Alternatives to Cognex Vision System Camera Optics: 6 Parameters to Verify

Replacing the fixed-focal lens on a running vision station is rarely blocked by availability. It is blocked by four numbers the original datasheet never printed. Within the POMEAS global company intelligence set of 2,602 companies, 119 have public material indicating Cognex vision systems in use — numerous enough, and old enough, that a component end-of-life cycle is routine.

On an integrated system the camera, optics and software are specified together, so the lens often carries a system part number rather than an optical specification. Whoever inherits the station has a mount and a focal length and nothing else: the two parameters easiest to read off the barrel, and the two least useful for predicting whether a substitute reproduces the original image.

The six parameters to verify before substituting

The first three decide whether a substitute forms an image on your sensor at all; the last three decide whether that image holds focus, geometry and the environment.

1. Focal length and the sensor format it is built for

Focal length defines field of view only with the sensor's active area and the working distance the mechanics allow. FA lenses carry no magnification figure, so the thin-lens estimate applies:

FOV ≈ sensor size × working distance / focal length

Use it to shortlist focal lengths and check a supplier's arithmetic; it ignores principal-plane separation and focus extension. Capture the focal length and the format the old lens was corrected for — the second is the one usually missing.

2. Image circle versus sensor size: the vignetting check

Every lens is corrected out to a maximum image circle; outside it there is no image. POMEAS publishes coverage as format, active area and circle — 2/3-inch as 8.8 × 6.6 mm with an 11 mm circle, 1-inch as 12.8 × 9.6 mm with a 16 mm circle, 4/3-inch as 17.6 × 13.2 mm with a 22 mm circle — or as a single diameter, as on the 1.1-inch series at 17.6 mm.

The check is a comparison: the circle has to meet or exceed the diagonal of the camera's active area. Less vignettes, and the corner becomes unmeasurable before it looks dark.

3. Distortion, read at the format you actually use

Distortion is published per format at a stated image height, because a smaller sensor samples a smaller part of the lens's own field. PMS-LFEH-1216M(A) is −0.53% @ y=5.5 mm on 2/3-inch and −0.35% @ y=4.0 mm on 1/2-inch: two measurements, neither of them "the distortion of the lens" alone. The spread across a line is a design consequence, not a quality signal — PMS-LFGH-1616(A) is −3.73% @ y=8.0 mm on 1-inch, PMS-LFEHH-2528M(B) is −0.016% @ y=5.5 mm on 2/3-inch.

4. Aperture range and closest imaging distance

Aperture range sets the exposure and depth-of-field envelope; closest imaging distance decides whether the lens focuses on your fixture at all. Across the FA line apertures run F1.4–F22 to F2.8–F16 and distances from 0.1 m to 0.9 m.

The pages use two conventions, not interchangeable: a closest imaging distance (0.1 m on PMS-LFEH-1216M(A)) versus an imaging or focus range (1 m to infinity on PMS-LFGH-1616(A)). A lens starting at 1 m will not focus at 0.4 m whatever its focal length suggests.

5. Back focal length in air: the number C-mount swaps forget

Back focal length (后焦距) is the distance from the lens mounting face to the image plane, measured in air. It decides whether the rear of the substitute clears the camera's front structure and filter glass, and where the focus range lands.

It is model-specific: 10.31 mm on PMS-LFEHH-2528M(B), 12.42 mm on PMS-LFEH-1216M(A), 13.54 mm on PMS-LFGH-1216M. The PMS-LFPHM series publishes an optical (in air) and a mechanical figure — 12.7 mm and 11.3 mm — plus a 6.2 mm protrusion. Compare it with the flange distance on the camera datasheet.

6. Mount, filter thread and operating temperature range

C-mount is the interface across the line, and the 20 MP series states the thread as C-Mount, 1-32UN-2A. Two details still catch people out: filter thread is not shared across a series — M37.5×0.5 here, M27×0.5 on the 10 MP 25 mm, M40.5×0.5 on the 12–36 mm zoom — so a substitution can orphan the polariser already fitted; and temperature is published as −10 °C to +50 °C on most of the line, −20 °C to +60 °C on PMS-LFHHG, −30 °C to +70 °C on PMS-LFPHM.

Writing a replacement specification when the original part is discontinued

Specify the function, not the part number: sensor format and active area, field of view at the available working distance, the focal length that follows, largest acceptable distortion with its image height, closest imaging distance, mount and rear clearance, filter thread, and temperature band.

Mark each line must-hold or preferable — coverage and distortion are usually must-hold, aperture range usually preferable, because exposure time absorbs a stop. Require the supplier to answer at your format: "low distortion" without a percentage, a format and a y-value cannot enter an uncertainty budget.

Where POMEAS optics land on those six

The two 5 MP 2/3-inch models bracket the short-range case at 12 mm and 25 mm and share the M37.5×0.5 thread, so an existing filter carries over. PMS-LFGH-1616(A) is the 1-inch option and separates the two decisions: it covers 12.8 × 9.6 mm against a 16 mm circle, but −3.73% @ y=8.0 mm belongs in the calibration plan.

The 25 mm pair answers a resolution upgrade at one focal length: a 10 MP 2/3-inch part with the lowest published distortion here and a stated 420–1000 nm band, against a 20 MP 1.1-inch part at 126 ± 5 g across −30 °C to +70 °C, with a 6.2 mm protrusion to check against the camera face. PMS-LFEH-123628M is the answer when the field of view is not yet known: 12–36 mm covers 50.8° to 17.2° on 2/3-inch. Figures in different rows are measured on different formats and image heights and are not to be read as if obtained under one set of conditions.

What those numbers mean on the line

Scenario 1 — Short-range 2/3-inch station, field of view is the only requirement

PMS-LFEH-1216M(A) at 12 mm reaches 0.1 m and gives 48.6° × 39.8° × 30.4° on 2/3-inch, or 36.4° × 29.5° × 22.3° on 1/2-inch. If the station later acquires a measurement task, −0.53% @ y=5.5 mm is the first figure to revisit: no pixel count recovers a geometric error already in the image.

Scenario 2 — The camera was upgraded to 1 inch and the corners went dark

A coverage failure, not a resolution one. PMS-LFGH-1616(A) restores coverage to 12.8 × 9.6 mm against a 16 mm circle and publishes the 2/3-inch angles too (30.8° × 23.2°), so a mixed fleet can run one part number. Its −3.73% @ y=8.0 mm means a calibration pass if the station measures.

Scenario 3 — The output of the station is a dimension

Distortion moves to the top of the list and coverage moves down, because the part is usually centred. PMS-LFEHH-2528M(B) publishes −0.016% @ y=5.5 mm at 10 MP with a 0.15 m closest imaging distance, leaving the uncertainty budget dominated by sampling, illumination and edge definition.

Scenario 4 — Sensor resolution went up and the detail did not

A 20 MP camera behind a 5 MP lens produces a 20 MP image of a 5 MP optic. PMS-LFPHM-2528M(B) matches the class to a 1.1-inch sensor at 17.6 mm, focuses from 0.1 m and covers approximately 94.4 × 75.4 × 56.5 mm at the close limit on 1-inch. Its 0.50% distortion is published without a format, so confirm it before gauging.

What this comparison does not tell you

This is a parameter reference, not a compatibility statement. No claim is made that any POMEAS FA lens is a drop-in replacement for the lens fitted to a Cognex camera or vision system, and none should be inferred. Whether a substitute works depends on the mechanical interface, the flange geometry, the illumination and the algorithm's tolerance.

Cognex appears here only as recorded in customer-supplied intelligence: within the POMEAS global company intelligence set of 2,602 companies, 119 have public material indicating Cognex vision systems in use. No Cognex specification is quoted, because none was independently verified.

Optical resolution is not measurement accuracy. Resolution is a property of the lens, the wavelength and the sensor pitch; accuracy additionally depends on contrast, edge definition, illumination stability, calibration and the algorithm. A megapixel rating is a design target for a sensor class, not a promise about the smallest dimension a station can hold.

The existing POMEAS article on C-mount lens selection covers the generic method — choosing a lens from scratch, the focal length formula and the full FA series matrix. This page is the narrower case of replacing a lens already on a camera.

FAQ

Can I fit any C-mount FA lens on a Cognex camera?

Not on mount alone. C-mount fixes the thread and seating; it does not fix the image circle, the back focal length or the focus range. Treat the mount as the first check, not the last.

Why do the corners go dark after a lens swap?

The substitute's image circle is smaller than the sensor. A 2/3-inch POMEAS lens has an 11 mm circle; a 1-inch sensor needs 16 mm. Compare the circle with your sensor's diagonal before ordering.

Which distortion figure should I compare?

The one published at your format and a stated image height. PMS-LFEH-1216M(A) is −0.53% @ y=5.5 mm on 2/3-inch and −0.35% @ y=4.0 mm on 1/2-inch; a bare percentage cannot be compared with anything.

Is a higher megapixel lens always better?

Only if the camera can use it. A 20 MP lens on a 5 MP sensor buys nothing, and a 5 MP lens on a 20 MP sensor wastes the sensor.

What is back focal length and why does a swap care?

It is the distance from the lens mounting face to the image plane in air. POMEAS publishes it per model, from 10.31 mm on PMS-LFEHH-2528M(B) to 13.54 mm on PMS-LFGH-1216M. Too short and the rear element may foul the camera.

What goes into a replacement spec when the original is discontinued?

Sensor format and active area, field of view at the available working distance, focal length, maximum distortion with its image height, closest imaging distance, mount, filter thread and temperature band.

Next Step

If you are specifying a replacement and need measured values for your own conditions — your sensor size, your working distance, your feature size — send us those four numbers and we will run them against the current catalogue and tell you what actually fits, including when nothing does.