Replacing a fixed Cognex DataMan or KEYENCE SR reader: what has to match

In short

Start from what the installed reader delivers: its field at the working distance, the codes it reads, and the data and timing your programmable logic controller (PLC) receives. Take these from your drawings, settings backup and PLC program, check the gaps on the unit, and add wiring, power and mounting. Compare each item with one exact variant of the new reader; our analog tool compares camera parameters only, so the models it shows are starting points. Switch only after the new reader matches the old one on the same parts, with your PLC and reject.

Record the station before you compare readers

Start with what your drawings, settings backup and PLC program already record, then check each item on the unit and the line. Take the exact model and firmware from the label and the setup software. List every enabled code type, including any direct part mark (DPM) mode for codes marked on the part itself rather than on a label.

Measure the smallest and largest module (one cell of a 2D code or the narrowest bar of a 1D code), the symbol size and the quiet zone, the clear margin around it. Add contrast and finish, the working distance range, how far the code wanders sideways and in depth, its tilt, the codes per trigger and the exact string the PLC expects. Record the built-in light, its colour, the front cover and any external light, because they come with the exact variant. A datasheet line that says Data Matrix covers one entry of this list; every other entry needs its own check.

Match field and pixels per module at the same distance

Pixels per module is the number of image pixels across one module. Our calculators divide the horizontal pixels by the field width to give pixels per millimetre; multiply that by the module size. Compare both readers at the same physical field and working distance.

Each reader's own lens sets its field and focus at your distance, and more pixels spread over a wider field can leave fewer on each module. Check the result against your decoder's documented range, as shown in code size, camera resolution and pixels per module.

Line speed sets several separate limits

A presentation is one arrival of a code at the station, with its trigger and its result. Check presentations per second, images per presentation, exposure short enough to limit blur, processing time, the delay to result and reject, and how often the light may pulse.

The MV-ID3016XM datasheet lists a maximum frame rate of 60 frames per second (fps) and a maximum reading speed of 110 codes/s (datasheet). These are different quantities, and both are manufacturer maximums under stated settings, not results on your mark. Work out exposure with exposure, motion blur and strobe and confirm timing on the bench.

The PLC contract is more than a protocol name

The PLC contract is everything the control logic expects from the reader: signals, network data, timing and reject behaviour. The DataMan 370 manuals list EtherNet/IP, PROFINET, SLMP, Modbus TCP and CC-Link, with a trigger by discrete input or by command. A matching protocol name on the new datasheet only starts the check.

Compare where results sit in cyclic data, exchanged every bus cycle, and in acyclic data, read on request. Compare which device opens the connection, and the device description file the PLC imports (GSD for PROFINET, EDS for EtherNet/IP) with its firmware. Then compare delimiters, result timing, the no-read string and how the reject stays in step with the part. This holds even within one brand: Cognex notes that moving a DataMan to PROFINET Class B firmware requires a hardware configuration update in the Siemens TIA Portal project. Plan PLC work from the start.

Wiring, power, mounting and settings

The old cable and bracket fit only if each of these points matches the exact variant:

• Supply voltage and current match, including Power over Ethernet (PoE) if the old reader used it.
• Connector type, coding and pinout match the existing cable.
• Inputs are opto-isolated (separated by an optocoupler) or non-isolated as your wiring expects, and the outputs can drive their loads.
• Trigger polarity, pulse width and latency suit whatever fires the reader.
• Mounting-hole pattern, housing, protection rating, cable exit and clearance fit the station.

The MV-ID3016XM record, for example, lists three opto-isolated inputs, three opto-isolated outputs, a 24 VDC supply and Fast Ethernet (datasheet); set these lines beside your wiring diagram.

Back up the installed settings and note the firmware first; Cognex Setup Tool can export a DataMan's parameters. Use that export as the list you rebuild in the new reader's software, retest every rule, and keep the old unit until the new one passes.

What our analog tool checks

The analog tool takes an exact competitor model and runs Hikrobot models through a strict filter. Device kind, category, colour, spectrum, shutter type and interface family must match exactly. Width and height in pixels must each fall within 0.88 to 1.12 times the original, pixel size on each axis within the same band, maximum frame rate at least 0.8 times and interface speed at least 1.0 times (tool implementation). That rate is camera frames per second, not codes per second.

Each axis counts on its own. Our data for the DataMan 362 give 1280 × 1024 px (Cognex documentation) and MV-ID3016XM has 1408 × 1200 px (datasheet), so the width ratio of 1.10 passes and the height ratio of 1.17 fails (calculated). These limits are the tool's current screening settings, not an engineering rule, and a reader with another pixel count or interface can still suit your station if the field, the PLC contract and the paired test hold.

A missing field stays unknown, and a failed gate or a blocking unknown keeps a model out of the strict result. A model that passes every gate without a confirmed sensor part appears as a functional candidate, with compromises such as lower resolution listed beside it.

When the strict filter finds nothing, the page shows reference cards instead: near matches marked "Reference comparison, suitability not assessed". Our tools API returns only the strict result, so an empty API answer and a page with cards come from the same search. No gate looks at code types, DPM, light, discrete inputs and outputs, power, mounting, pinout or PLC behaviour, so every model the tool shows still goes through the checks above.

What the tool showed for DataMan and SR readers

On 25 September 2026 the strict filter accepted no Hikrobot model for any of three exact models we entered, and the page showed three reference cards for each.

None of these cards is a tested replacement. MV-ID3030XM, for instance, appears for DataMan 374Q/X although the strict filter rejects that pair, as the worked example shows.

Our data for the SR-X300 hold 1920 × 1200 px and 100BASE-TX, which is Fast Ethernet at 100 Mbit/s (KEYENCE documentation). Colour, pixel size, full-frame rate and shutter type are missing from it, and those unknowns keep every model out of the strict result. The lists follow the tool index, so run your exact model again when you plan the swap.

Prove the swap on your parts

Accept the new reader with a paired test: the old and the new reader read the same parts at the worst distance, angle, speed, surface and light. Give each part a traceable expected string. Log correct reads, no-reads, wrong reads, duplicates, late results and results assigned to the wrong part separately, then check what the PLC and the reject do with each outcome.

Check the decoded data against the expected string on every part, a task the GS1 DataMatrix guideline keeps separate from grading the printed symbol. That guideline does not set how many parts this acceptance needs. Agree the sample size and the allowed outcomes under your plant, customer or sector requirements.

Worked example: DataMan 374Q/X on paper

Would an MV-ID3030XM keep the sampling of an installed DataMan 374Q/X? Assume both see the same 100 mm field at the station's working distance and the code has a 0.30 mm module.

The sampling matches on paper, provided both lenses really give a 100 mm field at your distance, and each decoder's documented range decides whether 6.14 px per module is enough.

The strict filter still rejects the pair: the API recognises DataMan 374Q/X but returns the status partial and no candidates. Gigabit Ethernet at 1 Gbit/s on the DataMan against Fast Ethernet at 0.1 Gbit/s on the MV-ID3030XM fails the interface gates (Cognex documentation; datasheet), and the DataMan's colour is unknown to the tool. On the page, MV-ID3030XM appears only as a reference card.

Next, measure the field the installed reader really covers, for example from its own image of a ruler at the code plane. Send us that distance and we confirm the MV-ID3030XM field there; then recalculate and take the pair through the checks above, including the network change, and the paired test.

Calculate for your case

Open the analog tool with DataMan 374Q/X or SR-X300 already entered. The link carries only the model name; the page runs the search when it loads and shows the reference cards from the table above. The five-candidate cap in our API examples applies to the API only. For your own reader, type the full designation from its label.

Common mistakes

• A reader is chosen for more pixels while its field at your distance is wider, so fewer pixels land on each module and small codes stop reading. Compare pixels per millimetre at the measured field.
• The protocol name matches, but the PLC still expects the old byte layout, delimiter and no-read string, so results land in the wrong place or the reject fires on the wrong part. Map every field, import the description file for the exact firmware and test reject timing.
• The drawing says DataMan 370, a family of several models, so the tool returns an unknown model and the wrong variant gets quoted. Read the full designation and firmware from the unit itself.
• A reference card or a pair from an old cross-reference table goes on the order, and the reader arrives with a different field or interface. None of the old-table pairs we checked passes the strict filter; put every such name through the checks and the paired test.

When this rule breaks

If the old reader has already failed, you cannot measure its field or run a paired test. Rebuild the specification from the backup, drawings, PLC program and parts with known strings, and test the new reader against those strings.

If the product, the mark or the line speed changes at the same time, you are choosing a new reader, not replacing one. Start from choosing an industrial code reader.

If codes fail after the swap, work through why a code does not read before changing hardware.

Next step

Send us the model and firmware from the label, the settings backup, the PLC data layout with its GSD or EDS file, the wiring and mounting drawings, and a photo of the worst code beside a ruler at the working distance. Add line speed and presentations per second; for a KEYENCE SR reader, add the exact variant. We check exact Hikrobot variants against their datasheets and your list, and tell you what is settled on paper and what your parts must prove.

Related

• Choosing an industrial code reader: fixed or handheld, label or direct part mark
• Working out the camera resolution a code needs
• Why a code does not read, and what to check first
• Setting exposure for a moving part, and deciding whether to pulse the light

How this material was prepared

Prepared by MVisionPro from the stated sources and the MVisionPro calculation engine. A physical test is claimed only when the material says so explicitly. Read the editorial method.

Editorial status: verified. Content updated 2026-09-25.