Troubleshooting Siemens KTP700 HMI Lag in TIA Portal V13 SP1

1. Problem Summary

Engineers using TIA Portal V13 SP1 Upd4 with WinCC Professional V13 SP1 Upd4 and a simulated KTP700 Basic Panel frequently report that HMI variables refresh slowly despite setting the acquisition cycle to 100 ms. Typical symptoms include:

• IO fields displaying stale values, lagging the real PLC state by 1-5 seconds.
• Output fields updating in visible "jumps" rather than smoothly.
• Some screens responding promptly while other screens on the same project appear sluggish.
• Increased lag when multiple HMIs are online simultaneously or when the panel shares a heavily loaded PROFINET segment with distributed I/O.
• Tag values remaining frozen on screen even though the PLC is updating the source DB.

The simulated runtime (WinCC Runtime Professional on a PC) is not the same as the actual device. Performance characteristics observed on the PC simulator — CPU class, available memory, and network stack latency — differ from those seen on the physical KTP700 Basic PN. Always verify the fix on the real panel before signing the change off.

2. Affected System Configuration

The reference configuration for this article is:

If your hardware differs (for example KTP700 Basic DP with PROFIBUS, or an S7-1500 controller), the underlying principles still apply but specific connection counts and update modes may differ.

3. Root Cause Analysis

HMI lag on the KTP700 is typically multifactorial. The following are the dominant contributors, ranked by frequency observed in field reports.

3.1 Acquisition Cycle Set Below Panel Capability

Setting the HMI tag acquisition cycle to 100 ms on a KTP700 Basic Panel is counterproductive. The KTP700 Basic CPU cannot service 100 ms polls on dozens of tags without queuing requests in the internal tag manager. Cross-vendor testing on entry-level HMIs has shown that polling rates above the panel's native tag throughput saturate the OPC/tag manager queue. Apply this rule:

Set the acquisition cycle to 2–10× the maximum acceptable operator-perceived update latency. For a 500 ms perceptual threshold, set the acquisition cycle to 1 s. For operator comfort screens where 1 s update is fine, use 2 s.

3.2 Update Mode Misconfiguration

Each HMI tag in TIA Portal has an Update property with three options:

Tags left on "Cyclic continuous" with a 100 ms cycle multiply the panel CPU load unnecessarily. Switching screen-bound tags to "Cyclic in operation" reduces the active tag count dramatically — typically by 60-80% on a multi-screen project.

3.3 Tag Density per Screen

A single screen displaying more than 50 tags with 100 ms acquisition cycles requires the panel to issue roughly 500 acquisition requests per second. The KTP700 Basic PN's internal tag manager cannot sustain this rate under the S7 protocol without dropping or delaying responses. Reduce visible tag count or split screens into logical sub-screens with navigation buttons.

3.4 PLC Connection Resource Exhaustion

The SIMATIC S7-1200 has a hard limit on S7 communication connections:

Each HMI panel consumes 1 connection (or more if multiple connections are configured for redundancy). Once connections are exhausted, new tags queue at the panel and latency increases nonlinearly. Verify in TIA Portal under PLC → Properties → Communication.

3.5 PROFINET Real-Time and TCP/IP Coexistence

PROFINET RT traffic shares the physical cable with the S7 communication used by HMI tag polling. Heavy I/O update traffic (for example PROFINET IO from ET200SP with 250 µs update time) can saturate the unmanaged switch buffer and cause TCP retransmissions on the HMI connection. Use a managed switch with QoS priority for PROFINET RT to avoid this.

3.6 KTP700 Basic Panel Hardware Constraints

The KTP700 Basic PN is the entry-level panel in the SIMATIC HMI lineup. It uses an ARM Cortex-A8-class CPU with limited RAM compared to the Comfort series. The Basic panel refresh rate is suitable for non-critical visualization — not for high-speed process control. For higher tag throughput, a TP700 Comfort or PC-based WinCC Runtime is required.

4. Acquisition Cycle and Tag Update Configuration

4.1 Setting the Acquisition Cycle

Navigate in TIA Portal:

1. Project tree → HMI [KTP700] → HMI tags → [Tag Name]
2. Open Properties → Properties tab → Acquisition cycle
3. Set the cycle from the dropdown (default is 1 s). Standard cycles include 100 ms, 250 ms, 500 ms, 1 s, 2 s, 5 s, and 10 s.

Recommended values by application type:

4.2 Selecting the Update Mode

In the same Properties view:

1. Open Update property.
2. Select Cyclic in operation for screen-bound variables.
3. Reserve Cyclic continuous for tags used in global scripts, persistent faceplates, or trend views that span screens.
4. Use On change only when the PLC triggers a job mailbox update via S7 function (rare in Basic panels and firmware-dependent).

4.3 Group Tags by Update Path

For faceplate tags (user-defined faceplates with multiple tags), use area pointers instead of polling each tag. Area pointers allow a single S7 read of a contiguous DB area to update many tags atomically. In TIA Portal:

1. PLC → Properties → Connections → HMI Connection → Area Pointers
2. Define a pointer for the data block area covering the faceplate tags.
3. Reference the area pointer in the HMI faceplate tag configuration.

This reduces the per-tag request count and improves effective update latency.

5. Tag Throughput and Network Timing Analysis

Use the following calculation to determine if your tag polling load is sustainable on a KTP700 Basic PN.

Tag throughput formula:

tags_per_second = active_tags / acquisition_cycle_seconds

Example A: 50 tags, 100 ms cycle
= 50 / 0.1 = 500 tags/sec  ← KTP700 Basic will queue

Example B: 50 tags, 500 ms cycle
= 50 / 0.5 = 100 tags/sec  ← Sustainable

Example C: 50 tags, 1 s cycle
= 50 / 1.0 = 50 tags/sec   ← Comfortable

The S7 protocol uses one request/response pair per tag (2 TCP frames). Each S7 read is at minimum ~80 bytes, so the wire bandwidth required is:

bandwidth_bytes_per_sec = tags_per_second × 2 × 80
At 500 tags/sec = 80,000 bytes/sec = 640 kbit/s
At 100 tags/sec = 16,000 bytes/sec = 128 kbit/s

On a 100 Mbit/s PROFINET segment the wire bandwidth is rarely the constraint. However, the panel CPU must still process each request, parse the response, and update the screen object. The KTP700 Basic's processor typically saturates above 300-400 tags/sec on a sustained basis. Above this rate, response frames queue and visible lag appears.

5.1 SVG Timing Diagram — Polling Cycle

At a 500 ms cycle, the panel CPU has roughly 5× more headroom to handle other tasks (script execution, alarm handling, trend archive writes). Most user-perceived lag is resolved at this setting without any hardware change.

5.2 PLC Scan Time Impact

The S7-1200 OB1 scan time directly affects the worst-case HMI read latency. If OB1 takes 20 ms and the panel issues a read at the start of a cycle, the PLC may not have the latest data until the next cycle. To check:

1. TIA Portal → PLC → Online & Diagnostics → Cycle time
2. Note the min/max/current cycle time and the cycle time monitor graph.

If the max cycle time exceeds 50 ms, optimize the PLC program before troubleshooting HMI lag further. Common optimizations: move heavy math to cyclic OB30/OB35 (S7-1500) or reduce the number of monitored tags in the PLC.

6. PLC-Side and Network Optimization

6.1 Verify S7 Connection Count

1. Project tree → PLC [S7-1200] → Properties → Communication
2. Note the connection count assigned to HMIs, PG, and OPC.

If the count is at the CPU limit, reduce HMI connections or upgrade to a CPU with more resources (CPU 1215C = 16 connections).

6.2 Use Integrated S7 Connection per HMI

Configure one S7 connection per HMI panel rather than one connection per tag group. WinCC TIA Portal supports integrated connections which multiplex multiple tag groups over a single S7 connection. In the HMI tag editor:

1. Select all tags → Properties → Connection
2. Verify all tags reference the same S7 connection.

6.3 Reduce Cross-PLC Traffic

If the S7-1200 is communicating with other PLCs (S7-300, ET200, third-party devices) over PROFINET, ensure the Update time of PROFINET IO devices is appropriate. Setting IO update to 1 ms on devices that do not need it consumes bandwidth that HMI polling could otherwise use. A 4 ms update time is usually sufficient for non-safety I/O.

6.4 Enable PROFINET Diagnostics

Use TIA Portal's Online & Diagnostics view to confirm:

• PROFINET device status (online/offline, error count)
• Port statistics (discarded frames, CRC errors)
• Diagnostic buffer entries

Replace any damaged cables or faulty switches — physical layer issues can cause TCP retransmits that look like HMI lag but originate in the network.

6.5 Network Cable Specification

Use only Cat 5e or higher rated cable for PROFINET, maximum segment length 100 m. Industrial cable (e.g., 6XV1 840-2AH10) is recommended for factory environments with EMI. Avoid running PROFINET cables parallel to VFD output cables for distances greater than 1 m without separation.

6.6 PLC-Side Ladder Example for On-Change Update

Where "On change" update is required, trigger the HMI from the PLC using a job mailbox:

// Trigger HMI tag refresh on value change
// DB_HMI.RefreshTrigger is a BOOL used as on-change bit
// DB_HMI.RefreshID is the job mailbox ID for the HMI

A "MotorState"           // Detect state change
= "DB_HMI".RefreshTrigger   // Set the trigger

// HMI picks up the trigger and reads the new tag value once

This pattern is useful for high-priority alarm or fault indications that must be reflected immediately but are otherwise infrequent.

7. Screen Design Best Practices

7.1 Reduce Screen Object Count

A KTP700 Basic PN screen with 100+ graphics objects (sliders, gauges, trend views) requires significant repaint work. Limit active objects per screen to less than 50. Archive or navigate away from heavy screens when not in active use.

7.2 Use Appearance Instead of Visibility Animation

Visibility animations trigger a re-evaluation of associated tags every acquisition cycle. Excessive use of visibility tied to fast-acquisition tags compounds lag. Use Appearance animation (which only triggers on tag change) instead of Visibility where possible. The two are visually equivalent for status colors.

7.3 Limit Trend Views

Each trend view can issue multiple tag requests per second. Limit trends to 1-2 per screen and set the archive cycle to 1 s or longer. Do not run more than 4 trend views across the entire project.

7.4 Disable Animations During Commissioning

Temporarily disable all animations (movements, color changes) and verify whether lag persists. If lag disappears, animations were the cause. Re-enable one at a time to identify the offending element.

7.5 Avoid Complex Faceplates on Basic Panels

User-defined faceplates with many internal tags add a tag manager overhead. A faceplate instance counts each tag inside it. Ten instances of a faceplate with 5 tags = 50 tags added to the polling queue. For Basic panels, prefer template screens over faceplates where possible.

7.6 Script Discipline

VBScript in WinCC Professional V13 SP1 Upd4 executes on the runtime PC. Inefficient loops in scheduled scripts can starve the tag manager. Avoid polling in scripts — read tags once at script start and cache the value locally.

8. Diagnostic Procedure

8.1 WinCC Runtime Performance Diagnostics

WinCC Professional V13 SP1 includes a Performance Analysis tool. Use it to identify:

• Slow tag update times (per tag)
• High CPU load on the runtime PC
• Frame rate issues (especially under simulation)

On the physical KTP700, navigate to Start → Settings → System → System Information → Memory to view CPU and memory load. If the CPU load exceeds 80% on a sustained basis, the panel is saturated.

8.2 PROFINET Frame Analysis

Use a managed switch with port mirroring and capture traffic with Wireshark. Apply a display filter for the S7 protocol (port 102) to isolate HMI traffic. Typical observations:

• 100+ S7 read frames per second = overpolling
• TCP retransmissions = network issue
• PROFINET alarms = IO device problem

8.3 Cross-Reference with PLC Buffer

Check the S7-1200 diagnostic buffer for Communication errors:

• TIA Portal → PLC → Online & Diagnostics → Diagnostic buffer

Repeated "Connection lost" or "Resource exhausted" entries indicate the panel's polling rate exceeds what the PLC can serve. Increase the acquisition cycle or reduce the number of active connections.

8.4 Force Table Test

Use the PLC's Monitor & Force Table to force a tag value change. Verify the HMI reflects the change within the configured acquisition cycle. If lag exceeds the cycle, the issue is at the PLC or network layer; if lag is within the cycle, the panel's tag manager is the bottleneck.

9. Step-by-Step Resolution Procedure

Apply the following changes in order. After each step, re-test on the physical panel before proceeding.

1. Audit acquisition cycles. Set all fast-changing tags to 500 ms minimum. Set status bits to 1 s.
2. Audit update modes. Change screen-bound tags from "Cyclic continuous" to "Cyclic in operation".
3. Verify S7 connection count on PLC is less than or equal to available connection resources. Consolidate multiple connections into one per HMI.
4. Reduce tags per screen to less than 40 where possible. Split screens with too many tags.
5. Disable animations temporarily and verify lag behavior.
6. Check PROFINET cable integrity and switch port statistics for errors.
7. Re-test on the physical KTP700 PN (not just the simulator).
8. Document final values in the project notes for future reference.

10. Verification and Acceptance Test

After applying the changes, perform the following acceptance test:

1. Place the panel into runtime and open the affected screen.
2. Force a value change in the PLC via Monitor & Force Table.
3. Verify the IO field updates within the new acquisition cycle (e.g., 500 ms).
4. Observe for 5 minutes under normal operating conditions — confirm no regression in other screens.
5. Check the HMI's system information page for sustained CPU load less than 80%.
6. Review the PLC diagnostic buffer for new errors.
7. Sign off the change in the project change log.

If lag persists after the acceptance test, escalate to TIA Portal V13 SP1 Upd5 or a later version (V14 SP1, V15.1, V16). Later TIA Portal versions include performance improvements for HMI runtime. Also consider migration to KTP700 Comfort or TP700 Comfort for higher performance, or use of WinCC Runtime Professional on a PC for higher tag throughput.

11. Performance Comparison: KTP700 Basic vs Comfort

If your application requires less than 200 ms visual response across many tags or a high tag count (>500), migrate to a Comfort panel or WinCC Runtime on a PC. The Basic panel is a deliberate cost-optimized choice for non-critical visualization.

12. Troubleshooting Matrix

References to Manufacturer Documentation

For detailed engineering rules, consult:

• Siemens Industry Online Support — Main portal for SIMATIC manuals, KB articles, firmware notes, and FAQs.
• TIA Portal V13 SP1 Help: search "HMI tag acquisition cycle" in the installed help portal.
• KTP700 Basic PN Operating Instructions — entry in Siemens support portal under HMI operator panels → 7" panels.
• SIMATIC S7-1200 System Manual — entry in Siemens support portal under SIMATIC S7-1200.
• S7-1200 Communication function manuals (S7 communication, OPC UA) for advanced integration.

For cross-vendor reference on HMI communication delay settings, see:

• Delta HMI Communication Delay Time FAQ — documents a 4 ms inter-request delay between device feedback and the next HMI request.
• Schneider Electric Magelis HMI Communication Lag FAQ — documents that too many actions on a Vijeo Designer HMI can cause variable refresh lag.

FAQ