SINUMERIK 840D Channel Setup for Dual-Toolpost VTL Retrofit

David Krause16 min read
Motion ControlSiemensTechnical Reference
Licensed PE Working through this on a live machine? A Maine-licensed engineer can take it from here — included with IMD hardware, by the hour for everything else. Book an engineer

1. Retrofit Context and Requirements

A vertical turning lathe (VTL) with two independent tool posts and four NC axes is a common retrofit candidate when the original CNC becomes obsolete or unsupportable. Each tool post on a dual-turret VTL is conventionally operated as a quasi-independent machine: an NC alarm, program stop, or operator action on one post must not interrupt production on the other post. This independence requirement drives the controller architecture decision more than any other factor.

The two viable Siemens controller families for this application are:

  • SINUMERIK 840D sl – a single NCU hosting multiple channels and mode groups, paired with a SINAMICS S120 drive line-up.
  • Two SINUMERIK 802D sl controllers – each handling one tool post, each with its own drive cluster.

Both architectures satisfy the independence requirement. They differ in spares strategy, cabinet footprint, drive topology complexity, and HMI ergonomics. The remainder of this document specifies the 840D sl multi-channel architecture in detail, contrasts it with the dual-802D alternative, and provides commissioning and verification steps an integrator can apply directly to a 4-axis VTL retrofit.

2. Architecture Options at a Glance

Architectural decision: The choice between a single 840D and two 802D units is governed by cabinet space, common-spares policy, and the ability to share a drive cluster. On a 4-axis VTL with a common spindle motor contactor panel and shared hydraulics, a single 840D sl with two channels typically yields a smaller cabinet, lower panel cost, and a single PLC (PLC Basic / PLC 317 in the NCU) – but it requires correct channel and mode group configuration to preserve the independence that the legacy twin-NC architecture provided.

2.1 Option A – SINUMERIK 840D sl with two channels

A single SINUMERIK 840D sl NCU (e.g. NCU 720.3 or 730.3) hosts two NCK channels, each acting as a logically separate NC. Each channel is bound to its own mode group, MCP, and HMI focus. The SINAMICS S120 drive line-up is shared via DRIVE-CLiQ; axes are assigned to channels through machine data.

2.2 Option B – Two SINUMERIK 802D sl controllers

Each tool post receives its own 802D sl controller and its own drive cluster (typically SINAMICS V60 or S120 Combi). The two controllers are completely independent; cross-post coordination is implemented over discrete I/O or a small PROFINET link. Cabinet size and panel wiring roughly double relative to Option A.

3. Channel and Mode Group Fundamentals

Three Siemens-specific terms govern multi-toolpost operation on an 840D sl. Their distinction is the most common source of configuration errors on retrofits.

3.1 Channel (NCK channel)

A channel is an independent program execution context inside the NCK. It owns its own:

  • Part program counter and interpreter state
  • Active G-code group state (modal functions)
  • Tool table activation
  • Channel-specific machine data ($MC_…)
  • Channel-specific setting data ($SC_…)

A reset, M30, or NC alarm in one channel does not affect the program counter or modal state of another channel unless a cross-channel write ($NC…-variables to a different channel) is in use. Channel assignment of axes is set in:

; AXCONF_MACHAX_NAME_TAB / AXCONF_MACHAX_CHANAX_NAME_TAB
MD10000 $MN_AXCONF_MACHAX_NAME_TAB[0]  = "X1"
MD10000 $MN_AXCONF_MACHAX_NAME_TAB[1]  = "Z1"
MD10000 $MN_AXCONF_MACHAX_NAME_TAB[2]  = "SP1"
MD10000 $MN_AXCONF_MACHAX_NAME_TAB[3]  = "X2"
MD10000 $MN_AXCONF_MACHAX_NAME_TAB[4]  = "Z2"
MD10010 $MN_AXCONF_MACHAX_CHANAX_NAME_TAB[0,0] = "X1"
MD10010 $MN_AXCONF_MACHAX_CHANAX_NAME_TAB[0,1] = "Z1"
MD10010 $MN_AXCONF_MACHAX_CHANAX_NAME_TAB[0,2] = "SP1"
MD10010 $MN_AXCONF_MACHAX_CHANAX_NAME_TAB[1,0] = "X2"
MD10010 $MN_AXCONF_MACHAX_CHANAX_NAME_TAB[1,1] = "Z2"

Channel-to-axis binding is one of the first items verified at commissioning: any axis declared in MD10000 but not referenced in MD10010 is visible to no channel, and any axis name referenced in MD10010 without a matching MD10000 entry raises alarm 20080 "Axis name %1 not defined".

3.2 Mode group

A mode group is the unit to which an operating mode (AUTO, JOG, MDI, REPOS, REF) is bound. Two channels assigned to different mode groups can execute in different modes simultaneously: e.g. channel 1 running AUTO while channel 2 is being jogged for setup. Two channels assigned to the same mode group share a mode – they cannot be in AUTO and JOG at the same time.

For a true dual-toolpost retrofit where each post is operated by its own operator, each channel should be in its own mode group. The number of mode groups is set by:

MD12711 $MN_MM_MODE_GROUP_NAME_TAB[0] = "TP1"
MD12711 $MN_MM_MODE_GROUP_NAME_TAB[1] = "TP2"
MD12985 $MN_NUM_MODE_GROUP_NAMES      = 2

Channel-to-mode-group assignment is the next binding step:

MD10060 $MC_CHAN_NAME[0]                  = "CHAN_TOOLPOST1"
MD10060 $MC_CHAN_NAME[1]                  = "CHAN_TOOLPOST2"
MD12970 $MN_MM_NUM_CHAN_NAMES             = 2
MD12701 $MN_MM_CHAN_NAME_TAB[0]           = "CHAN_TOOLPOST1"
MD12701 $MN_MM_CHAN_NAME_TAB[1]           = "CHAN_TOOLPOST2"
; Channel-assigned mode group index:
MD12711 (channel-side override) - via $MC_CHAN_MODE_GROUP

Verify the binding after each NCK power-on by reading /Channel/Configuration/name in the HMI diagnostic area or by issuing $AC_MODE_GROUPNO from the part program.

3.3 Cross-channel synchronization

When both posts must remain coordinated (e.g. synchronized turning of a long workpiece), channels exchange wait markers and synchronized actions. Wait markers are set and tested from the part program:

; Channel 1 program (Toolpost 1)
N20  SETM(1,2)         ; set marker 1 in channel 2
N30  WAITM(1,2)        ; wait for marker 1 in channel 2
N40  G1 X100 F0.2

; Channel 2 program (Toolpost 2)
N20  G1 X100 F0.2
N30  SETM(1,2)         ; release channel 1
N40  WAITM(2,2)

Synchronized actions (IDS = … DO …) fire on real-time conditions across channels when higher-frequency coupling is required; they are not strictly necessary for a simple independent-post VTL but should be reserved for synchronized turning or transfer operations.

4. SINAMICS S120 Drive Line-Up and Axis Sharing

The drive cluster on a 4-axis VTL retrofit typically consists of:

  • One Line Module (Active Line Module or Smart Line Module) sized for the sum of motor ratings.
  • Two Motor Modules for the two tool-post feed axes (X1/Z1, X2/Z2) – each motor module is a double-axis module where Z and X share a DC bus on the same physical unit.
  • One Motor Module (or two, if each post has its own spindle) for spindle drives.
  • One or more NX10 / NX15 modules for distributed I/O and additional axes if cabinet separation is required.

All drive objects are exposed on DRIVE-CLiQ as Servo DO types and are visible to the NCU as logical axes. Channel assignment is purely a function of MD10010; no drive-side change is required when moving an axis between channels.

Failure isolation: A drive-side fault (overcurrent, encoder error, DRIVE-CLiQ topology violation) propagates to every channel that references the affected axis. To preserve independence on a shared drive line-up, separate the tool-post axes across at least two Motor Modules so that a single Motor Module fault cannot stop both posts.

For a typical VTL configuration with two independent feed groups, the drive-side assignment is:

Drive Object Motor Module Slot Logical Axis Channel
Servo DO #1 Slot 2 / Port 1 X1 Channel 1 (Toolpost 1)
Servo DO #2 Slot 2 / Port 2 Z1 Channel 1 (Toolpost 1)
Servo DO #3 Slot 3 / Port 1 X2 Channel 2 (Toolpost 2)
Servo DO #4 Slot 3 / Port 2 Z2 Channel 2 (Toolpost 2)
Servo DO #5 Slot 4 / Port 1 SP1 (main spindle) Channel 1 or shared

5. MCP and HMI Distribution

5.1 MCP assignment

Each mode group requires its own MCP focus. On an 840D sl retrofit with two operator stations – one per post – use two MCP 483C IE panels (or MCP 310C for compact pendants). Each MCP is connected to PROFINET and bound to a mode group:

; In PLC (PLC 317 inside the NCU) – assign MCP PN device to a mode group
; Use the MCP/MG coupling FB or HMI sl configuration:
; HMI sl > System > MCP/MG coupling:
;   MCP1 (192.168.1.10) -> Mode Group 1
;   MCP2 (192.168.1.11) -> Mode Group 2

If only one MCP exists, the operator must select the active mode group via the MCP channel-select key. For a true independent-post retrofit, two physical MCPs are recommended.

5.2 HMI assignment

SINUMERIK Operate runs on a PCU 50.5 (or PC-based). Two TCUs (Thin Client Units) may be connected to one PCU. Each TCU is bound to a mode group through HMI sl configuration:

; HMI sl configuration file (or via Commissioning > HMI > TCU assignment)
; tcu1 -> mg1.tp1
; tcu2 -> mg2.tp2

Once bound, each operator station presents the program, tool table, and alarm list of only its assigned channel. Cross-channel visibility is read-only.

6. Machine Data Reference (Excerpt)

The following MD excerpt is sufficient to bring up a 2-channel, 2-mode-group, 4-axis configuration. All MDs require an NCK reset or, where indicated, an NCK POWER ON.

MD Symbol Value Activation Purpose
MD10000 AXCONF_MACHAX_NAME_TAB X1, Z1, SP1, X2, Z2 Power On Axis name list visible to NCK
MD10010 AXCONF_MACHAX_CHANAX_NAME_TAB (channel,channel-axis)-pairs Power On Bind logical axes to channels
MD12701 MM_CHAN_NAME_TAB CHAN_TOOLPOST1, CHAN_TOOLPOST2 Power On Channel names
MD12970 MM_NUM_CHAN_NAMES 2 Power On Number of channels
MD12711 MM_MODE_GROUP_NAME_TAB TP1, TP2 Power On Mode group names
MD12985 NUM_MODE_GROUP_NAMES 2 Power On Mode group count
MD20150 GCODE_DEFAULT_VALUES[7] 2 Reset Default G group on reset
MD27880 PART_COUNTER 1 (each channel) Reset Channel-local part counter
MD30350 SIMU_AX_VDI_OUTPUT 0 Power On Disable sim-axis VDI handshakes

For the PLC side, MCP/MG coupling is set in the HMI configuration and assigned through the PLC interface DB19 (operator-panel interface) for each mode group.

7. Failure Isolation Analysis

The independence requirement maps cleanly onto the following fault matrix:

Fault Source Channel 1 (TP1) Impact Channel 2 (TP2) Impact
Channel 1 NC alarm (program syntax, modal error) Stops Continues
Channel 2 NC alarm Continues Stops
X1 axis encoder fault (drive-side) Stops Continues (only X1 lost)
Z1 motor module fault Stops Continues (TP2 axes on separate module)
Motor Module shared between X1/X2 (anti-pattern) Stops Stops
DRIVE-CLiQ ring break Stops Stops (whole ring down)
PL / PLC stop (latch open, emergency stop) Stops Stops
NCK software watchdog Stops Stops (whole NCK halts)
PROFINET MCP1 loss (one physical MCP unplugged) TP1 inoperable from MCP1 Continues
HMI sl crash on TCU1 TP1 visibility lost Continues (TCU2 independent)
Architectural rule: Any single failure that takes down both posts defeats the goal of the retrofit. Therefore, place each post's axes on physically separate Motor Modules, separate encoder chains, and (where practical) separate PROFINET segments for I/O. A single DRIVE-CLiQ ring is acceptable when its failure is rare and the post pair can be restarted together; a dual-ring topology is preferred when production cannot tolerate simultaneous post loss.

8. Channel Architecture Topology (SVG)

SINUMERIK 840D sl — 2 Channels / 2 Mode Groups / 4 Axes Mode Group 1 (TP1) Channel 1 — X1, Z1, SP1 MCP1 | TCU1 (Operator TP1) Mode Group 2 (TP2) Channel 2 — X2, Z2 MCP2 | TCU2 (Operator TP2) NCU 720.3 / 730.3 PLC 317 | HMI sl SINAMICS S120 Line-Up Line Module | MM (X1/Z1) | MM (X2/Z2) | MM (SP1) | NX10

9. Commissioning Procedure

  1. Pre-power checklist. Verify cabinet PE bonding, DRIVE-CLiQ wiring (use only Siemens-approved cables, max 70 m per segment), and PROFINET device discovery. Confirm the NCU and PCU firmware match the SINUMERIK 840D sl release notes for the chosen software version (e.g. 4.95 SPx or 6.x as deployed in your region).
  2. NCK first power-up with default MD. Clear all NC and drive MD to factory defaults. Bring the NCK to RUN state. Resolve any topology errors reported by alarm 25000 / 25201 before proceeding.
  3. Define channels and mode groups. Apply the MD excerpt in Section 6. Issue NCK Power On. Verify with $AC_CHANNO from each operator panel that the channel focus follows the MCP.
  4. Configure axes to channels. Populate MD10000 and MD10010. Power-on. Use HMI Commissioning > Axes to confirm each axis displays under the correct channel.
  5. Assign MCPs and TCUs. Configure MCP/MG coupling in HMI sl. Test mode-group switching with both MCPs. Verify that pressing the channel-select key on MCP1 swaps the visible channel without affecting channel 2 state.
  6. Drive commissioning per axis. For each axis, run the SINAMICS automatic controller optimization with the mechanical axis coupled. Record the result files into the NCU CompactFlash card under /siemens/sinumerik/….
  7. Independence test 1 – NC alarm isolation. From channel 1, deliberately trigger an NC alarm (e.g. program a non-existent tool). Confirm channel 2 keeps running and operator can jog X2/Z2.
  8. Independence test 2 – drive fault isolation. With the machine in safe state, simulate an encoder fault on X1 (e.g. unplug encoder cable). Confirm only channel 1 halts; channel 2 continues and reports no alarm.
  9. Cross-channel synchronization test. Execute the SETM/WAITM pair from Section 3.3 in both channels and verify timing on the oscilloscope trace of X1/Z1 vs X2/Z2.
  10. Operator acceptance test. Run a complete part on each post independently. Record cycle time, alarm log, and tool-change sequence. Sign off against the customer acceptance protocol.

10. Comparison: 840D Multi-Channel vs Dual 802D

Criterion 840D sl, 2 channels, 1 drive line-up Two 802D sl, separate drives
Axes supported per control Up to 30 (NCU 730.3) – 4 is well below limit Up to 5–6 per 802D – 2 axes + spindle per control fits
Channels 2 logical NCs in one NCK 1 NCK per 802D
Drive cluster One SINAMICS S120 line-up, shared DRIVE-CLiQ ring Two SINAMICS V60 / S120 Combi clusters, independent
Cabinet footprint Smaller (one NCU, one line-up) Larger (two NCKs, two line-ups)
Spares commonality with other 840D machines Excellent if customer standardizes on 840D sl 802D spares are separate
Cross-post synchronization Native: wait markers, synchronized actions Discrete I/O or small PN link – extra design
Single point of failure (NCK) NCK software fault halts both posts Each post fully isolated
PLC programming One PLC 317 with channel-aware segments Two independent PLC projects
HMI Two TCUs against one PCU; clean focus switching Two independent HMIs, no shared focus
Commissioning effort Higher initial (mode group + channel design) Lower initial – two parallel 802D projects
Reuse of legacy MCPs / handwheels Requires MCP 483C IE or MCP 310C 802D MCP directly, simpler wiring
Field service training Higher (multi-channel diagnostics) Lower (each 802D is a stand-alone product)

11. Recommended Architecture

For a VTL retrofit where:

  • Spare-part commonality across the fleet is a stated goal,
  • The cabinet footprint is constrained,
  • The customer service team is already trained on 840D sl, and
  • The two tool posts occasionally need to be coordinated (e.g. on long shafts or when transferring a workpiece),

the single 840D sl with two channels and two mode groups is the recommended architecture. It satisfies the independence requirement through the channel/mode-group separation, supports synchronized turning through native SETM/WAITM and synchronized actions, and consolidates the spare-part inventory to the 840D sl family.

The dual-802D architecture is preferred when:

  • The two posts are operated by separate shops with separate maintenance contracts,
  • Maximum electrical isolation is required (one post may be operated with the other fully de-energized for extended periods),
  • Cabinet space is not a constraint, and
  • The customer already stocks 802D spares.
Field note: The most common retrofit mistake is binding both channels to a single mode group. This makes the two posts mutually exclusive in operating mode – an operator cannot jog one post while the other is running AUTO. Confirm mode-group separation in the HMI Commissioning > Mode Group screen before the first acceptance run.

12. Spare-Parts and Lifecycle Considerations

Both controller families are subject to Siemens product lifecycle announcements. Before retrofit commitment, confirm the active product status on the official Siemens support portal:

For 802D sl, verify that the deployed firmware version is still within the active service window. Older 802D (non-sl) variants are in the legacy phase and should be replaced rather than retrofitted for production machines.

13. Practical Commissioning Checklist

# Item Pass / Fail Notes
1 NCK reports no topology alarms after first Power On
2 Two channels visible in HMI Commissioning > Channels
3 Two mode groups visible in HMI Commissioning > Mode Groups
4 X1, Z1, SP1 bound to channel 1; X2, Z2 bound to channel 2
5 MCP1 → mode group 1, MCP2 → mode group 2
6 TCU1 → mode group 1, TCU2 → mode group 2
7 Channel 1 NC alarm does not stop channel 2
8 Channel 2 NC alarm does not stop channel 1
9 X1 encoder fault does not stop channel 2
10 SETM/WAITM synchronization works under load
11 Emergency stop drops all drives safely, no single-axis latch fail
12 Operator acceptance run on both posts signed off

14. Troubleshooting Matrix

Symptom Likely Root Cause Resolution
Alarm 20080 "Axis name not defined" MD10000 missing entry Add axis to MD10000, Power On
Axis visible in both channels MD10010 bound to wrong channel Re-bind to correct channel, Power On
MCP key press has no effect on second post Both channels in same mode group Add second mode group, reassign channels
Channel 1 RESET affects channel 2 program counter Cross-channel write via $NC_… variables Audit program for shared state, refactor to local
Channel 1 alarm stops channel 2 Drive-side fault shared between axes Re-assign axes to separate Motor Modules
SETM/WAITM deadlocks on first cycle Mismatch in marker number or direction Verify SETM(m, ch) sets marker m in channel ch
DRIVE-CLiQ alarm 25000 at startup Topology mismatch after NCK replacement Accept topology via SINAMICS commissioning tool
HMI shows only one channel after Power On MD12970 mismatch with configured channels Increase MD12970, Power On
MCP "Channel advance" key swaps to wrong mode group MCP/MG coupling incorrectly configured Re-bind MCP PN device in HMI configuration

15. Quick Parameter Reference

; === Channel definition (NCK global) ===
MD12970 $MN_MM_NUM_CHAN_NAMES          = 2
MD12701 $MN_MM_CHAN_NAME_TAB[0]        = "CHAN_TOOLPOST1"
MD12701 $MN_MM_CHAN_NAME_TAB[1]        = "CHAN_TOOLPOST2"

; === Mode group definition (NCK global) ===
MD12985 $MN_NUM_MODE_GROUP_NAMES       = 2
MD12711 $MN_MM_MODE_GROUP_NAME_TAB[0]  = "TP1"
MD12711 $MN_MM_MODE_GROUP_NAME_TAB[1]  = "TP2"

; === Axis catalog (NCK global) ===
MD10000 $MN_AXCONF_MACHAX_NAME_TAB[0]  = "X1"
MD10000 $MN_AXCONF_MACHAX_NAME_TAB[1]  = "Z1"
MD10000 $MN_AXCONF_MACHAX_NAME_TAB[2]  = "SP1"
MD10000 $MN_AXCONF_MACHAX_NAME_TAB[3]  = "X2"
MD10000 $MN_AXCONF_MACHAX_NAME_TAB[4]  = "Z2"

; === Axis-to-channel binding (channel axis index) ===
MD10010 $MN_AXCONF_MACHAX_CHANAX_NAME_TAB[0,0] = "X1"
MD10010 $MN_AXCONF_MACHAX_CHANAX_NAME_TAB[0,1] = "Z1"
MD10010 $MN_AXCONF_MACHAX_CHANAX_NAME_TAB[0,2] = "SP1"
MD10010 $MN_AXCONF_MACHAX_CHANAX_NAME_TAB[1,0] = "X2"
MD10010 $MN_AXCONF_MACHAX_CHANAX_NAME_TAB[1,1] = "Z2"

; === Channel-to-mode-group binding (HMI Commissioning screen) ===
; Channel 1 -> Mode Group 1 (TP1)
; Channel 2 -> Mode Group 2 (TP2)

; === MCP/MG coupling (HMI sl configuration) ===
; MCP1 (192.168.1.10) -> Mode Group 1
; MCP2 (192.168.1.11) -> Mode Group 2

Can a single SINUMERIK 840D control two independent tool posts so that an alarm on one does not stop the other?

Yes. Configure two NCK channels and assign each to its own mode group. With channel 1 mapped to X1/Z1/SP1 and channel 2 mapped to X2/Z2, an NC alarm, reset, or program stop in one channel does not affect the program counter, modal state, or mode of the other channel. Place each post's axes on separate SINAMICS Motor Modules so a single drive fault cannot stop both posts.

Do two 802D controllers need two separate drive clusters?

Yes, in practice. Each 802D sl binds to its own SINAMICS V60 or S120 Combi drive cluster over a dedicated interface. Sharing a single SINAMICS S120 line-up between two 802D units is not a standard configuration and would require additional commissioning effort and a custom PROFINET topology. For a retrofit where drive consolidation is a goal, the 840D sl multi-channel architecture is the cleaner choice.

How many mode groups do I need on a dual-toolpost VTL?

Use two mode groups. One mode group can host only one operating mode (AUTO, JOG, MDI, REF, REPOS) at a time across all its assigned channels. With a single mode group, you cannot run AUTO on toolpost 1 while the operator jogs toolpost 2 for setup. Two mode groups, one per post, remove that restriction.

How do I synchronize the two posts when needed?

Use wait markers and synchronized actions. From the part program, SETM(m, ch) sets marker m in channel ch, and WAITM(m, ch) waits for that marker. For tighter coupling (synchronized turning, transfer between posts), use synchronized actions (IDS = … DO …) that fire on real-time conditions across channels.

What is the most common commissioning mistake on this retrofit?

Binding both channels to the same mode group. The two posts then share an operating mode and behave like a single NC with two channels — the independence goal is lost. Always verify mode-group separation in HMI Commissioning > Mode Group and confirm via $AC_MODE_GROUPNO at each operator station before the first acceptance run.

Back to blog