Omron NX102 CP1E CP1H Migration Guide 2026

Omron NX102 in 2026: The CP1E and CP1H Migration Path for Southeast Asia Machine Builders

A packaging line in Rayong, a sterilizer build in Penang, a bottling skid in Ho Chi Minh City — most of them were programmed on an Omron CP1E or CP1H. As of August 2026 those controllers are no longer in volume production, and the migration question is now landing on the desk of every panel builder who still owns one.

This guide is written for the buyer on the receiving end of that question: the SEA machine builder, the OEM controls engineer, and the small-system integrator who has been told "spec the replacement" without being told what to watch out for. It draws on the publicly announced Omron discontinuance windows, the NX102 catalog data we carry, and what is actually different when you move a CP1 program across to Sysmac Studio.

If you would like a quote on a specific NX102 or CP1E / CP1H part number, send the MPN, quantity, and target delivery window to our inquiry desk. We typically respond with availability, lead time, and MOQ within one business day for SEA time zones. Samples of NX-ID and NX-OD slices can be arranged against qualified panel-build accounts to support a first migration pilot.

Why CP1E and CP1H Are No Longer Routine Buys in 2026

Omron's CP1E family has been on the company's discontinuance schedule since 2018, with staged last-buy deadlines. By mid-2026 the majority of CP1E-N (relay output) and CP1E-ED (transistor output) part numbers have already passed their final-order cutoff. New CP1E units are no longer coming off the production line; what is left in the channel is whatever surplus franchised distributors, panel shops, and decommissioned equipment resellers are willing to release.

CP1H is a slightly different story. It was placed on Omron's "discontinuance in stages" notice later than CP1E, with the last-buy window originally set for September 2025 through March 2027 depending on the specific suffix (X, Y, XA, YA). In practice, that means a CP1H you order in late 2026 may still be sourced from Omron's factory — but only for specific suffixes, and only until the per-SKU cutoff. Buyers who have not pinned down a CP1H part number against the current schedule are finding out the hard way that some variants are already past the line.

The SEA machine-builder market has not been waiting around. As of mid-2026, global factory-automation order books were running ahead of 2025 by roughly 9 to 12 percent on a year-over-year basis per Q2 vendor earnings reported across the major PLC vendors, with SEA specifically absorbing a disproportionate share of the regional capacity-build (food processing, packaging, semiconductor backend test). Component-side pressure is showing up too: power-device lead times were quoted at 52 weeks in August 2026 across some semiconductor franchises, and the broader industrial market is seeing a third 2026 component price hike cycle take effect in late August. That backdrop is what is pushing the NX102 conversation into the open at every SEA panel shop we talk to.

The buyer-side impact is threefold:

New machine builds that spec CP1E have to be redesigned to NX102 or NX1 before the first PO is cut.
Service spares for installed CP1E base units are increasingly thin, with genuine-stock pricing rising. We are seeing CP1E-N14DR-D and CP1E-N40DR-D list price move up by roughly 35 to 60 percent above 2023 levels across SEA surplus channels, with MOQ on genuine units commonly 1 unit against verified test reports.
CP1H orders placed against the wrong suffix can slip into a 26-week allocation rather than a 4-week delivery.

What the NX102 Actually Is, and Where It Fits

The NX102 is Omron's entry-level NX-series controller. In the Sysmac platform hierarchy it sits below the NX1P (compact, all-in-one CPU with onboard I/O) and below the NJ-series (motion-capable mid-range), but above the CP1 family in almost every specification that matters to a modern machine builder.

Two NX102 CPU modules carry the bulk of the demand we see in our catalog, both rated as premium-grade stock:

NX1021220 — CPU with 16 DC inputs / 16 transistor outputs, EtherCAT and EtherNet/IP master, 4 MB program capacity, OPC UA server, basic motion via EtherCAT slaves.
NX1021120 — same architecture with screwless push-in terminal wiring, otherwise functionally identical.

Both ship with the same NX-ECC EtherCAT coupler architecture, both support the NX-series safety add-on (NX-SL3300 / SL3500) and the high-speed analog / temperature slices (NX-AD / DA / TS). The two SKUs are differentiated mainly by wiring method and a small difference in onboard I/O density on certain firmware revisions.

For a CP1E replacement, the practical differences versus the CP1E-N14DR-D you might have specified five years ago are these:

Programming environment moves from CX-Programmer to Sysmac Studio. The IEC 61131-3 languages (ladder, ST, FBD) all carry over, but the project structure and tag database are different.
I/O is no longer onboard screw-terminal only. You will normally build an NX102 station with the CPU plus a power supply (NX-PA / PD) plus one or more digital input / output slices (NX-ID / OD). High-density 32-point slices are now the norm. Typical MOQ on slice orders is one unit with same-week shipment from SEA-region stock.
Network is EtherCAT for device-level and EtherNet/IP for plant-level. Serial RS-485 / RS-232C is supported via NX-CIF slices, not onboard.
Motion control is now an EtherCAT slave exercise, not a pulse-train exercise. The NX102's built-in positioning is suitable for low-axis-count indexing applications; anything beyond four axes with interpolation should look at NJ or NX701.

Where Genuine CP1 Surplus Stock Still Makes Sense

There are still three situations where holding or buying CP1E / CP1H units is the right call in 2026:

First, you have a long-life machine that has been in service for eight or more years, where the panel layout is fixed and the customer has explicitly refused a redesign. Here the only realistic option is to source a small buffer of CP1E-N / CP1H CPUs against the specific suffix you have on the floor. The genuine-stock market is thin and the prices have moved up by a double-digit percentage since 2024, but for a 15-year-life skid in a chemical plant that math works. Genuine units come with a test report; MOQ is typically 1 unit against a verified suffix and serial-number trace.

Second, you are building a non-critical accessory skid where the customer's spec sheet still names CP1E. Sometimes the spec was written in 2021 and never updated. A small one-off build of, say, three units can be cleared with two genuine CP1E CPUs from surplus stock rather than triggering a redesign conversation with a customer who did not ask for one.

Third, you have an active CP1E spare parts commitment. If you signed a 5-year MRO contract in 2023 that names CP1E-N as a covered line, you need a path to source it through 2028. The honest answer for most panel builders is to renegotiate the contract to NX102 now, while there is still time, rather than ride the genuine-stock curve into 2027.

In all three situations the practical question is the same: who actually has the part, and can you prove it is factory-genuine rather than relabeled? That is where a distributor carrying both current NX102 stock and verified CP1 surplus becomes valuable.

What Changes When You Move the Program Across

The technical part of the migration is not as painful as the version-control part. The mapping is:

CX-Programmer POU files become Sysmac Studio POUs. Ladder logic carries over almost line-for-line. Structured Text and Function Blocks carry over with minor syntax adjustments.
I/O addresses change. CP1E used CIO-area bit addressing (0.00, 100.01, etc.). NX102 uses tag-based addressing on slice terminal numbers. This is the single biggest source of bugs in a first migration.
Timers and counters move into Sysmac Studio's IEC 61131-3 task system rather than the CP1's fixed scan model. Programs that depended on a specific scan time now see a slightly different execution order. Most programs do not care; some PID loops do.
Memory areas (DM, HR, AR) become Sysmac Studio global variables with retain attributes. Anything that was implicitly retained in the CP1 needs an explicit retain flag in NX102.
Expansion racks (CP1W-20EDR / CP1W-TS002 etc.) do not carry over. The NX102 uses EtherCAT slices, which are physically smaller, terminate differently, and require a different cabinet layout.

For an SEA machine builder with one or two controls engineers, the realistic migration cost on a CP1E-N program of 200 to 500 rungs is two to four engineering days plus one weekend of commissioning. For a CP1H program with motion and serial comms, plan a week to ten days.

Pricing and Lead Time Reality in 2026

Quoted lead times for NX102 CPUs across major franchised channels in mid-2026 ranged from 6 to 10 weeks for the most common SKUs (NX1021220 and NX1021120), with the longest tail on the NX102-9xxx variants that bundle additional motion capability. SEA-region availability from our SEA-side warehouse is typically 1 to 3 weeks against active stock with MOQ 1 unit; cross-region lead time from the EU or US channel is the longer 6 to 10 week figure.

Genuine-surplus CP1E-N units were quoted at 1 to 3 weeks where available, with pricing approximately 35 to 60 percent above 2023 levels for the most common part numbers. CP1H units still on the active Omron schedule were quoting 4 to 6 weeks for the open-suffix SKUs and 14 to 18 weeks for SKUs approaching the last-buy window.

The NX702 and NJ501 lines that SEA integrators sometimes spec for higher-axis-count machines have been running closer to 14 to 18 weeks, with allocation pressure showing up specifically on NJ501-1340 and NJ501-4320 SKUs.

These lead times are published by franchised distributors and reported through industry channels. They move month to month, but the directional picture is that NX102 is in steady supply while NJ and NX702 are tighter, and CP1E is in genuine-stock territory only.

The SEA-Specific Layer: What SEA Machine Builders Should Plan Around

Three things make the SEA context different from a US or EU factory floor:

First, the installed base skews older. A large fraction of CP1E and CP1H machines in Thailand, Vietnam, Malaysia, Indonesia and the Philippines were commissioned between 2014 and 2020. Many of those machines have only just been paid off. Migrating them now means a CapEx conversation with a customer who is also dealing with rising component lead times in the rest of the panel. SEA industrial PLC spending is on track to grow roughly 8 percent year-on-year through 2027 per multiple regional market reports published in mid-2026, but the mix is shifting toward controllers with documented cybersecurity features rather than entry-level bare-metal PLCs.

Second, panel-shop capacity is concentrated. The top twenty SEA panel shops handle a disproportionate share of regional machine-build work, and several of them have a CP1E template they have re-used forty times. Migrating that template to NX102 is a one-time investment that pays off across dozens of future builds. The incentive to do the work now is high; the budget is often not. OMRON's August 2026 SEMICON Taiwan showcase and the wider push toward integrated AI-vision on Sysmac-platform controllers is a tailwind for that conversation — the CP1E has no path to the new functionality.

Third, the regulatory environment is tightening around cybersecurity and functional safety. Malaysia's DOSH requirements and Vietnam's growing adoption of IEC 61508 SIL principles are pushing machine builders toward controllers with documented security features. The NX102's OPC UA server and built-in EtherCAT safety network (when paired with an NX-SL safety controller) address this in a way CP1E cannot. For new builds going into regulated end-markets, NX102 is increasingly the default not because CP1E is gone but because the compliance paper trail is shorter.

A Practical Decision Path for the Next Build

If you are specifying a new machine in late 2026:

Lowest-cost single-purpose skid, no motion, no remote I/O, no safety, < 40 I/O points total: NX102 with onboard I/O plus a single expansion slice. NX1021220 base + NX-OD3258 + NX-ID3344 will handle most cases. Typical MOQ 1 unit with 1 to 3 week SEA-side availability against active stock.
Machine with 4 axes of simple motion and EtherCAT remote I/O: NX102 plus external Sigma-X or 1S-series servo drives on EtherCAT. The NX102's built-in positioning handles this without the NJ upgrade.
Machine with 6+ axes, complex motion, or integration with a higher-level NJ/NX supervisory controller: NJ501-1340 or NX102-9xxx, but expect 14+ week lead times and plan accordingly.
Service spare for an installed CP1E base: verify the exact suffix, then quote against genuine-surplus stock with a clear test report. Plan to migrate the panel at the next major maintenance event rather than the next minor one.
Service spare for an installed CP1H base: confirm the suffix against Omron's per-SKU last-buy schedule. Some suffixes are still open through 2027; others are already closed. The answer is in the specific part number, not the family name.

What We Carry for This Migration

Our catalog currently stocks the two NX102 CPUs that cover the majority of entry-level machine builds — NX1021220 and NX1021120 — alongside the NX-ECC EtherCAT couplers, NX-ID / NX-OD digital slices, and the NX-PA / PD power supplies they require. We also hold verified CP1E and CP1H units from factory-genuine surplus channels, with test reports available on request. SEA-region availability is generally 1 to 3 weeks against active stock; cross-region lead time is 6 to 10 weeks. MOQ is 1 unit on most SKUs, with tiered pricing kicking in at 5 and 25 units.

For machine builders planning a Q4 2026 or Q1 2027 commissioning window, the right time to lock in NX102 stock is now, while the quoted 6 to 10 week lead times still give you a reasonable buffer against cabinet-build slip. Send your MPN list, quantity, and target delivery date to our inquiry desk for a same-day quote; samples of NX-ID and NX-OD slices can be arranged against qualified panel-build accounts to support a first migration pilot.

The CP1E will keep running machines in SEA for another decade. But as of 2026, it is no longer a controller you can freely spec on a new build. The NX102 is the realistic replacement, the Sysmac Studio transition is a one-time cost rather than a recurring one, and the lead times are workable if you plan ahead.