Problem Overview
The Siemens Sinumerik 810D control with MMC100.2 HMI boots, services load, and then the operator screen displays "MMC100.2 version 05.03.31 Waiting for communicationpartners". The boot is interrupted before ShopMill can render, leaving the panel effectively black. Alarm 120202 "Wait for connection to NC/PLC" is raised in parallel. The fault usually appears immediately after the on-board lithium cells (NCK battery and CMOS battery on the MMC/PCU) are replaced on an aged machine such as a Hardinge VMC, Fadal 4525, or similar 3-axis vertical machining center equipped with the 810D powerline or 810D CCU3 architecture.
The condition is recoverable in most cases. The HMI process is alive and able to render the Siemens Services, Diagnostics, and Start-up areas if the operator repeatedly cycles the Machine area key, which proves the underlying link from the CCU to the NCK/PLC is at least partially intact. The boot halt is a startup-ordering problem, not a fatal link failure.
Observed Symptoms
| # | Symptom | Stage |
|---|---|---|
| 1 | NCK and CMOS batteries replaced; control re-powered | Maintenance |
| 2 | Alarm 120202 "Wait for connection to NC/PLC" logged at boot | Boot / Services |
| 3 | Popup: "Warning: Application 'dino' didn't post initcomplete! Press <enter> to continue" | Services load |
| 4 | ShopMill fails to load; screen black | HMI load |
| 5 | Banner: "MMC100.2 version 05.03.31 Waiting for communicationpartners" | HMI idle |
| 6 | 7-segment display on CCU shows 8 with no further state change |
NCU |
| 7 | Pressing Machine key repeatedly toggles into Service / Diagnostics | Workaround |
| 8 | On-board archives (.ARC) for NC and PLC are present on the drive | Recovery asset |
| 9 | Original MHI environment is bootable, password unknown | Recovery asset |
| 10 | NC/PLC general reset performed; PS LED does not begin to blink at the S4 = 2-3-2 step | Procedure failure |
8 on the MMC is the Sinumerik default "HMI ready, no NC/PLC link established" state. It is not a hard NCK fault on its own - it only confirms that the HMI is alive and that the named communication partner (NCU, PLC) has not signalled ready. The root cause is almost always in the NCU/PLC startup handshake, not in the PCU/MMC.
Root Cause Analysis
Alarm 120202 in Sinumerik 810D/840D is raised by the HMI when the HMI-MMC (or the modern PCU/TCU equivalent) cannot complete the OPC-style or NCVAR.INI-driven handshake with the NCK and PLC within the configured timeout. The default timeout window on MMC100.2 software version 05.03.31 is short (typically 30-60 s after the last Services-area phase). Three root-cause families dominate field reports:
- NCK / PLC data loss after battery replacement. With dead batteries, the persistent NCK machine data, PLC retentive flags, and the licensed option set (CF card slot on the NCU) may not survive a power-down. The NCK boots into an "empty" machine configuration and asks the HMI for commissioning data that no longer exists.
-
HMI application
dinofailed init-complete handshake. dino is the display initialization component of the MMC HMI. A failure here means the HMI never marks the link ready, so the NCK times out trying to confirm the partner is present. - MMC vs. NCU software / version mismatch. The unit may be a MMC102 (despite the silk-screen label) running a software set that does not match the boot sector the NCK expects. Cross-flashed or partial image restoration commonly produces this exact symptom.
The most common contributor is a partial or failed general reset (NC + PLC clear) on the NCU. A clean reset is the prerequisite for the HMI to re-initialise the link and for the on-board archives to be reloaded successfully.
Pre-Procedure Verification
Before clearing the NC or PLC, confirm the following so the reset does not lock the machine out of recoverable data:
-
Archive the on-board images. From Start-up → Series Commissioning → Create Service Commissioning Archive, copy the active NC and PLC archives to an external PG/PC over Ethernet or to a PC card. Files of interest:
*.ARCin/user/sinumerik/hmi/data/and PLC project*.dpin the PLC area. -
Image the MMC drive. If the MMC is a PCU 50 / PCU 50.3 or an MMC100.2 with a CompactFlash or HDD, connect a USB or Ethernet and create a sector image. Tools:
WinSCPover the documented MMC IP, or a Linuxddimage from a bootable MHI rescue stick. -
Verify drive integrity. A
chkdsk /fon every MMC partition must return clean. Bad blocks on the partition holdingmmc2.exeor the registry will mimic a communication-partner fault even when the NCU is fine. - Identify the NCU variant. Open the cabinet and read the NCU label. 6FC5410-0AY03-0AA0 is a 810D CCU3. 6FC5410-0AY01-0AA0 is an earlier 810D. The switch positions below are valid for both, but the LED colour mapping (PS / PF / PR) is identical.
- Identify the HMI module. Confirm whether the unit is MMC100.2 (CPU 586, 32 MB RAM, 4 MB Flash) or MMC102 (StrongARM, 64 MB RAM, 16 MB Flash). The image and the licence are tied to the module, not just to the disk.
NC and PLC General Reset (Reference Procedure)
The Sinumerik 810D / 840D NCU uses a 4-position rotary switch S4 and a 2-position mode switch S3 on the front of the NCU. The PS (Power Supply) and PR (PLC Run) LEDs above the 7-segment display report state. Use the following reference sequence for a full NC + PLC clear.
PLC General Reset
- Set S3 on the NCU to position 1.
- Power on the control. Wait for the 7-segment to show a stable number.
- When the control is fully powered up, press the Reset pushbutton on the NCU front.
- Wait for the NCU display to show
6. - Set S3 to 0 and press Reset.
- Wait for the NCU display to show
6again. - Turn S4 to position 2 — PS LED is ON.
- Turn S4 to position 3 — PS LED is OFF.
- When the PS LED comes on, turn S4 to position 2 then immediately back to position 3.
- PS LED should now be blinking at ~2 Hz. This is the PLC clear-in-progress state.
- Set S4 to position 0 — PR LED should become solid green within a few seconds.
- If the PR LED does not light solid green, the procedure was not performed correctly. Repeat PLC steps 1 through 4 (with the 0 → 1 → 0 transitions) before retrying the S4 sequence.
NC General Reset
- Set S4 to position 2.
- Press Reset on the NCU.
- Wait for the 7-segment to step through
1,2, ...,6and to land on6with the PF (Power Fail / PLC Fault) LED OFF. - Power-cycle the control.
- On reboot, the HMI should now be able to complete the communication handshake and either drop into Start-up mode or boot ShopMill directly.
Recovering the Original MHI Environment
If the active Windows / MHI image will not boot, the MMC 100.2/102 carries a hidden Original MHI Environment partition with a baseline HMI. Booting from it is the fastest path to a working display, but it is password-protected.
| Known OEM password | Source / comment |
|---|---|
SUN--- |
Common default for Hardinge and Mazak MMC builds |
EVE--- |
Common default for some Fadal MMC builds |
111 |
Service / development default on early MMC100.2 |
EVENING |
OEM backdoor on certain MMC102.2 builds |
| OEM-specific | Provided by the machine builder — required for non-default service contracts |
If none of the above work, the password hash is stored in \mhi\system\hmi.ini and can be cleared by editing the registry hive HMI under SOFTWARE\Siemens\SINUMERIK with a regedit offline mount from a Linux rescue environment. A direct edit of mhisetup.ini on the Original MHI partition with a new Password= entry and re-seal is documented for field service on the 810D.
Reloading the On-Board Archives
- After a successful NC + PLC clear, the HMI will boot to either the Start-up area or directly into ShopMill / ShopTurn, depending on configuration.
- From the active HMI, navigate to Start-up → Archive → Read In.
- Select the on-board NC archive (
*.ARC) from/card/user/sinumerik/hmi/data/archive/or from the same path on the internal HDD. - Read in the NC archive, then the PLC archive, in that order. Acknowledge all popups.
- Power-cycle the control. The next boot should reach the operator screen without alarm 120202.
MMC100.2 vs MMC102 - Identify Before You Reflash
| Property | MMC100.2 | MMC102 |
|---|---|---|
| CPU | Intel 386EX / 586 | Intel StrongARM SA-1100 |
| RAM | 32 MB | 64 MB |
| Flash | 4 MB | 16 MB |
| Typical HMI version | 05.03.x | 06.04.x and later |
| Max ShopMill version | 5.x | 6.x / 7.x |
| Boot message seen in fault | "MMC100.2 version 05.03.31" | "MMC102 version 06.xx.xx" |
mmc2.ini.
Verification Checklist
Run the following checks after the procedure to confirm the fix is complete and stable:
- Alarm 120202 is no longer present in the alarm log at next boot. Verify under Diagnostics → Alarm Log.
- ShopMill loads fully without the "dino didn't post initcomplete" warning.
- Jog the spindle in manual mode at low RPM. NCK → HMI feedback must show the actual RPM without timeouts.
- Run a dry-run NC program (e.g.
G91 G00 X1.0) and confirm the active block display updates in real time. - Power-cycle the control three consecutive times. Each boot must reach the operator screen without the "Waiting for communicationpartners" banner.
- Replace the NCK and CMOS batteries with fresh cells (e.g. Panasonic BR-2032 on the MMC, Tadiran TL-5104 on the NCU) and record the install date on the cabinet label.
Troubleshooting Matrix
| Observed | Likely cause | Next step |
|---|---|---|
7-segment stuck on 8, MMC shows "Waiting for communicationpartners" |
NC/PLC handshake failed at boot | Full NC + PLC general reset per sequence above |
| PS LED never blinks during PLC clear | Reset button not held long enough, or 24 V missing on X121 | Hold Reset 2 s, verify 24 V at NCU terminals |
| "dino didn't post initcomplete" warning | HMI registry corruption on the MMC disk | Re-image the MMC from a known-good backup or re-install MHI |
| Banner says MMC102, silk-screen says MMC100.2 | Module swap, mismatched image | Match image to the silk-screen, regenerate the licence pairing |
| Original MHI boots but password unknown | OEM default not in your possession | Use OEM backdoor table above, or edit hmi.ini offline |
| NCK boots, PLC does not | PLC project corrupt in retentive memory | PLC general reset, then re-load the .dp archive |
| Alarm returns within minutes of successful boot | Battery is again dead after install - wrong cell | Verify battery is TL-5104 / BR-2032, not a non-rechargeable substitute |
| Alarm returns after every power-down | NCK / CMOS battery not making contact | Clean contacts, reseat, verify polarity, check holder spring tension |
Field-Proven Cautions
- Do not pull the NCK or CMOS battery while the control is energised. Loss of NCK memory mid-write can corrupt the licence on the CF card slot and require a Siemens re-licence procedure.
- Do not skip the S3 = 1 → S3 = 0 transition at the start of the PLC clear. Without it, the PS LED sequence will not behave as documented and the clear will not register.
- Do not power off the control during the "dino" init-complete phase. Power-loss here is the most common cause of needing to re-image the MMC entirely.
- The "Machine" key cycling workaround is for diagnostics only. It does not clear the alarm and the control will not run a part program until the HMI-NCK link is fully established.
- If the machine has a Hardinge, Fadal, or any third-party PLC project, the OEM-specific ladder logic must be reloaded from the OEM archive after a PLC general reset. A standard Siemens archive will not contain the custom PLC blocks.
What does Siemens 810D alarm 120202 actually mean?
Alarm 120202 is "Wait for connection to NC/PLC". The MMC100.2 HMI raises it when the OPC/MPI handshake to the NCK and PLC does not complete within the boot timeout, typically 30 to 60 seconds. It is a link-state alarm, not a hardware fault alarm on its own.
Why does the HMI show "MMC100.2 Waiting for communicationpartners" after a battery change?
Replacing the NCK and CMOS batteries can wipe the persistent NCK machine data and the PLC retentive image. The NCK then boots into an unconfigured state, and the HMI's dino display initialiser cannot complete the init-complete handshake, leaving the HMI stuck on the "Waiting for communicationpartners" banner.
How do I do a Siemens 810D NC and PLC general reset?
Use the S3 and S4 rotary switches on the NCU front panel. PLC clear: S3 = 1, power on, press Reset, wait for 7-segment to show 6, S3 = 0, Reset, then cycle S4 from 2 → 3 → 2 → 3 until the PS LED blinks. NC clear: S4 = 2, press Reset, wait for 7-segment to land on 6 with PF LED off, then power-cycle.
The PS LED does not blink during the PLC clear - what is wrong?
The Reset pushbutton is being released before the firmware has registered the state change, or the 24 V supply to NCU connector X121 is missing. Hold the Reset button for at least 2 seconds at every step, confirm 24 V at X121, and check the emergency-stop loop on X111 is closed before retrying.
How do I unlock the Original MHI environment on a Hardinge or Fadal 810D?
Try the OEM default passwords first: SUN--- (Hardinge / Mazak), EVE--- (Fadal), 111 (early MMC100.2), or EVENING (certain MMC102.2 builds). If none match, edit \mhi\system\hmi.ini on the Original MHI partition offline and replace the password hash, then reboot.
My silk-screen says MMC100.2 but the boot banner says MMC102 - is that the problem?
Yes, almost certainly. The HMI image and the licence are paired to the module variant. A swapped module must be paired with the matching image, or the communication handshake to the NCK will fail. Reinstall the image that matches the silk-screen and regenerate the licence pairing in mmc2.ini.
Do I need a Siemens re-licence after a battery swap?
Usually no, because the licence is stored on a CompactFlash card in the NCU slot, not in battery-backed SRAM. However, if the CF card is removed or the NCU is re-initialised without it, the licence will be missing and the control will not boot past the early handshake stage. Keep the CF card in its slot whenever power is applied.