SINUMERIK 840D Spare Parts Guide: NCU, PCU, OP Panels and What to Order

The alarm lights up, the axes freeze, and the operator is already at your desk. On a Siemens SINUMERIK 840D machine, that moment usually points at one of a small group of parts: the NCU, the PCU, an operator panel, or the machine control panel. Get the wrong module and you waste a day. Get the right one and the machine is cutting again before the second shift starts.

This guide is for maintenance and procurement people who need SINUMERIK 840D spare parts without the OEM wait and without guessing at MLFB numbers. It covers what each major unit does, what tends to fail, how to read the label so you order a compatible replacement, and where a tested used unit makes more sense than a new one. For the broader rules of buying used industrial stock, see our buyer's guide to used industrial parts.

Typical 840D cabinet layout: NCU, PCU and drive modules sharing one backplane.

What sits inside a SINUMERIK 840D system

SINUMERIK 840D is not one box. It is a control system made of several modules that talk to the drives, the operator, and the PLC. Knowing which block failed is half the job. The other half is matching the exact order number, including version and hardware level.

Drive power is a separate story. Many 840D machines still run on SIMODRIVE 611 modules. Newer cabinets pair the control with SINAMICS. If the NC is healthy but the spindle or axes will not enable, look at the drive line before you condemn the NCU. Our note on what happens when a SIMODRIVE 611 power module dies covers that path.

NCU: the part that actually runs the machine

When people say "the Sinumerik is dead," they often mean the NCU. It holds the NC software, axis configuration and the link to the drive bus. On classic 840D / 840D powerline systems you will see NCU 561, 571, 572 and 573 families under numbers like 6FC5357-0BB…. On 840D sl the hardware moves toward compact NCU units such as the 6FC5210-0DA21-2AA1 style modules.

Common real-world failures are capacitor age on older boards, flash or compact flash corruption, fan failure, and connector damage from repeated rack pulls. Heat in a poorly ventilated cabinet shortens life more than runtime hours alone. Intermittent "NC not ready" after a warm weekend is a classic early warning.

Before you order a replacement NCU, copy the full MLFB from the side label, not from an old packing list. You need the complete string, including the version letter where it is printed. Memory size and option packages are not always interchangeable even when the front of the module looks identical. If the machine archive is healthy, a matching NCU can be restored from backup after the hardware swap. If the only copy of the project lived on the failed unit, recovery gets expensive fast. Keep a current series startup archive off the machine.

In our EU stock we rotate tested NCU units for 840D service work, including boards such as the 6FC5357-0BB35-0AA0 NCU (version F). Every unit is checked before it ships. Send the exact number from your label if you want a match confirmation before you buy.

PCU: when the HMI side dies but the NC might still be alive

A blank operator screen does not always mean the NCU failed. On many 840D cabinets the PCU is a separate industrial PC that hosts the HMI. The NC can still be running in the background while Windows on the PCU has crashed, the disk has failed, or the panel link is gone.

PCU 50 and PCU 50.3 units show up constantly in used inventories. A typical service part is the 6FC5210-0DF21-2AA0 PCU 50. Watch CPU generation, RAM and the exact hardware version. A newer PCU is not automatically a drop-in if the image, panel interface or installed software release does not match the machine.

Practical checks before you condemn the PCU:

• Does the panel get 24 V, and does the backlight try to come on?
• Can you reach the machine over Ethernet with a service laptop while the screen stays dark?
• Are keyboard or MCP commands still recognised?
• Is the compact flash / hard disk making noise, or is the boot stuck in a loop?

If the NC answers and only the HMI is dead, replace or image the PCU path first. Swapping the NCU "just in case" creates a second problem when the original configuration no longer matches.

Siemens NCU nameplate

Operator panels and machine control panels

Operator panels fail in boring, predictable ways: backlight end of life, broken touch layers, cracked fronts from coolant and impact, and ribbon or LVDS cable faults from door flex. For 840D you will meet OP010, OP012, OP015 and related fronts under numbers such as 6FC5203-… and 6FC5610-….

Examples from current catalogue listings include the OP012 front 6FC5203-0AF01-0AA0 and the 840D operator panel 6FC5203-0AC00-0AA2. Layout matters. Key arrangement, language variants and US versus EU keyboards are easy to miss on a rushed order.

The MCP is a different part. If softkeys on the screen work but cycle start, feed override or mode select do nothing, look at the machine control panel and its cable before you rebuild the whole HMI. We stock units such as the MCP 483 6FC5203-0AF22-1AA0 for exactly that case.

For general Siemens panel economics (budget, lead time, when used stock wins), the same logic as our used Siemens HMI panel guide applies: a tested used front often beats an eight-week OEM quote when the machine is down today.

Match screen size, connector and key layout, not only the first digits of the part number.

How to read a Siemens MLFB so you stop ordering the almost-right part

Siemens order numbers look noisy until you split them. Take a typical control unit string such as 6FC5357-0BB35-0AA0:

• 6FC points at SINUMERIK / CNC hardware rather than a drive or PLC family.
• The next digits identify the product group (NCU type, PCU type, panel family).
• The groups after the hyphens lock hardware version, memory, interfaces and design level.
• Version letters on the label (A, B, F and so on) are not decoration. Firmware and hardware revisions ride there.

Rules that save return shipments:

1. Photograph the label on the failed unit before you unplug anything.
2. Write down the full MLFB, version, and any option or memory notes on the sticker.
3. Confirm the machine software release from the HMI or from the last archive.
4. Ask the supplier to confirm form, fit and function against that string, not against a "similar 840D NCU."

Official documentation and manuals live on Siemens Industry Online Support. Use it to check equipment manuals and hardware descriptions when the label is damaged. For encoder-side problems on the same machines, our Heidenhain encoder troubleshooting guide is the better starting point than swapping NC hardware first.

What usually fails first on aging 840D machines

From service work and the parts that move through industrial stock, the failure order is fairly consistent:

1. Cooling and power quality: dead fans, clogged filters, weak 24 V supplies and dirty mains. Fix these or the next NCU dies early too.
2. Operator panels: high touch hours and harsh environments.
3. PCU storage and power supplies: disks, compact flash and aged electrolytic capacitors.
4. NCU hardware: less common than panel and PCU issues, but catastrophic when it happens.
5. Periphery and options: handhelds, I/O modules, NX modules, panel power converters. Example stock in this group includes periphery such as 6FC5111-0CB02-0AA0 and option modules like SINUMERIK NX10 6SL3040-0NC00-0AA0.

A useful diagnostic habit: separate "no display" from "no NC." If drives still show life and emergency logic still reacts, start at the panel and PCU. If the entire NC stack is dark and the drive bus never comes up, start at supply rails and the NCU.

840D still runs a large share of European production machines for a reason: when you can still get parts, the platform stays economical.

Repair, new OEM, or tested used?

There is no single right answer. There is a right answer for your downtime cost.

For discontinued or long-lead SINUMERIK hardware, tested used stock is often the only path that keeps a good machine in service. That is true across Siemens CNC controls generally, which is why we keep a live CNC controls collection and a dedicated Siemens collection rather than mixing random auction lots with no test notes.

A practical swap checklist

1. Record all alarm texts and LED states before powering down.
2. Photograph every connector and the module label.
3. Save or verify the latest archive (NC + PLC + HMI data where applicable).
4. Confirm 24 V and cabinet temperature after the old unit is out.
5. Fit the replacement with correct grounding and seating on the rack or rails.
6. Restore the archive and check software/hardware compatibility messages on boot.
7. Test in a safe mode: emergency chain, axis enable, spindle enable, then a dry run.
8. Only then return to production feed rates.

If you are not sure whether the fault is control-side or drive-side, do not shotgun-replace both. One tested module at a time keeps the diagnosis honest.

Related reading on Samonde

• Siemens SIMODRIVE 611: common failures and solutions
• When a SIMODRIVE 611 power module dies
• SINAMICS fault codes and fixes
• Used Siemens HMI panels and factory budgets
• How to choose used industrial spare parts
• Heidenhain encoder problems and error messages

SINUMERIK 840D machines are still earning their keep on a lot of floors. The platform stays viable when you can identify the failed block quickly, order by the full MLFB, and keep a short list of known-good spares for the parts that stop production. That is the whole game: less guesswork, shorter downtime, same machine.