Troubleshooting
Why PanelView Plus Piloted List Selectors Do Not Update
Short answer
A piloted list selector marks a state only when the live tag value equals one of its configured state values. A missing selection indicator means the settled PLC value is outside the state list, so fix the mapping before blaming refresh. Rapid multi-write bursts are sampled, not replayed; test that the settled value appears without navigating away.
On this page
- Why does the selector show no selection indicator?
- Why do the selectors refresh only after leaving the screen and coming back?
- Why do some selectors update while others on the same screen never do?
- Does writing the tag several times per scan stop the HMI from updating?
- What should the display update rate actually be set to?
- How do I test one failing selector against the PLC?
- When is it a terminal or project fault instead of configuration?
- Questions asked
On a PanelView Plus, a piloted list selector that sits blank or stale is a value-to-state mapping problem until proven otherwise. The object marks an entry only when the live tag value equals one of its configured state values, so a settled value outside the state list leaves nothing to mark — no selection indicator, no back-color square. The pattern reported on a PanelView Plus 1000 (2711P-RDB10C, 5.10 runtime) — some selectors updating, others blank with no indicator, values appearing after a screen change — is that signature, not a refresh fault. Rapid writes, four or five per PLC scan, are sampled rather than replayed and cannot be displayed individually. The order of attack is fixed: verify the settled value against the state list, fix the mapping, then prove the settled value on the terminal before touching update rates.
Why does the selector show no selection indicator? #
Because the settled tag value matches none of the selector's configured state values: FactoryTalk View Studio builds a piloted list selector from states — value and text pairs — and marks the one state whose value equals the live tag value, so a value with no equal leaves no indicator.
FactoryTalk View Studio's ME help documents the setup under "Set up states for a control list selector" and "Set up general properties for a piloted control list selector"; each state pairs a value with the text shown for it, and the indicator is that state's back color. When the square is missing, check three things:
- the exact tag address bound to the object, not a similarly named tag;
- every configured state value against the value the PLC actually settles on;
- the tag's data type at both ends, because a tag read with the wrong type lands outside the state list even when the PLC value is intended — BCD is the classic case and the break appears above 9, the pattern explained in why an HMI counter display jumps from 9 to 16.
Two neighboring behaviors are separate from the indicator. Entries that must become unselectable as the machine changes state are handled by conditionally disabling piloted list selector options from PLC states. The list's scroll position has its own control, the piloted list selector top position property, which is independent of the state values.
Worked example: twelve states, one burst #
A selector configured with 12 states carries values 0 through 11, so its highest valid value is 12 − 1 = 11. The PLC writes 0, 3, 7, 11 and 9 — five writes in one scan — then holds the last value.
- Settled on 9: 9 ≤ 11, a state exists, and the indicator must appear at the next display update. If it does not, the fault is past the mapping — binding, access settings, or the display itself.
- Settled on 12: 12 > 11, so 0 of the 12 states match, and no update rate will ever draw an indicator. The fix is a state for 12, or a change to the logic that writes 12.
The decision: a missing indicator on a value above the top state is a configuration fault. Correct it and retest before suspecting the terminal.
Why do the selectors refresh only after leaving the screen and coming back? #
Opening a display reads its tags, so re-entry always gives one clean read, which is why the workaround appears to fix the screen while proving nothing about the cause — the real fault is the open screen not sampling the tag.
In FactoryTalk View ME, the display update rate is a property of the display, and every tag on that display rides the same rate — one rate per screen, not per object. That design carries a diagnostic: a rate that is too slow delays every object on the screen equally, and it can delay an indicator but never remove one. When re-entry is the only thing that refreshes a valid, settled value, suspect the object's binding first and the screen's rate second.
Why do some selectors update while others on the same screen never do? #
A display-level fault would affect every object on the screen equally, and it does not — one blank selector beside working neighbors points at that object's own tag binding, its state values, or its state access settings, which is exactly what to compare selector against selector.
Put one working selector and one failing selector side by side in the same conditions and compare, in this order:
- tag binding — the exact address, not a near name;
- state values — the failing object's list against the live value it settles on;
- state access, write on enter and handshake configuration — confirmed in the project, because a remembered "the same for all of them" is not a verified setting.
| Observed on the terminal | First cause to test | Action |
|---|---|---|
| No indicator; settled value outside the state list | Value-to-state mapping | Add the state or stop the logic writing that value |
| Indicator appears only after re-entering the display | Open-screen update path | Check the binding, then the display update rate |
| One selector blank while its neighbors track | That object's binding or state values | Compare the two objects' settings directly |
| Question marks instead of a value | Communication path | Fix the communications path before touching selector states |
Question marks are a different failure — the terminal cannot read the tag at all — covered in why a PanelView Plus logic module shows question marks on live values.
Does writing the tag several times per scan stop the HMI from updating? #
It hides the intermediate values, never the final one: a display samples its tags at the display update rate, so a burst of writes between two samples is invisible by design, and the value still held at the next sample is the only one that can be drawn.
Illustrative arithmetic with the reported numbers: at a 0.1-second update rate and an illustrative 10-millisecond PLC scan (0.01 s), the display samples once every 0.1 s ÷ 0.01 s = 10 scans. Four to five writes per scan means roughly 4 × 10 = 40 to 5 × 10 = 50 tag values pass between samples, unseen and unrecorded. That is sampling, not failure.
The acceptance test is the settled value: run the burst, let the tag settle on a state value, and require the indicator without navigating. If the operator must see a specific intermediate value, latch or hold it in PLC logic for at least one display update period — 0.1 s at the reported rate. Any such latch is a change to machine behavior, so validate it against the control design before it runs.
What should the display update rate actually be set to? #
Whatever the project file says — the reported figures contradict each other, 0.1 seconds configured against a claimed 0.05-second maximum, and both cannot be true, so read the setting in the display's properties in FactoryTalk View Studio before acting on a remembered number.
The rate decides lag and nothing else. Its unit is seconds, and it applies to all tags on the display, so the decision rule is simple: a blank selector with no indicator is never a rate fault, because sampling cannot remove a state match that exists. Change the rate only after the mapping is proven and the whole screen trails the PLC; then lower the seconds value in steps and confirm the terminal and its link sustain it. On PanelView Plus 5000 terminals built in View Designer, the equivalent screen refresh setting is expressed in milliseconds, and that platform's guidance discusses tag update frequency beyond 500 ms — the two settings are not interchangeable.
How do I test one failing selector against the PLC? #
Drive the tag to each configured state value, hold each across several display updates, and require the indicator on the terminal without navigating away — that one pass separates mapping faults from update faults and takes minutes per selector.
- Open the affected display on the terminal and go online with the controller; watch the exact tag bound to the failing selector and the tag bound to one that works.
- Write each configured state value to the tag in turn and hold it; record which values draw the indicator and which do not.
- Flag every live value with no matching state — each one is a mapping fault regardless of anything else you find.
- Run a burst of writes, let the tag settle on a valid state, and require the indicator without leaving the screen.
- Change one thing at a time — a binding, a state value, an access setting — and repeat the same test after each change.
- Accept the fix only on the terminal; a pass in Test Application is a logic check, not an acceptance test.
When is it a terminal or project fault instead of configuration? #
When a value that provably equals a configured state still draws no indicator on the terminal — at that point binding, mapping, access and rate are excluded, and the remaining suspects are the deployed runtime project and the terminal itself.
Work the exclusions in the order above, then redeploy the runtime project cleanly to the terminal and repeat the settled-value test once. Keep the evidence for the support case: the exact tag values observed online, the state lists as configured, and the behavior on the terminal versus Test Application. A selector that passes on a redeployed project and still fails on one terminal while its siblings behave is a hardware support case, not a configuration one.
Questions asked
Do the write-on-enter and handshake settings affect the missing indicator?
No. Write on enter and the handshake control how an operator's selection is written back to the PLC and acknowledged; the indicator is driven only by the tag value read from the controller. Leave both at their configured values while you troubleshoot the indicator, and verify those values in the project rather than from memory.
Will rebuilding the selectors from scratch fix it?
Not if the state values are retyped with the same mismatch. Record each selector's state list first, correct the values against the settled PLC values, and rebuild only the objects that still fail the settled-value test. A rebuild that recreates the same state list reproduces the same blank indicator.
How long must a PLC value hold for the selector to catch it?
At least one full display update period. At the reported 0.1-second rate, a value held for 0.1 seconds is sampled with certainty; anything shorter races the sampler and may never be drawn. Latch display-critical values in logic for one update period or longer.
Is a difference between Test Application and the terminal meaningful?
Yes. Test Application checks project logic on the engineering workstation; the terminal runs the deployed runtime. A mapping fault appears in both. A selector that indicates in Test Application but not on the terminal points at the deployed runtime project or the terminal itself, so redeploy before editing objects.