The Tudor 390: The Vintage Movement Watchmakers Turn Away

Here's a quote from a watchmaker's bench thread about the Tudor calibre 390:

"I admire your courage for attempting a 390... That was the first, and hopefully the last."

And another, from a different source:

"The Tudor 390 is widely known to be a problematic movement to service, even for trained watchmakers."

If you own an early Tudor Submariner — a 7922, 7924 or 7928 — this is what's inside it, and this post is about what that actually means for you.

---

First, a myth to clear up

The Tudor 390 is not an in-house movement. This gets stated confidently and it isn't true.

The Ranfft movement database indexes it literally as "Tudor 390 (FEF 380)." Tudor's own technical catalogue is titled "caliber 390 (FEF 350, 380)." NOS balance staffs are sold as fitting "Tudor 390, FEF 350, 380, 389."

It's a Fleurier (FEF) hand-wound base with a Tudor-designed automatic winding module bolted on top.

(One dealer listing claims it's "based on Valjoux 722." That's wrong — ignore it. And we couldn't definitively resolve whether the base is the FEF 350 or the 380; Tudor's own catalogue title cites both, so it probably draws on the family generally.)

That construction — a bespoke automatic module on a base that was never designed for one — is the source of everything below.

---

The problem

The automatic winding module

The core issue is stated plainly on the bench: the assembly "lacks integration between the base movement and automatic winding system."

Reassembly requires simultaneous precise alignment of the rotor, axle pin, cross clip and base-calibre components. It is not a movement that tolerates being put back together approximately.

The consequence is unusual and worth understanding: this is as much a serviceability problem as a wear problem. The module is intolerant of imperfect lubrication and imperfect reassembly. So a 390 that gets a mediocre service fails again shortly afterwards. Owners then conclude the movement is unreliable, when what's actually unreliable is the outcome of servicing it.

One experienced technician's qualification is the fairest summary: the auto-winder "can be problematic if not serviced (lubricated) correctly."

What you'd notice:

• Poor or no self-winding, despite the movement running perfectly well when hand-wound
• Watch stops overnight or after a rest day
• The rotor turns freely but power reserve never builds
• Noise or roughness from the rotor
• Rotors frequently found gouged from previous clumsy servicing — a sign of the problem's history as much as its present

The mainspring bridle — an under-diagnosed second fault

This one is subtler, and it's probably misdiagnosed more often than it's caught.

The 390 uses a Tudor-specific mainspring bridle geometry. Tudor listed a dedicated part for it — "A/1396 MAINSPRING BRIDLE ONLY," specific to calibre 390.

Correct 390 mainsprings are scarce. So watchmakers substitute a generic automatic mainspring. But a generic bridle is too short for the 390 barrel, and as one bench report describes it: "The rest of the spring is getting caught up under the bridle and causing issues."

The bridle can no longer slip properly. On a slipping-bridle automatic, the bridle is the safety valve — it's what stops the rotor from over-tensioning the mainspring. Disable it and the barrel just keeps stuffing energy in every time the rotor turns.

What you'd notice:

• Audible "galloping" — an irregular, hurried sound
• The watch gains time dramatically when you're moving, and behaves normally at rest
• On a timegrapher: amplitude spikes of 60°+ when worn or on a winder, against a resting amplitude of around 270°

That second symptom is the diagnostic one, and you don't need any equipment to spot it.

Because generic substitution is common, expect this to be a widely under-diagnosed cause of "my 390 runs fast when I wear it." If your watchmaker's answer to a galloping 390 is to regulate it slower, the underlying fault is still there.

---

How to self-test

The winding test

Let the watch run down completely. Don't touch the crown. Hold it flat in your palm and swirl it in a horizontal circle, like swirling a wine glass, about two rotations per second.

A healthy automatic starts within a handful of swirls. Count 50 — it should be running. Continue to 200 total (about two minutes), then set it dial-up and see how long it runs.

• Won't start after 100+ swirls → the autowind isn't coupling at all
• Starts but runs only minutes → the module is slipping badly
• Runs, but the rotor spins freely with no sense of load → the coupling to the barrel is broken

Cross-check by hand-winding: if the watch keeps excellent time when hand-wound but dies whenever you rely on the rotor, the going train is fine and the module is the problem. Which, on a 390, is the expected answer.

The bridle test — worn versus at rest

This needs no tools at all.

1. Sync to time.is and log the offset precisely.
2. Campaign A: wear the watch normally for three days, logging the cumulative offset each day.
3. Then let it run down, hand-wind it fully, and:
4. Campaign B: leave it dial-up on a table, undisturbed, for three days, logging the cumulative offset each day. Top it up to full each morning.

Compare the two rates.

• Similar rates → whatever the watch's timekeeping is like, the bridle isn't the cause
• Dramatically faster when worn, normal at rest → that's the bridle fault, and it's fixable with the right mainspring

If your watchmaker is about to fit a generic mainspring, this test is the argument for not doing that.

General health

The positional variation test is the most informative thing an owner can do without equipment. Fully wind, then leave the watch 24 hours in each of six positions (dial up, dial down, crown up, crown down, crown left, crown right), rewinding each morning, and log the rate for each.

The delta — the spread between fastest and slowest position — is what matters. On a 60-year-old movement, vintage tolerances are much looser than modern ones, so don't panic at a figure that would be alarming on a BB58. But a delta that's grown compared to your own earlier measurement is meaningful, and a delta above about 20 s/day suggests worn pivots or a hairspring problem.

---

The solution

Option 1: Find the right watchmaker

Some will decline. That's documented — one shop turned away a 1960s Tudor Sub with "this watch is too old and he doesn't want to touch it," and "60 year old watch here and sometimes there just isn't enough parts to make these things work anymore."

The same watch was then serviced successfully by a specialist for $1,200, including crown and stem, crystal, cleaning and a new second hand.

So the job is findable. It is just not findable at the first shop you walk into.

Ask directly: "Have you serviced a Tudor 390 before?" Given that this movement punishes imperfect reassembly, the answer matters more than the price — and a straight "no, but I'd like to try" from an honest shop is more useful than a confident yes from one that is about to learn on your watch.

We don't publish a list of names: a shop that was right three years ago may have changed hands or stopped taking vintage work. Snowflake dial rot and the service trap sets out the five questions worth asking any vintage Tudor watchmaker before you hand the watch over.

Option 2: Convert it to manual wind

This is the documented escape hatch, and it's less drastic than it sounds: fit a complete FEF 380 calibre onto the 390 mainplate, converting the watch to hand-wound.

You lose the automatic function. You gain a movement that is straightforward to service, uses available parts, and won't fail again in two years. For a watch that's going to be worn occasionally rather than daily, it's a serious option.

Keep the original 390 parts. Reversibility is worth a lot on a vintage watch.

Option 3: Donor movements

Complete 390 movements trade openly on eBay and through specialist dealers as "for parts." Given that "parts have just about dried up" through official channels — a long-term owner's words on the Vintage Rolex Forum — a donor is often the only realistic source for module-specific components.

NOS balance staffs are cross-listed for "Tudor 390, FEF 350, 380, 389," so some going-train parts are more available than you'd expect.

---

A word for the defence

Not everyone agrees with the pessimistic reading. Revolution's history of Tudor describes the 390 as "one of the most robust movements one could imagine; bulletproof almost!", citing unserviced examples still keeping excellent time.

That's an editorial view rather than a bench source, and it directly contradicts three independent watchmakers. But it's not baseless either, and the reconciliation is straightforward:

The going train is durable. The automatic module is the failure point.

Which is exactly why the manual-wind conversion works so well as a fix — you're keeping the good half and removing the problematic one.

---

Before you spend anything

One thing to weigh before authorising a big service bill on a vintage Tudor: the movement is rarely what determines the watch's value. The dial is.

If your 7016 or 7021 has an original Snowflake dial in decent condition, that dial is worth more than everything else combined — and a Rolex Service Centre will very likely replace it and not give it back. That's the subject of the next post in this series, and it's worth reading before you make any service decisions.

---

Sources: Omega Forums — On my bench: Tudor Submariner 7928 · Omega Forums — Tudor 390 restoration · Watch Repair Talk — Tudor c.390: rebanking, low amplitude · Watch Repair Talk — FEF 380 mainspring · Rolex Forums — How's your Tudor 390 running? · Rolex Forums — 1960s Tudor Submariner, what are my options · EmmyWatch — Tudor 390 · Revolution — The Tudor Oyster story