Outside vapor deposition explained for busy engineers
Outside vapor deposition sits at the center of modern preform manufacturing. When you understand how OVD builds soot, strips OH, and consolidates into dense glass, you can predict draw yield, coating behavior, and field loss before the tower heats up. In this guide, we walk through mechanisms, line choices, QA gates, and floor-ready numbers. We keep the language plain and the steps practical so your team can move from theory to stable output without guesswork.
Introduction: design "preform-first," fix problems upstream
Designing from the preform back reduces scrap and shortens cycles. The glass recipe, soot density, dehydration time, and consolidation temperature define the window for draw tension and coating cure. When the outside vapor deposition run builds clean soot and removes OH early, the consolidated billet collapses without trapped bubbles and the draw sees a steady neck-down. Your attenuation at 1310 and 1550 nm lands near target with a low tail. When OVD drifts, you fight breaks, noisy OD, and micro-bends after cabling. In the sections below, we tie each OVD stage to a check you can run on the floor.
Core concepts: what OVD does and why it works
What OVD is.
In outside vapor deposition, vapor-phase precursors form silica "soot" that deposits on the outside of a rotating bait rod. The laydown grows a porous body with core and cladding dopants set by gas flows. That soot body is then dehydrated and consolidated into a solid, transparent preform.
Why OH control matters.
Hydroxyl groups create a loss shoulder near 1383 nm and raise background attenuation. The OVD answer is chlorine-based dehydration while pores are still open. Clean gas trains and dry furnaces keep OH low. You verify with fast IR checks and trend the 1383-nm signature.
What "good" looks like on loss and geometry.
A healthy single-mode line often hits around 0.35 dB/km at 1310 nm and 0.25–0.30 dB/km at 1550 nm on bare fiber, with spec maxima giving safety margin. Geometry matters the same way: tight OD and concentricity stabilize the neck-down and improve coating centricity.
Why throughput ties back to glass.
Draw lines commonly run 10–20 m/s, and proof programs screen strength at high line speed. Those numbers only hold if the preform collapses cleanly and the surface is free of sleeks. When the glass is right, speed is easy. When the glass is wrong, speed hides problems.
Deep dive: the outside vapor deposition flow, end to end
Step 1: Gas handling and recipe setup
Start with ultra-pure SiCl₄, GeCl₄, SiF₄, and O₂. Flow accuracy sets Δn between core and cladding. Track moisture at every joint and seal. Even a small leak shows up as a 1383-nm shoulder later. Document cylinder lots, filters, and MFC calibration. Your aim is stable soot chemistry and repeatable layer thickness per pass.
Engineer's checklist
Moisture meters at skid outlet and torch inlet
MFC zero and span check before shift
Dopant flow cross-check against RNF from last good lot
Torch traverse speed and bait-rod rotation audit logged
Step 2: Soot laydown on the bait rod
The torch hydrolyzes chlorides into nano-scale silica that deposits on the outside of the rotating bait rod. Build the core first using GeO₂ to raise index, then transition to cladding using fluorine to lower index if the design calls for low-water-peak or bend-friendly behavior. Keep soot density inside your consolidation model. Denser soot shortens sinter time but raises trapped porosity risk if dehydration lags. Lower density is easier to dehydrate but takes longer to consolidate.
Watch-outs
Layer-to-layer thickness drift
Sudden soot color change (gas purity)
Temperature hotspots at traverse reversals
Early ring width creep on trench designs
Step 3: Chlorine dehydration
Strip OH from the soot body using chlorine-bearing flows at elevated temperature. The reaction converts OH to volatile species that leave the glass. Dehydrate before pores close, or you trap water and carry a 1383-nm shoulder into draw. Keep a log of time, temperature, and flow. Add a quick IR scan on a cane sample to catch drift.
Control knobs
Cl₂ partial pressure and carrier flow
Dehydration dwell time per diameter
Furnace leak-back test and oxygen background
Cane sampling for IR at fixed axial intervals
Step 4: Consolidation into transparent glass
Sinter the soot preform into a fully dense, transparent cylinder. Zone temperatures and residence time prevent bubbles and limit residual stress. Surface finish matters: sleeks turn into break initiators on the tower.
Thermal profile tips
Use a thermo-viscosity sweep to set zones
Stage temperature ramps to avoid skin sealing
Maintain low-turbulence flow around the neck-down
Record bubble counts per axial window as a capability metric
Step 5: Collapse, machine, and pre-draw QA
After consolidation, finish geometry, check straightness, and run refracted-near-field (RNF) to confirm the index profile and concentricity. Log inclusions and surface defects. If capacity allows, pull a pilot cane and run a short proof draw to validate tension and UV cure before the first full heat.
Release gates
RNF Δn within band at multiple axial stations
OD and concentricity within release spec
Surface chip count below limit
Bubble map green across whole length
The unique OVD "5-step implementation" you can deploy this week
Lock the gas train
Purge lines, change filters, leak-check manifolds, and log moisture. This keeps outside vapor deposition chemistry steady and repeatable.
Calibrate laydown thickness per pass
Run a short laydown at target Δn. Cut and measure soot density and ring thickness. Adjust traverse and rotation speed before longer campaigns.
Dehydrate while pores are open
Time the chlorine step so OH leaves before densification. Track the 1383-nm signature with quick IR checks on cane to confirm progress.
Set consolidation zones from viscosity data
Measure viscosity vs. temperature on a sample. Use it to program furnace zones. Aim to reduce bubbles and residual stress without skin sealing.
Pilot draw and go/no-go release
Draw a short length at reduced speed, stabilize tension and UV dose, then ramp. Use OD, attenuation, and break rate as the release trio for production.
OVD versus VAD, MCVD, and PCVD: a five-dimension comparison
| Dimension | OVD (outside vapor deposition) | VAD (vapor axial deposition) | MCVD (modified CVD) | PCVD (plasma CVD) |
|---|---|---|---|---|
| Scale and throughput | High; fast soot laydown and large bodies | High; long axial boules | Medium; tube scale limits | Medium; uniform layers |
| Profile complexity | Good for step and simple graded cores | Good axial uniformity | Very high; rings and trenches | Very high; fine control |
| OH/impurity control | Strong with timed dehydration | Strong with tuned dehydration | High with clean gas | Very high; clean plasma |
| Lead-time flexibility | Medium; furnace queues dominate | Medium; long growth rigs | High for small lots | High for small lots |
| Plant footprint | Medium/large furnaces | Large growth rigs | Small benches | Medium benches |
When to choose OVD
If your mix is volume single-mode with standard cores and low-water-peak cladding, outside vapor deposition delivers the best cost per fiber-kilometer while keeping room for bend-friendly clads.
Technical details that decide loss, breaks, and cost
Soot density vs. consolidation time
Lower soot density speeds laydown but extends consolidation. Higher density shortens sinter time but can trap micro-porosity if dehydration lags. Pick the balance based on furnace capacity and your bubble map.
Dehydration chemistry
Chlorine reacts with OH, forming volatile species that leave during heat. Time this while pores remain open. The earlier you drive down OH, the less you fight the 1383-nm shoulder after collapse.
Index profile with dopants
Germanium raises the core index. Fluorine lowers the cladding index and supports low-water-peak designs. Trench or ring profiles for bend-insensitive fiber need tight Δn and repeatable ring widths. RNF mapping catches drift before the tower does.
Geometry and concentricity
Tight OD and concentricity stabilize draw tension and coating centricity. When geometry drifts, the neck-down wobbles and UV dose varies. That shows up as micro-bends after cabling.
Coating windows and environment
Standard UV-cured acrylates cover typical telecom ranges. Match the coating set to line speed and post-cure strength targets. Pair outside vapor deposition glass with a coating system that fits the duct, temperature, and bend radius you expect.
Practical checklists you can run on shift
Seven "green-tag" checks before releasing an OVD lot
IR scan near 1383 nm shows no OH shoulder growth.
RNF map holds Δn and concentricity across axial stations.
Bubble and inclusion counts clear your limit after consolidation.
Surface inspection reports zero sleeks, chips, or contamination.
Dehydration log: time, temperature, and chlorine flow verified.
Thermo-viscosity sweep aligns with furnace zones.
Pilot draw at low speed confirms OD control and UV cure.
Five KPIs to trend weekly
Breaks per million meters at proof
Attenuation tails at 1310/1550 nm
OD drift and coating concentricity over time
Preform yield per OD × length
Consolidation queue time vs. laydown hours
"Outside vapor deposition" in the field: where choices show up
Access builds with low-water-peak targets
Push hard on dehydration and keep fluorine in the cladding to pull the 1383-nm loss down. Validate typical loss early on bare fiber so cabling steps do not surprise you.
Data-center ducts and high-count cables
Tight ducts and small bend radii stress macro-bend loss. Shift OVD recipes toward trench-friendly profiles and hold Δn margins. Confirm macro-bend targets on wrap tests before release.
Harsh corridors and high temperature
If the cable sees higher temperatures, verify coating chemistry. Specialty coatings extend the operating range. Align line speed, cure dose, and post-cure strength to the route.
The unique OVD "7-item readiness list" for new product ramps
Spec review: attenuation, cutoff, zero-dispersion, and bend targets fixed.
Recipe lock: Ge and F flows tied to a reference RNF profile.
Gas hygiene: moisture limits at skid and torch signed off.
Dehydration timing: chlorine dwell matched to diameter.
Consolidation program: zones set from viscosity data.
Metrology plan: RNF stations, IR checks, and bubble maps defined.
Pilot protocol: draw speeds, proof level, and release criteria agreed.
Real-world scenario: scaling an OVD line for a low-water-peak single-mode
Context
A mid-size plant wants a low-water-peak single-mode line for access networks. The team chooses outside vapor deposition to hit cost targets and keep a path to bend-friendly claddings.
Moves
They tighten moisture control in the gas skid, add an in-process IR check after dehydration, and reset consolidation zones using a fresh viscosity sweep. They start a 12 m/s pilot draw to tune tension and UV dose, then ramp to 18 m/s with 700 MPa proof.
Results
Typical attenuation lands near 0.35 dB/km @ 1310 nm and 0.25–0.30 dB/km @ 1550 nm on bare fiber, with cabled values holding under contract limits. Yield per 200 mm × 3 m preform meets plan. The consolidation queue becomes the pacing factor, so they schedule laydown to match furnace hours.
"Outside vapor deposition" 5-step implementation flow (print-ready)
Plan
Fix attenuation targets, bend targets, and OD. Select dopant levels and index profile. Define RNF release bands.
Prepare
Purge gas lines, replace filters, and verify MFCs. Load the laydown and traverse programs. Stage chlorine cylinders.
Produce
Run soot laydown with live thickness checks. Transition core to cladding cleanly. Start dehydration while pores are open.
Consolidate
Execute zone program with dwell times from viscosity data. Map bubbles and watch for skin sealing. Finish surfaces.
Prove
Pilot draw at reduced speed. Check OD, attenuation, and cure. Ramp to production speed and proof. Release when the trio holds stable.
OVD versus VAD: five-dimension decision table
| Criterion | OVD | VAD |
|---|---|---|
| Cost per fiber-kilometer at volume | ●●●●○ | ●●●●○ |
| Lead-time flexibility | ●●●○○ | ●●●○○ |
| Index profile complexity | ●●○○○ | ●●○○○ |
| OH control potential | ●●●●○ | ●●●●○ |
| Footprint and utilities | ●●●○○ | ●●●●○ |
Dots are relative within this table and help frame trade-offs.
Practical tables you can use on shift
Typical single-mode targets you can tape to the furnace
| Parameter | Typical target | Why it matters |
|---|---|---|
| Attenuation @ 1310 nm | ≤ 0.35 dB/km | Metro budgets and OTDR margin |
| Attenuation @ 1550 nm | ≤ 0.25–0.30 dB/km | Long-haul and DWDM spans |
| Zero-dispersion wavelength | ~1302–1322 nm | Dispersion management |
| Draw speed | 10–20 m/s | Throughput vs. defect risk |
| Proof test stress | ≈ 700 MPa | Strength screen at line speed |
| Preform OD × length | ~200 mm × 3–6 m | Yield planning per setup |
Seven KPIs that link OVD to tower stability
Proof breaks per million meters
Attenuation P95 at 1310 and 1550
OD drift over length and time
Coating concentricity variance
Preform yield vs. OD × length
Consolidation defects per axial meter
Moisture level trends at the torch
Three tool types that de-risk new OVD recipes
Multiphysics furnace modeling to flatten consolidation gradients
Waveguide solvers to model bend loss on trench designs
DOE/statistics packages to map laydown rate, soot density, and bubble defects against attenuation tails
FAQ
How much does an OVD preform cost per fiber-kilometer?
Costs vary with OD, length, and recipe. Track cost per fiber-km, not cost per preform. Larger OVD bodies lower unit cost when consolidation keeps pace and break rates stay low. Plan with conservative yield numbers and update after each campaign.
How long does an outside vapor deposition cycle take?
From laydown to a consolidated, collapsed billet, expect several days to a couple of weeks. Laydown can be fast; consolidation and QA drive the schedule. A short pilot draw reduces risk before the first full heat.
What attenuation should I expect if OVD runs clean?
Plan for ~0.35 dB/km at 1310 nm and ~0.25–0.30 dB/km at 1550 nm on bare fiber, with margin against your contract maxima. Keep an eye on the 1383-nm region to confirm the dehydration step worked.
How fast can I draw fiber from an OVD preform?
Most plants run 10–20 m/s on draw and set proof to screen strength at high line speed. If breaks rise, check bubbles, surface sleeks, and cure dose before lowering speed.
What proof stress should I set?
A common setting is ≈ 700 MPa (100 kpsi) across the full length. Some programs go higher for special deployments. The key is repeatability and clean strength tails.
Where is OVD better than VAD or MCVD?
Use outside vapor deposition when you need high volume and standard cores or low-water-peak cladding. If you must sculpt narrow rings or complex trenches, inside-tube methods may cut tuning time.
What trend should guide my next OVD recipe?
High-count data-center builds and tight ducts keep pushing bend performance. That favors trench-style claddings and stable Δn control in OVD. Design for small bend radii and verify on wrap tests before release.
Does coating selection tie back to OVD?
Yes. Coating cure at line speed depends on a stable neck-down and a smooth glass surface. OVD surface quality, furnace zones, and consolidation cleanliness show up immediately in proof and in post-cable micro-bend tests.
Summary: make OVD boring, and everything else gets easier
Build every campaign around outside vapor deposition discipline. Lock the gas train, calibrate soot density, and strip OH while pores are open. Consolidate with zone control that limits bubbles and stress. Verify Δn by RNF before the first full heat, and run a short pilot draw to nail tension and cure. Do this every time and you will cut breaks, hit your 1310/1550 nm targets, and lift yield per preform. When the outside vapor deposition step is predictable, the tower is quiet, the numbers hold, and shipments leave on time.