VOC gas sensor inside an enclosure: vents, membranes and lag

Short answer: house the sensor in a sealed stainless box and put a PTFE membrane breather in the wall 0.4 in [10 mm] from the inlet, with a clear opening about 10x the sensor's gas port. For a 2 mm [0.08 in] port that is about 31 mm² — one 6.3 mm [0.25 in] circle or two 5 mm [0.2 in] holes. Behind that vent, a sensor with a free-air T90 of 10 s (the time to reach 90% of a step change) still answers in the tens of seconds.

Seal the box with no vent and the same step change stretches to minutes, the final reading settles low, and the humidity element climbs at power-up. The sensor behind the numbers here is a metal-oxide VOC unit that also measures temperature and humidity. Its 2 mm port and 10 s response are example inputs, so swap in your datasheet values and redo the two sums.

What do you need to house a VOC sensor in a sealed box?

You need four parts: a sealed stainless box, a PTFE membrane breather with a clear opening about 10x the sensor's gas port, a rated cable gland for the sensor leads, and standoffs that hold the port within about 10 mm of the membrane. Every item on the bench list serves one of two jobs: keeping the box's rating or keeping the gas path fast.

• Box. Any of the 304 stainless wall-mount enclosures from 6 x 4 x 4 in up gives you the gasketed door and removable panel. The worked example uses the 8 x 6 x 4 in 304 stainless wall-mount enclosure, 8 x 6 x 4 in (203 x 152 x 102 mm) outside, which leaves about 183 in³ (3.0 L) of interior volume for the board and the air gap.
• Membrane breather. An ePTFE membrane with 31 mm² or more of clear opening for the example port. Threaded breather bodies and adhesive patches both work; choosing a pressure equalization membrane vent covers the hardware differences.
• Cable gland. One rated to the same class as the box and sized to grip the cable's outside diameter; outdoor sensor cable connections for a sealed enclosure covers gland choice and drip loops.
• Standoffs. 10 mm [0.4 in] spacers that hold the board with the port square to the membrane.
• Bench kit. A reference temperature and humidity probe, the test gas, and a logger running at 1 Hz for the response test.

How big should the membrane vent be?

Size the vent's clear opening to about 10x the sensor's port area; a 2 mm port needs about 31 mm² of membrane, which is one 6.3 mm [0.25 in] circle or two 5 mm [0.2 in] holes. Area does the work; thread size is just packaging.

The example port measures 2 mm [0.08 in] across, so its area is π/4 x (2 mm)² = 3.1 mm². Ten times that is 31 mm², and a round opening of that area measures √(4 x 31 / π) = 6.3 mm. Order a 6.5 mm [0.26 in] hole for margin; it gives 33 mm². Scale with the port: a 4 mm [0.16 in] port needs 126 mm², a 12.7 mm [0.5 in] circle.

The rule comes from transport resistance, which falls in proportion to open area. At 10x the port, the membrane adds a small share of the resistance the sensor's own inlet already has, so the datasheet response survives. Cut the opening back toward 1x and the membrane's resistance rivals the inlet's own; the wait roughly doubles. Published work on fluoropolymer sensor membranes shows the same lever: the membrane's diameter, not the housing volume, sets how fast gas reaches the element behind it.

The membrane itself has to be ePTFE — PTFE expanded into micropores that pass gas and vapor and stop liquid water. Solid PTFE sheet is a seal and a gas barrier, so a machined PTFE plug protects the box and kills the response. Oversize the vent wherever oil mist or paint overspray can reach it, because fouling cuts the open area. The difference between solid and expanded PTFE, including what passes and what stays out, is the subject of this PTFE versus ePTFE permeation comparison.

Where do you put the vent and the sensor?

Mount the membrane in the side wall 10-20 mm [0.4-0.8 in] from the sensor inlet, port facing the membrane: a 10 mm [0.4 in] air gap adds about 6 s of diffusion lag, while 50 mm [2 in] adds about 2.5 min.

The lag comes from diffusion, and diffusion scales with the square of distance. Take D = 0.08 cm²/s for the example, the order of the diffusion coefficient of toluene vapor in air at 25 °C (77 °F); ethanol, a common calibration vapor, runs about 0.12 cm²/s. Across a 10 mm gap the wait is (1 cm)² / (2 x 0.08 cm²/s) = 6 s. Stretch the gap to 50 mm and it becomes (5 cm)² / 0.16 cm²/s = 156 s, about 2.5 min. Water vapor moves about 3x faster, near 0.25 cm²/s, so the humidity element leads the gas element through the same vent.

1. Mark the vent hole on the side wall at the height of the sensor port, at least 15 mm [0.6 in] clear of folds, welds and edges, so the breather's gasket seats on flat sheet.
2. Cut stainless slowly with a hole saw or step drill and cutting oil. Deburr both faces and wipe the swarf off; tooling dust left on 304 shows up later as rust stains.
3. Mount the board on 10 mm [0.4 in] standoffs on the removable panel, inlet square to the membrane, and hold the gap at 10-20 mm. Keep desiccant bags, cable bundles and potting compound out of that line.
4. Lay the vent and gland holes out in the cutout designer and send the layout in with the quote request. Hole spacing and edge-distance fundamentals sit in the cutouts, glands and windows guide.
5. Treat the cut as the moment the rating goes on the line; what a panel keeps after you cut it and mount devices shows which parts decide the number the finished assembly holds.

How do you seal the box without losing the rating?

Fit a breather and a gland that their manufacturers rate for the box's own rating, because the weakest penetrator sets what the finished assembly holds: a Type 4X / IP66-rated breather with its gasket seated flat and pulled down evenly.

Our stainless boxes are rated NEMA 4X / IP66 by the manufacturer; not UL listed, and NEMA publishes the type definitions without testing enclosures. The second digit of IP66 means powerful jets under IEC 60529 — a 12.5 mm nozzle at 100 L/min — and that is the duty every hole you add has to survive.

6. Fit a membrane breather whose maker documents the same hose-down class as the box, gasketed and tightened per its instructions. A breather rated below the box pulls the whole assembly down to its own rating; whether breather vents keep a NEMA 4X rating works that rule through.
7. Pass the sensor cable through a gland rated to the same class and sized to the cable OD, with a drip loop outside so water runs off the sheath instead of into the fitting.
8. Close the door on a clean, undamaged gasket and pull the screws down evenly, working opposite corners; a loose corner shows up as fogging after the first washdown.

For the gas-sensor case in particular, the write-up on venting a gas sensor without losing the rating covers which vent styles hold up in washdown service and which trade the rating away.

How do you test the response after the box is closed?

Step the boxed sensor from clean air into test gas and log at 1 Hz: against the example sensor's 10 s free-air T90, accept up to about 30 s behind the membrane and hold the humidity element within ±5% RH of a reference probe. Run the bare sensor first, then the closed box, and read these numbers off the two logs:

1. Free-air reference. Bare sensor at 25 °C (77 °F) and 50% RH, logged at 1 Hz, stepped into the test gas; the example sensor gives T90 ≈ 10 s.
2. Boxed T90. Repeat the same step with the box closed. The example band accepts up to about 30 s, three times free air; a trace that runs to minutes means cut the gap or open the area a size.
3. Sensitivity change. Compare the steady-state boxed reading with the bare reading for the same gas concentration; in this example, accept within 30%.
4. Temperature and humidity deviation. Hold the boxed humidity element against the reference probe for 5 min; the example band is ±5% RH. The sensor's own heater adds self-heat around the package — 2-3 K in this example, so measure yours.
5. Long-term stability. Rerun the first two steps at the yearly shutdown and after any paint or silicone work near the vent.

An unvented box fails this test in a recognizable way: the humidity element reports the trapped air rather than the plant air, which the article on a humidity sensor in a sealed enclosure works through. Keep heaters and power parts off the sensor's standoffs; the placement rules for SHT-class temperature and humidity sensors apply just as much to a combined gas package.

Which mistakes cost response time or the rating?

The expensive mistake is sealing the sensor behind solid plastic or leaving the box unvented: the same step change that took 10 s in free air then takes minutes, settles low, and the humidity reading climbs at power-up. The rest of the list costs you bench time instead:

• Sensor buried in the box. 50 mm [2 in] of dead air adds about 2.5 min by the same arithmetic; hold the inlet at 10-20 mm [0.4-0.8 in] from the membrane.
• Vent under the rule. Halving the clear opening doubles the transport lag, so an opening at 1x the port already doubles the wait. Hold the 10x line: 31 mm² for the example port.
• Sealed with no vent. The sensor adsorbs from a fixed 3.0 L of air with nothing to replenish it: minutes of lag, a final reading that undershoots, and an RH trace that climbs as the electronics warm and drive moisture off the boards and gasket.
• Fouled membrane. Oil mist and paint overspray cut the open area; shelter the vent, or re-test T90 after any spraying near the box.
• Unrated penetrations. A plain drilled hole or an undersized gland becomes the assembly's weak point at the first hose-down; every hole gets a rated fitting.
• Blocked inlet. Potting or coating over the port turns the port itself into the barrier, and no membrane downstream fixes it.

What do you put on the order?

Order the 8 x 6 x 4 in 304 stainless wall-mount enclosure with one vent hole and one gland hole factory-cut on the same side wall, and buy the membrane breather and gland against the ratings in this table:

Send the drawing instead of cutting on site: factory-cut holes arrive deburred, and the spacing is checked by the layout tool before the sheet is cut. The custom stainless cutouts page lists what to send with the request. The box is delivered in 10 business days, anywhere in the world, counted from your approval of the drawing.

Catalog dimensions to compare

Compare these catalog dimensions with your component layout. Open each enclosure for its rating, material and usable space, then check those specifications against your application.

Questions