How-to
How to Disable a Remote HMI Display Through PLC Control
Short answer
Put the interlock in the PLC, not the panel. Read each panel's current screen through Remote Display Number, force display 21 through Replace Display Number while the disable bit is latched, and on release write the stored number back, hold one second, then write 0 so the panel never reports it is being controlled remotely.
On this page
Latch a disable bit in the PLC, move the far panel's Remote Display Number into a holding register, move 21 into that panel's Replace Display Number, and on release move the stored number back, hold 1.00 s, then move 0 — that final 0 is what stops a PanelView Plus locking with a message that its display is being controlled remotely. The panel supplies only two connections from the Display tab of its FactoryTalk View global connections: Remote Display Number reports the screen the operator is on, and Replace Display Number forces a screen change while it holds a nonzero value. The worked installation below ran FactoryTalk View ME 5.10 on PanelView Plus firmware v6.10 or later against an SLC 5/05 over Ethernet, with TON T4:0 preset to 1.00 s. Treat the result as an operator lockout only: emergency stop and safety circuits never pass through it.
The two connections that do the work #
Both fields sit in FactoryTalk View Studio under Global Connections, on the Display tab, and both point at controller integers you choose:
| Connection | Data direction | Behavior |
|---|---|---|
| Remote Display Number | Panel to controller | Continuously writes the number of the display currently shown |
| Replace Display Number | Controller to panel | While the integer is nonzero the panel changes to that display number; at 0 no replacement and no navigation occur |
| Close an On-Top Display | Panel | Closes an On-Top display; leave it out of this scheme entirely |
Number every operator display in the project from 1 upward, and put the message screen on an unused number above the highest real screen, 21 in this example. Create display 21 as a Replace display, not On-Top or Overlay: the Replace type swaps into the display queue cleanly, while a closeable display stacked on top of the current one is what fought the original attempt. The working two-panel installation ran FactoryTalk View ME 5.10 against an SLC 5/05 over Ethernet, with PanelView Plus terminals on firmware v6.10 or later.
A register map for two panels #
One working set of SLC-500 addresses; only the pattern matters, so map it onto your own controller:
| Address | Role |
|---|---|
| N7:0/0 | Disable request, set by the silo panel pushbutton |
| N7:0/1 | Disable latched |
| N7:1 | Stored display number, captured before the switch |
| N7:10 | Ground panel current display |
| N7:11 | Ground panel Replace Display Number |
| N7:20 | Silo panel current display |
| N7:21 | Silo panel Replace Display Number |
| T4:0 | Release delay timer, 1.00 s preset |
The N7 words are integers cleared to 0 at startup, the two N7:0 bits are Boolean, and T4:0 is a timer. The ground project publishes its current display to N7:10 and follows N7:11; the silo project does the same with N7:20 and N7:21. Nothing outside the controller ever writes N7:11 or N7:21 — one writer per Replace tag, always.
Build the lockout in eight steps #
- Number every display in each project from 1 upward.
- Create display 21 as a Replace display holding only a message such as "Display disabled by remote HMI" — no buttons, no input fields.
- In each project's Display tab, point Remote Display Number at that panel's read tag and Replace Display Number at its write tag per the map above, and delete any Close an On-Top Display entry.
- On the rising edge of N7:0/0, through a one-shot, run three operations in this order: move N7:10 into N7:1, move 21 into N7:11, latch N7:0/1.
- Condition every ground-panel command in the controller on the inverse of N7:0/1. The message screen is the visible barrier; this gating is the real one.
- When N7:0/0 drops while N7:0/1 is set, move N7:1 into N7:11 and start TON T4:0.
- When the T4:0 done bit sets, move 0 into N7:11 and unlatch N7:0/1.
- Verify end to end: navigate the ground panel to any screen, press Disable on the silo panel and confirm the switch to display 21, then clear the button and confirm the ground panel returns to the screen it left with no lockup message.
The order inside step 4 is not cosmetic: if the capture lands after the switch, N7:1 stores 21 and the release returns the panel to the message screen instead of the operator's screen.
The release timing, with the arithmetic #
Worked example, ground operator sitting on display 4:
- Press Disable. N7:1 = N7:10 = 4, then N7:11 = 21. The panel reads 21 on its next tag update and shows the message screen.
- Release the button. N7:11 = 4 restores display 4, and TON T4:0 starts.
- Timer preset. SLC timers count in the selected time base, so 1.00 s is a preset of 100 at the 0.01 s base, because 100 × 0.01 s = 1.00 s, or 1 at the 1.0 s base.
- Timer done. After 1000 ms the done bit sets, N7:11 = 0, and the latch clears. The panel reads 0 on its next poll and hands navigation back to its own buttons.
Keep the 1.00 s gap. It separates the restore write from the release write so both land; a release that arrives first strands the panel on the message screen. Shortening it buys nothing, and 1.00 s is the value the working installation settled on.
Why the first attempt locked the panel #
The original wiring put the expression "If {N7_1}=1 Then 21 Else 0" into the display connection and then added Close an On-Top Display to clear the message screen. Both choices fight this scheme: Close an On-Top Display and Replace Display Number contend over the display queue, and on release the expression writes 0, which triggers no navigation at all, so the message screen never leaves by itself. The lockup message is diagnostic in itself — a panel reporting that its display is being controlled remotely is telling you its Replace Display Number still holds a nonzero value, and the cure is always the same sequence ending in a write of 0.
The single-writer rule governs the whole Display tab: one mechanism commands a panel's display. Buttons that stopped navigating once these connections were configured are the same failure seen from the operator's side, and goto display buttons locked by HMI global connections covers the other face of that rule.
One project or two #
Commit to two projects whenever the panels must behave differently — here, the silo panel carries the disable button and the ground panel must not. Build identical graphics with different tag suffixes, Pushbutton_1 in the ground project and Pushbutton_2 in the silo project, and download each project to its own terminal. One .mer on both panels writes identical addresses, so the controller cannot attribute a press, and gating button visibility with a shared tag gates both panels identically, which fixes nothing. Deleting one button before each download works, but it turns every release into two downloads anyway.
Where the terminals really are interchangeable peers of one machine, one .mer is the right answer, and it is also how one spare HMI panel covers seven machines: identical behavior, cheap sparing. When the requirement is instead to see the same screens at both stations rather than lock one out, that is a mirroring problem, and whether a panel can fully mirror another controller remotely is a separate question with its own limits.
When it misbehaves #
| Symptom | Cause | Fix |
|---|---|---|
| Panel locks with "display being controlled remotely" on re-enable | Replace Display Number never returns to 0 | The timer-done rung writes 0 before unlatching |
| Panel does not switch to the message screen | Replace Display Number tag or its communication shortcut wrong in that project | Ground project follows N7:11, silo project follows N7:21 |
| Panel returns to its startup screen instead of the operator's screen | Capture rung placed after the switch | Move N7:10 into N7:1 before writing 21 |
| Both panels show the disabled message | The per-panel Replace tags are crossed | Check each project's Display tab against the map |
| Navigation still fights the operator | Close an On-Top Display still configured | Delete the entry; display 21 is type Replace, so nothing is stacked to close |
Questions asked
How many controller connections does this scheme need per panel?
Two per HMI in the working installation, one for the tag traffic and one for the display control path. Count them against the controller's connection limit when you size the system: running short shows up as tags that stop updating before any panel reports an error.
What happens if both operators press their disable buttons at nearly the same time?
The controller decides. Make the two latch paths mutually exclusive so a second request is ignored while a latch is set and its release sequence runs. A race between panels then resolves to whichever write landed in the earlier scan.
Why not keep the whole thing in one panel expression, like If tag = 1 Then 21 Else 0?
An expression can pull the panel to the message screen, but on release it writes 0, and 0 releases control without navigating anywhere, so the panel stays stranded on the message screen. The panel also keeps no register holding the screen the operator left, so only the controller can capture and restore it. That is why the sequencing belongs in the PLC.
What does the disabled panel do after a power cycle?
On restart the panel loads its startup display and then reads N7:11, which still holds 21 while the latch is set, so it is pulled back to the disabled screen. Even if it lands on a control screen first, its writes are gated in the controller by the latch, so no command goes through.
Which display numbers can I use for the message screen?
Display numbers are user-assigned integers from 1 to 999 in this FactoryTalk View ME version. Give every operator screen a number from 1 upward and put the disabled message on an unused number above the highest real screen, so screens added later cannot collide with it.