Watch PVD Coating: How to Spec Plating for an OEM Watch Run
Share
Two quotations land on your desk for the same black watch. Both write "PVD coating" on the plating line. One price is several times the other, and the difference is not the machine. One quote wrote thickness in microns, salt spray hours, an adhesion method and a colour reference. The other wrote three letters.
Boiled down, watch PVD coating is a vacuum process, not a liquid bath. The film is measured in microns, and specifying it comes down to four numbers: thickness, salt spray hours to ISO 9227, an adhesion rating, and a colour tolerance.
If you price a first metal line, compare plating quotes, chase colour-wear returns on Amazon, or cost a logo programme, this page is for you. We build OEM and ODM watches in Guangzhou.
Scope note: production surface finishing only, meaning PVD, electroplating, DLC and the base finishing underneath. Still deciding the base metal? Start with our watch case materials guide and our 316L, titanium and alloy comparison.
Already have a plating line on a quotation? Send it to us and we will tell you which numbers it is missing.
Key Takeaways:
- PVD is a process name, not a material, colour or thickness. A quote saying only "PVD" has specified nothing you can check.
- Four numbers decide whether a coating holds: thickness in microns, salt spray hours to ISO 9227, an adhesion rating from a named test, and a ΔE tolerance.
- Surface quality is set before parts enter the chamber. PVD reproduces what sits underneath, tool marks included.
- Colour is a rack decision, not a mold decision. Black runs easily, rose gold needs control, and small orders pay furnace time like large ones.
What PVD Is, and What It Is Not
PVD stands for physical vapour deposition. A solid target is vaporised inside a vacuum chamber and condenses onto the parts as a thin bonded film, typically 0.2 to 3 microns thick. Nothing is dissolved in a bath and no metal is reduced onto the surface electrochemically. It is a dry vacuum process, not a liquid bath, and it adds colour and hardness without changing case dimensions.
The four stages inside the chamber
- Base finishing. CNC marks, polish lines and blasting texture, all created on bare metal first.
- Cleaning. Ultrasonic degreasing, rinsing and drying. Compound left behind is the usual reason a film lets go later.
- Vacuum and activation. Parts are ion bombarded under vacuum, stripping the last monolayer so the film can grip.
- Deposition and cool down. The target is vaporised by arc or sputtering, the rack rotates, and the batch cools before opening.
What PVD is not
PVD is a process name, not a material, a colour or a thickness. Colour comes from the compound formed during deposition: titanium nitride yields gold, zirconium nitride a lighter gold, chromium nitride gunmetal, titanium carbonitride graphite, and DLC sits near black. Change the target or the gas mix and you change the colour, so a supplier can run black all week and still fumble a rose gold.
Keep PVD separate from the anti-reflective film on your crystal. AR is likewise vacuum-deposited, but it is an optical coating on glass specified by different criteria. Our sapphire and mineral crystal guide covers that side. For hand finishing, the FHH watch finisher trade page is the reference.
Thickness in context
Watch PVD coatings typically run roughly 0.2 to 3 microns depending on target, colour and brief. A human hair is around 70 microns across, so a 2 micron film sits under 3% of that. Practical consequence: coating does not change your dimensions or tolerances.
PVD vs Electroplating vs DLC: How to Choose
Three options come up in almost every plating conversation. The better question is which one matches your retail price and how the watch will actually be worn.
| Attribute | PVD | Electroplating (wet) | DLC |
|---|---|---|---|
| How it bonds | Vapour condenses and bonds under vacuum | Metal ions reduced from a bath onto the part | Carbon film, deposited by a PVD method |
| Typical thickness | roughly 0.2 to 3 microns | several microns and up for conventional builds | roughly 1 to 4 microns, in our experience |
| Typical hardness | Very high; titanium nitride commonly cited near 2,000 HV | Closer to the deposited metal, so markedly lower | Highest of the three by a wide margin |
| Colour range | Wide, driven by target and gas mix | Warm gold tones vacuum work reproduces less easily | Narrow, mostly near black |
| Colour stability | Stable, since colour is a compound | Can drift with bath chemistry and bath age | Stable, limited palette to drift within |
| Corrosion behaviour | Barrier only, so pinholes and edge coverage matter | Barrier, generally applied thicker | Barrier, plus notably low friction |
| Environmental load | Dry process, no bath to treat | Liquid effluent and sludge to manage | Dry process |
| Cost position | Mid | Typically the lowest up front | Typically the highest |
| Where it fits | Default for modern metal cases and hardware | Dress tones, budget-led programmes, low-abrasion parts | Tool and field watches, near-black briefs |
For scale on that hardness row, 316L stainless sits around 200 HV, which puts a titanium nitride film at roughly ten times the hardness of the steel underneath it.
When electroplating still wins
Wet plating keeps its place. Bath chemistry produces warm gold tones that some dress briefs want and that vacuum processes reproduce less convincingly, and it remains the lower-cost option where the case will not see much abrasion. The constraints are the effluent, the documentation burden, and wear closer to a soft metal than to a ceramic compound.
When DLC is worth the premium
DLC, or diamond-like carbon, is a carbon film deposited by a PVD method rather than a separate technology family. It earns its premium on tool and field watches, on near-black briefs where gunmetal PVD is not dark enough, and anywhere low friction matters. It is also the least forgiving of a poor base finish, because a hard, dark film shows every line in the polish underneath.
The hardness caveat
Hard is not the same as tough. A thin hard film on a more ductile substrate behaves like an eggshell: it resists scratching until the metal underneath deforms, then cracks instead of denting. Drop a coated case on a hard corner and you lose a chip of film while the base metal below is untouched.
Do not expect a coated part to behave like bare steel in later operations, either. You cannot bring a coated case back to a mirror the way you work 316L, because polishing cuts through microns of film in seconds.
The Four Numbers That Decide Whether a Coating Holds
A plating line has one job: to be checkable. If you cannot test it, you cannot reject it.
1. Thickness in microns
Standard wholesale film typically runs 0.2 to 0.5 microns, with 1 to 2 microns as a paid upgrade. Stepping from 0.5 up to 1 to 2 microns doubles to quadruples the film, which is a real change in service life. Do not accept the word "standard." A supplier's standard is whatever their default cycle produces, and that varies shop to shop. Write the agreed micron figure next to every coated part on the drawing, so the thickness travels with the file instead of living in an email thread.
2. Salt spray hours
Corrosion testing normally runs as neutral salt spray to ISO 9227. In our runs we see roughly three tiers: 24 hours entry level, 48 hours a common mid-market baseline, and 72 hours and above where corrosion returns drop off.
Worth knowing: ISO 9227 specifies the apparatus, solution and conditions. It sets neither an exposure duration nor a pass criterion. So "tested to ISO 9227" is not an answer. "48 hours neutral salt spray to ISO 9227, no base metal corrosion on significant surfaces, sampled to ISO 2859-1" is. Have the coupons from your own first production load shipped with the parts and kept, labelled, on your side: a report with no retained sample behind it settles nothing when a corrosion question comes up six months into the run.
3. Adhesion rating
Adhesion is assessed by cutting a lattice through the film, applying tape and rating how much lifts. The commonly referenced methods are ASTM D3359 and ISO 2409.
The detail most RFQs skip: adhesion failures trace back to cleaning more often than to the coating machine. A part carrying a trace of polishing compound goes in looking fine, comes out looking perfect, and lets go three weeks later. Invisible at dispatch, so the test belongs in the spec. It surfaces at case edges, the crown, the bezel edge, the buckle back and inside bracelet links. Our quality control and inspection levels page covers the sampling side.
4. Colour tolerance in ΔE
Colour is a tolerance against a reference, usually ΔE, one number for how far two samples sit apart in colour space. Answer two questions separately: within a batch, and between batches. Industry practice typically works near ΔE 1 within one load and ΔE 2 between loads, and most inspectors read anything past ΔE 3 as a mismatch once parts sit next to each other on an assembled watch.
The method matters more than the number. Keep a physical approved sample, sealed and signed by both sides. Photographs drift with the light and the screen. Memory drifts faster.
The failure pattern buyers actually see
Wear does not appear evenly. It starts at high-contact points: inside bracelet links, back of the buckle, outer bezel edge, the crown, and the lug tips.
A worked example: Say a 1,000-piece run ships on a coating spec nobody wrote down, and colour-wear returns land at 4%, which is 40 watches. Processing each one, meaning inbound shipping, inspection and repack or write-off, typically runs $8 to $15, so $320 to $600 across the run. Put a thickness and a salt spray duration on the same run and it comes back nearer 1%, or 10 watches, at $80 to $150. The difference is $240 to $450 of margin.
Not sure which thickness and salt spray tier your retail price supports? Send us your target retail and quantity and we will tell you what is realistic at your volume.
Colour, Cost and the Chamber Batch
Most plating cost arguments are about the chamber, not the colour.
Why black runs easily and rose gold does not
Black and gunmetal are the most straightforward to run: chromium-based targets, short cycles, and drift that barely moves what the eye sees. Gold and rose gold are harder, because tone depends on gas ratio, temperature and cycle length, and a short cycle reads pale while a long one runs orange. Holding a tone inside tolerance takes process control, and that costs furnace time.
The chamber is a batch, not a machine
A chamber is loaded with a rack, pumped down, run, cooled and opened. It does not care whether the rack held 80 parts or 800. Load cost is broadly fixed, so unit cost depends on how many parts shared it. Small runs either pay for a partial load or share one, and changing colour is a rack decision, not a mold decision.
A worked example: Take a brand ordering 300 cases in a custom bronze tone next to another ordering 3,000. The 300-piece order still needs its own rack, pump-down, test coupons and first-piece inspection. A single chamber load typically carries a few hundred to around 1,000 watch cases depending on case size and rack design, and the setup behind those steps costs much the same whether that rack leaves partly filled or full. The smaller buyer therefore pays a full load's setup across 300 cases, while the larger buyer spreads a comparable cost over ten times as many.
Two things follow for the smaller programme. Get the partial-load billing rule written into the quotation before you sign, because that single line decides what your next 300-piece top-up costs per case. Then draw your inspection samples from fixed rack positions every batch, since parts hanging close to the target do not pick up the same film build as the ones tucked behind them.
What actually drives plating cost
Not the colour. In our experience: furnace hours, including pump-down and cool-down; rack design and loading labour, which climbs with small parts; pre-treatment level; inspection and test coverage; and reject rate.
Lead time reality
Coating is rarely done in the same building as assembly. Sandblasting, anodising and plating commonly sit with specialist subcontractors under their own environmental permitting, and published accounts from other factories put two to three extra days on the schedule. Confirm that in writing.
What the Base Metal Does to the Coating
The same coating on three substrates is three different products.
On zinc alloy, the coating is the only barrier
Alloy has little intrinsic corrosion resistance. The film does all the work, and once it is breached the metal underneath has nothing to fall back on. That makes thickness and salt spray hours worth more here than anywhere else in the BOM, and makes adhesion critical, because a lifted edge on alloy is not cosmetic. Our 316L, titanium and alloy comparison walks through the failure sequence.
On 316L, the coating is reinforcement
Steel already resists corrosion. On 316L the film mostly delivers colour and surface hardness, and a scratch through to bare steel is an appearance problem rather than a structural one. On steel the same spec simply performs better, and a steel programme can run thinner without paying for it.
On titanium, adhesion is the whole game
Titanium passivates quickly, and the oxide layer it forms does not help a film grip. In our experience it needs more aggressive activation and tighter control of the gap between cleaning and loading. Get it right and the result is excellent. Get it wrong and the failure is a preparation problem wearing a coating problem's clothes.
Surface prep happens before colour, not after
The most common disappointment in first samples. PVD faithfully reproduces the surface underneath. It does not fill scratches and does not hide tool marks. A mirror-polished base comes out as mirror gold; a base with CNC lines comes out as gold with CNC lines, more visible because the film is uniform. When a first sample looks wrong, look upstream at polishing.
One Colour Across Case, Bezel, Bracelet, Crown and Hands
A watch is rarely one coated part. A typical programme coats the case, the bezel or its insert, the bracelet or buckle, the crown, the caseback, and sometimes hands and indices. Each can come from a different sub-process and, at worst, a different chamber.
Same chamber or separate chambers?
Same chamber, same load, same rack gives the closest thing to identical colour the process can deliver. Separate loads introduce drift, and drift is what a customer sees holding the buckle next to the case. Ask one direct question: which parts go in together? If the bracelet runs at the strap supplier and the case at the coating house, you need a physical reference moving between them before the first load.
Where mismatch shows first
- Buckle against bracelet, because they sit adjacent and the eye goes straight there
- Bezel against case, particularly with an aluminium or steel insert
- Crown against case, since a small part from another batch reads as a different metal
- Hands and indices against the case tone, adjacent in every product photograph
Our strap and bracelet manufacturing guide and bezel and insert materials page explain why those two are the usual offenders, our crowns and pushers spec guide covers why a small part causes outsized arguments, and the same applies to dials and hands.
Keep a physical approved sample
One sealed reference per part type, held by both sides, compared under the same light at every batch. Photographs are not references.
A worked example: A 1,000-unit programme specifies gunmetal on case and bracelet. The case shop coats on Monday; the bracelet supplier runs links in another chamber on Thursday, matching a photograph. Mismatch returns start around 8%, or 80 watches, each a full return. Send one sealed sample to both shops first and that usually falls to near zero.
Compliance: What a Coating Means for EU and US Programmes
On skin-contact parts, coatings are a compliance decision, and this is the dimension most plating pages skip.
Nickel release on skin-contact parts
Nickel release is a European compliance item on parts that sit against skin: on a watch, the caseback, buckle, bracelet and crown. An intact film sits between the metal underneath and the skin. Stated conservatively, an intact coating forms part of how nickel release is managed on those parts; a worn one is not something a compliance file can rely on. We hold nickel release documentation alongside our RoHS and REACH files, so ask at quotation stage.
RoHS and REACH documentation
Ask for report numbers, not a declaration on letterhead. Customs officers and marketplaces both ask for the document, and a statement with no number behind it survives neither. Our watch manufacturing cost breakdown covers what that testing runs and who usually pays.
US programmes
US buyers generally look to CPSIA, and on coated or printed parts to California Proposition 65 limits on lead. The advice does not change: list the parts carrying a coating or a print, and ask which were tested.
Who pays for testing
Settle this before signing, because policy differs between factories and programmes. Some include it in unit price, some bill per model, some expect the buyer to commission it. None is wrong. Not asking is.
How to Write the Plating Line in Your RFQ
Everything above collapses into one block you paste into an RFQ.
The line to copy
`` Process: PVD (arc or sputter, state which) Target / colour: e.g. TiN gold, CrN gunmetal, DLC Colour reference: physical approved sample no. __, sealed, held by both parties Thickness: __ microns, measured, on the following parts Salt spray: __ hours neutral salt spray to ISO 9227, pass criterion: __ Adhesion: cross-hatch and tape test per ASTM D3359 or ISO 2409, rating __ Parts coated: case, bezel, caseback, crown, buckle, bracelet, hands (list each) Pre-treatment: polish / brush / blast to approved sample, before coating Batch rule: all parts listed above coated in the same chamber load Test report: required with each shipment, with report number ``
The questions that expose a thin quote
Four questions, in writing, settle most comparisons.
- "What is your standard thickness in microns?" If the answer is "standard," you have learned something.
- "How many hours of salt spray, and what is the pass criterion?" Hours with no criterion is half a specification.
- "How is adhesion tested, and what rating do you ship?"
- "How is the approved sample sealed, and who holds it?"
A supplier answering all four in numbers is usually the one priced higher. Our guide to evaluating a watch factory carries the longer version, and our watch manufacturing guide maps the full OEM picture.
What to lock after sample approval
Lock the parameters, not just the recipe. The colour you approved came out of a specific target, gas ratio, cycle length and rack position. Shift the rack or shorten the cycle to hit a date and you get a colour that is close but not the same, and "close" is what generates the return.
The twelve-line cost checklist in our cost guide holds the plating spec at line four.
FAQ
Is PVD coating scratch-proof?
Not scratch-proof. A PVD film is scratch-resistant and typically several times harder than the steel underneath, but it is a thin ceramic layer. Drop a case on a hard corner and the film can chip while the metal below is undamaged.
How long does PVD coating last?
There is no honest number in years, because it depends on thickness, base metal and where the watch is worn. What we can say is where it fails first: bracelet links, buckle backs, bezel edges and crowns. Specify thickness and salt spray hours rather than a lifespan.
Is PVD the same as gold plating?
No. Gold plating is a wet electrochemical process that deposits metal out of a bath. PVD is a dry vacuum process where a compound such as titanium nitride forms the colour. Same visual goal, different bond, different thickness, different wear behaviour.
Can a worn coating be redone on a finished case?
In production terms, no, not economically. A worn film cannot be polished out or topped up the way a steel case can. Stripping back to base metal and running the cycle again costs more than the part does at OEM volumes. Price coated parts as consumed and spec the film up front.
Does PVD change case dimensions?
No. Watch PVD films are measured in microns, typically well under 3, against a human hair of roughly 70. Nothing measurable changes: a 41mm case stays 41mm, gasket seats stay as machined, and bezel and crown fits are unaffected.
What MOQ does a custom colour need?
More than a standard colour, because a custom tone needs its own rack, test coupons and first-piece inspection regardless of quantity. From what we ship, a custom colour programme starts to make sense at the same order of magnitude as full custom work, so plan around MOQ 300.
Conclusion: The Coating Line Is Where the Sample and the Run Diverge
The component series now covers the case, the crystal, the strap, the bezel, the dial and the crown, and this page ties their surfaces together. The material decision sets what the watch is. The coating spec sets what it is worth on a shelf and how long it keeps looking like the sample you approved.
Before you send that RFQ, fix three lines. Write thickness in microns rather than accepting the word standard. Name the salt spray hours and the pass criterion, since ISO 9227 sets neither. Keep a physical sealed sample for every coated part, not only the case.
Request a wholesale quote from JOEFOX. Send the plating spec line with your drawing, quantity and target retail, and we will tell you what each colour and thickness costs to run. OEM and ODM from MOQ 300, tooling and certification as separate lines.