Back to Home
Technical guides

Opto-mechanics

How to choose an optical table

Mass is not the specification. What matters is where the isolators resonate, how the top bends, and how quiet your floor already is.

Pneumatic optical table with air-spring isolator legs, air lines and levelling feet
ZTP pneumatic table — the air springs under the frame are what set f0

Isolate the floor

The legs do this, not the top. Their natural frequency decides which frequencies are reduced and which are amplified.

Resist your own forces

Shutters, stages and fans push on the table from above. Here the tabletop’s stiffness and damping matter, not the isolators.

Hold alignment

Flatness, thermal stability and a reliable thread grid are what keep a setup aligned between sessions.

Interactive isolation lab

Below √2·f₀ a table makes things worse

Support type

Transmissibility

75× down

Tabletop velocity

0.17 µm/s

Residual motion

1.33 nm

Gain at resonance

6.3×

Transmissibility curve

Amplifies below 2.1 Hz · isolates above it

Meets requirement
.001.01.11101000.5151050100Frequency (Hz)

1.33 nm residual against a 20 nm budget, with 75× attenuation at this frequency.

Red band: the table amplifiesYellow line: √2·f₀ crossoverT = √(1+(2ζr)²) ⁄ √((1−r²)²+(2ζr)²)

A rigid single-mass model of the isolators only. It says nothing about the tabletop’s own bending resonances, which set the high-frequency limit — for those, read the published compliance curve of the specific table rather than any formula.

Table catalog

The counter-intuitive part

A stiffer support is not a better one

Isolation only begins above √2 times the isolator natural frequency. Below that the mounts transmit more motion than the floor carries, worst of all at resonance where the gain is roughly 1/2ζ. A heavy table on rigid legs resonates somewhere in the tens of hertz — squarely inside the band that pumps, air handling and footfall occupy — so it can be measurably worse than the bench it replaced.

That single fact orders the whole product range. Air springs are not a luxury feature; they are the mechanism that drags f₀ down to 1–2 Hz so that everything above about 3 Hz is attenuated. Damping does not move the crossover — it only trades a lower resonance peak for slightly worse isolation far above it.

Four levels of support

Buy the one your floor and your measurement justify

Breadboard on a bench
No isolation

Breadboard on a bench

A flat, tapped, rigid surface — and nothing more. Correct when the bench is already quiet or the measurement is insensitive. Its natural frequency sits inside the building’s band, so it can amplify.

Honeycomb breadboard
Stiff for its mass

Honeycomb breadboard

A honeycomb core buys bending stiffness without the mass of solid steel, pushing the board’s own resonances up. Individually sealed holes keep spills out of the core.

Passive damped table
No compressed air

Passive damped table

Damped supports without an air supply. The natural frequency stays relatively high, so isolation starts higher up — but the resonance peak is small and there is no compressor to maintain.

Pneumatic table
Low f₀

Pneumatic table

Air springs put the natural frequency near 1–2 Hz, which is the only way to isolate the 10–50 Hz band buildings actually shake in. Needs clean, dry air and levelling.

Where a single frame will not do — a heavy instrument, an awkward footprint, an existing bench worth keeping — ZSR-P individual-isolator tables and ZDH standalone isolators let you isolate the payload rather than buy a whole table. Or browse the full catalog.

Then the top itself

What the tabletop specification is telling you

Compliance curve

The honest specification. It shows dynamic deflection per unit force against frequency, including the top’s own bending resonances. Compare curves, not thickness.

Honeycomb core

Bending stiffness comes from separating two steel skins. A thicker core pushes the first resonance up far more efficiently than adding solid mass.

Flatness

Typically quoted over the whole surface, often 0.02–0.05 mm on premium tops. It sets how much your mounts have to take up before anything is aligned.

Sealed holes

Individually cupped threads stop spills and swarf entering the core, where they cannot be cleaned out and will damp unpredictably.

Before you order

Six checks

  • Measure the floor before choosing isolators. A single afternoon with a geophone or accelerometer decides between a breadboard and a pneumatic table.
  • Identify the dominant disturbance frequency. Isolation is a function of frequency, so a table that solves 30 Hz may do nothing for 2 Hz building sway.
  • Check the payload against the isolator working range — an under-loaded air spring does not reach its designed natural frequency.
  • Confirm clean dry air, or choose passive damping. A pneumatic table fed by a wet compressor becomes an expensive rigid bench.
  • Leave the thread grid and edge access you will need in two years, not the layout you have drawn today.
  • Plan delivery: dimensions, floor loading, doorways and lifting points. A table that will not enter the room isolates nothing.

Before any quote

Contact us — we will measure your floor.

We can visit your site and run the vibration measurement ourselves, or help you carry out the tests with your own equipment. Either way you get real numbers for your room — the actual floor spectrum and the frequencies that matter — before we quote a table, rather than guessing at an isolator specification.

Configure it

Tell us the room and the experiment.

Size, support type and tabletop follow from the floor you have and the stability you need.