limit. A different order of problem entirely, and a reason to bring the watch in. |
The first two are the pair that confuses people, and the reconciliation is simple once it is said out loud. A crystal is cut to a tolerance and trimmed, and in that state, held at one temperature, it really does hold a few seconds a year. Then it is soldered into a movement, sealed into a case, strapped to an arm at thirty-something degrees and taken off onto a bedside table at fifteen. The specification did not get worse. The conditions did.
What Changes Between the Bench and Your Wrist
Four things, in roughly the order they matter:
- Temperature. The largest single factor by a wide margin, and the subject of the next section.
- Battery voltage. A cell delivers a slightly different voltage at the end of its life than at the start, and the circuit dividing the crystal signal is not perfectly indifferent to that.
- Ageing of the crystal. Quartz drifts a little in its first year and then settles. The change is small, and it runs one way rather than wandering.
- How the watch is used. A watch worn twelve hours a day and left in a drawer for the other twelve lives at two temperatures. One worn continuously lives at one.
Notice what is not on that list. Position and shock, which dominate the rate of a mechanical watch, barely register here. That gap is the whole reason quartz took over, and we went through it in mechanical versus quartz, where a well-regulated mechanical is doing well to hold ±5 to 30 seconds a day against the quartz figure of ±15 seconds a month.
Temperature Is the Biggest Single Factor
The quartz in a watch is cut as a tiny tuning fork, and its frequency is not flat across temperature. It peaks somewhere near normal wearing temperature and falls away on either side of that peak. The important part is that it falls away on both sides: a cold room and a hot car do the same thing to the rate, and both make the watch run slow rather than fast.
That single fact explains most of what gets reported to us.
Why the same watch gives two different answers. Measure it over ten days in February, when it spends every night on a cold shelf, and it loses. Measure it over ten days of an August in Durrës, sitting in a car through the afternoon, and it loses again for the opposite reason. Wear it continuously against your skin and it sits closest to its best temperature for the longest, which is why the watch that never comes off is usually the most accurate watch its owner has.
None of that is a defect, and none of it is worth a service visit. It is a watch behaving exactly as the physics says it will.
A Dying Battery Is the Most Common Cause of a Sudden Change
If the rate has changed rather than merely being imperfect, the battery is the first thing to check and it is right most of the time. A weakening cell produces inconsistent voltage, the movement runs slightly slow on it, and a watch that was accurate before starts losing 5 to 30 seconds a day. Our battery guide covers the full set of symptoms, including the second hand that begins jumping in four second steps, which is the same warning in a more obvious form.
The word doing the work here is suddenly. Steady drift of a few seconds a month has always been there and you have only just started counting it. A watch that has changed its behaviour is telling you something, and that is the one worth bringing in.
When Drift Is a Fault Rather Than Normal
The threshold. If your quartz watch loses more than 15 seconds a day on a fresh battery, or stops and restarts for no obvious reason, it is almost always the movement rather than the cell. That is roughly thirty times the outer limit the movement was built to hold, which is why it reads as a fault rather than as bad luck.
Past that point the usual cause is the movement itself: degraded lubricant in the gear train creating friction the motor has to fight, or a problem in the oscillator or regulator circuit that no battery will fix. We set that out alongside the other warnings in five signs your watch needs a service.
Quartz movements are not maintenance free, whatever the advertising said in 1975. They carry lubricated wheels and keyless work like anything else, and those lubricants age on the same timeline. A full quartz service strips the movement, cleans it ultrasonically, re-oils it, fits a fresh cell matched to the original reference and then runs it on test for at least 24 hours before it goes back. That last step is what separates a diagnosis from a guess.
How to Measure It Without Fooling Yourself
A good share of the drift reported at the counter turns out to be measurement error. The method that works takes ten seconds of effort and then patience:
- Pick one reference and keep it. A phone or computer clock that syncs over the network. Not another wristwatch, and not the oven.
- Set the watch to the second, then write down the date and the exact time you set it.
- Leave it alone for ten to fourteen days. Overnight tells you nothing at all.
- Compare against the same reference and divide the difference by the number of days.
The reason for the wait is arithmetic. Misread the hands by one second across a two day test and you have invented fifteen seconds a month of drift that is not there. Across a fortnight the same one second error is noise.
What Paying More Does, and What It Does Not
This part is worth being blunt about, because it is easy to sell against. Every watch on our shelves is quartz, and every one of them is accurate to within a few seconds a month. Paying more does not buy better timekeeping.
The Daniel Klein Premium at 64 euro (6,200 L) and the Hislon Classic at 199 euro (19,300 L) keep time the same way, because they are doing the same thing with the same physics. What the Hislon money buys is a sapphire crystal, a slimmer dress profile and the finishing that goes with it. The whole ladder is laid out in what more money actually buys in a watch. If someone tells you a dearer quartz watch is more accurate, they are selling rather than explaining.
If It Has Changed, Bring It In
The workshop is on Rruga Aleksander Goga in Durrës, open Monday to Saturday from 8:30 to 20:30, no appointment. Tell us what changed and roughly when, because "it started losing a minute a week in June" narrows the diagnosis faster than any bench reading. Most of the time it is a cell and you are out in a few minutes. When it is not, we say so before we open anything.
Frequently Asked Questions
Is it normal for a new quartz watch to gain or lose a few seconds a month?
Yes. A few seconds a month is what a finished quartz watch does in normal wear, and it is what a quartz watch we sell does regardless of what it costs. Japanese quartz movements are published at plus or minus 15 seconds a month, which is the guaranteed outer limit rather than the expected result, so a watch drifting two or three seconds a month is comfortably inside its own specification.
Why do specifications say the crystal is accurate to a few seconds a year?
Because that figure describes the bare 32,768 Hz crystal under steady conditions, not a watch on a wrist. Once the crystal sits inside a case that moves from a warm arm to a cold room to a hot car, and is driven by a battery whose voltage falls over several years, the year figure becomes a month figure. Nothing has gone wrong in that step. It is the difference between a component measured in isolation and the same component doing a job.
My watch was accurate and has suddenly started losing time. What changed?
Check the battery first. A weakening cell delivers inconsistent voltage and the movement runs slow on it, typically losing 5 to 30 seconds a day, and a second hand that jumps in four second steps is the same warning in a different form. The word that matters in the question is suddenly. Steady drift of a few seconds a month is normal, and a change in behaviour is the thing worth acting on.
How much drift means a repair rather than a battery?
If your quartz watch loses more than 15 seconds a day on a fresh battery, or stops and restarts for no obvious reason, it is almost always the movement that needs attention rather than the cell. That is roughly thirty times the outer limit the movement was built to, so it is a different order of problem rather than a worse version of normal drift. Bring it to the counter on Rruga Aleksander Goga in Durrës and we will test it.
How should I measure my own watch properly?
Set it against a clock that syncs over the network, write down the date and the time to the second, then leave it alone for at least ten to fourteen days before you compare again. Anything shorter measures your own reading error more than it measures the watch: one second misread over two days looks like fifteen seconds a month of drift that is not there.
A Watch That Keeps Its Own Time
The Navimarine Steel 5ATM Blue is 75 euro (7,300 L): stainless steel case, deep blue sunburst dial, date at 3, 5 ATM water resistance, brand new with a 1-year guarantee. Free delivery anywhere in Albania, paid to the courier at the door.
Rruga Aleksander Goga · Durrës 2001 · Albania · +355 67 571 6090