Explainer
How the LS system area screen change register works
Short answer
Writing a screen number to LS0008 commands a GP-series panel to display that screen, but only while the system area is enabled in the panel's setup and a screen with that number exists. LS0008 sits at offset +8 from LS0000, so a block mapped from D100 puts the command word in D108. The T-tag and W-tag chain exists because the mapped word is the same path the PLC uses to command screens.
On this page
Yes, the write to LS0008 is the screen-change command. On the GP2400, LS0000 through LS0020 form the panel's system area, and LS0008 holds the requested screen number: write 3 there and the panel displays screen 3, as long as the system area is in use in the panel's setup and a screen numbered 3 exists in the project. The programmer chained a momentary T-tag through LS2301 to a W-tag because a system area mapped into PLC memory makes that word the same command path the PLC uses to change screens. A direct navigation action would move the display but leave the PLC blind to the request, and the next PLC write would overrule it anyway.
What a write to LS0008 does #
LS0000 through LS0020 is a reserved block: every word in it has an assigned function, so it is not ordinary application storage. Inside that block, LS0008 is the screen-change command word, and the value it carries is a screen number from the project's screen list. The panel treats the value as a request and switches to that screen when a screen with the number exists; a number that matches no screen is ignored and the display stays where it was.
The hardware manual for this generation lists the system area without detailing each word inside it, which is why LS0008 looks undocumented at first. The per-word layout is published in the manufacturer's system-area documentation.
The setup that puts LS0008 in command #
A write to LS0008 only commands a screen change when the system area is in use. In GP-Pro EX the setting lives in the display unit settings: open System Settings, go to Display Unit, and in the System Area group the System Data Area setting defines which LS range serves as the system area. With the area in use, its words connect to a memory range in the PLC, and that connection is what makes PLC-driven screen changes possible: the program writes a screen number into its side of the mapped words and the panel obeys, exactly as it obeys a panel-side write.
Without it, a write to LS0008 lands in internal memory and commands nothing. That is the first thing to check on a panel that ignores writes the project clearly makes: the setup, not the tag, decides whether the command word is live.
Why the button drives a T-tag, a W-tag and a register #
The direct route exists: a screen-change action attached straight to the button. Writing the register instead has two real advantages in a mapped project.
First, coordination. LS0008 is the shared command path. The PLC can write the same word, so a machine-mode change can put up the right screen with no operator involved, and a button that writes the register makes its request down that same path. Local navigation and PLC-driven navigation stay on one mechanism instead of two that know nothing about each other.
Second, the chain separates the action from the touch object. The T-tag sets LS2301 momentarily while the button is held; the W-tag, triggered by that bit, performs the write once. Because the trigger is a bit rather than the button itself, other logic can key off the same bit, and the write can carry conditions no touch object could express.
The chain has one trap: a momentary set and a continuously evaluated tag are not interchangeable. A tag that holds the write active repeats it and reasserts the screen number after anything else tries to navigate, so the button would appear to fight the rest of the project.
When an inherited button does something other than its label suggests, trace the chain before touching it: search the project for every reference to LS2301 and every write to LS0008, the same discipline that finds a hidden object behind a function key that does the wrong action.
Numbering the mapped PLC address #
When the system area is mapped, the setup gives the PLC-side starting address of the block. LS0000 is the first word of that block, and LS0008 sits eight words after it, at offset +8. Any shorthand for the command word, including the advice that it is the eighth integer in the PLC, only resolves to a real register once that starting address is fixed. Never count words from a presumed base: read the configured start from the project's system setup and compute the offset from it.
Cross-reference the mapped address in the PLC program before anything writes to it: search for writes, initialization moves and mode logic that touch the word. Keep the command word distinct from the status words elsewhere in the mapped range. The mapping also runs in the other direction, and it is the same mechanism that lets tags toggle PLC bits from HMI screens across the display boundary.
Worked example: a D100 map and three writes #
The numbers here are an illustrative setup, not a documented default: the system area is in use and mapped to PLC words from D100, the project contains screens 1 through 10, and screen 12 does not exist.
The block begins at D100, so LS0000 corresponds to D100 at offset 0, and the command word sits at offset +8:
D100 + 8 = D108
so LS0008 is D108 on the PLC side. Three writes then behave as follows.
| Who writes | Value into D108 | Result |
|---|---|---|
| Button press, through the T-tag and W-tag | 3 | Screen 3 exists, so the panel displays screen 3 |
| PLC rung, machine enters run mode | 5 | Screen 5 exists, so the panel displays screen 5 |
| PLC rung, faulty input detected | 12 | No screen 12 exists, so the request is ignored and the display stays put |
In every row the panel does not care who wrote the number: it reads the command word and switches to that screen if the screen exists.
Now the comparison that answers the original question. Replace the chain with a direct navigation action to screen 3 and press the button: the display shows screen 3, but D108 keeps whatever value it held, because a direct action never touches the command word. The program's picture of the requested screen is stale, and the next PLC write to D108 moves the display again without warning. The register path is what keeps the two writers honest, and that is the reason the chain exists.
Editing the project: keep the chain or replace it #
Decide from the configuration, not from how the button looks. If the PLC writes the command word anywhere in the program, or any logic watches it, keep the register path and change only the destination value or the trigger, confirming the W-tag still addresses the mapped system word. If the button's only job is opening one fixed screen and nothing in the PLC reads or writes that word, a direct screen-change action is simpler; remove the old T-tag and W-tag only after searching the project and confirming nothing else keys off that trigger bit.
Whichever way you go, leave exactly one command path per button. A button that both navigates directly and writes the register issues two requests per press, and the second one quietly wins in testing while the first resurfaces in production.
Before editing, save an unchanged copy of the project and record the system-area settings, the mapped start address above all, so the offsets you rely on are the configured ones. Then verify the chain end to end: run the project in simulation wired to the controller using the approach in testing an HMI simulation against a live PLC, press and release the button, and confirm the destination screen appears and stays after release with the trigger bit back inactive and no write reasserting itself. Exercise PLC-driven screen changes and local navigation in both orders and confirm neither side overwrites the other.
Make the transfer and the test with the machine in a state where an unexpected screen change cannot mislead an operator, and confirm the mapped address before anything writes to it, because the other words in that block carry assigned functions too.
Projects this old stay maintainable: GP-Pro EX supports screens made for models as far back as the original GP510 series, so the GP2400 project opens in the current editor and the system-area settings can be read and corrected there.
Questions asked
Does the PLC see the button press, or only the screen change?
Only what the chain writes. LS2301 sits above LS0020, outside the LS0000 to LS0020 system area, so the momentary trigger stays inside the panel unless that bit is separately mapped. The PLC sees the value land in the mapped command word, and any rung that watches that word can react to the request. A direct navigation action would give the PLC nothing at all.
Who wins if the button and the PLC write the command word at the same time?
Whichever request the panel processes last. One write from each side resolves cleanly to the later value. Trouble starts when both sides write continuously: the display can flicker between the two numbers as each write reasserts itself. Give every screen number one owner per machine condition, and let the other side request through logic that waits for an idle or acknowledged state.
Is LS0008 the screen-change register on every panel?
Do not assume it transfers. The layout described here is the system area of the GP2400 generation: LS0000 through LS0020 with assigned functions and LS0008 as the screen-change command. Register positions and the enabling settings are defined per panel family, so read the system-area documentation for the specific model before reusing the address in a different project.
How do I find the mapped starting address in an inherited project?
Open the project in the editor and read it from the system settings for the display unit, where the system area is configured; the project file is the authoritative record. Record that start address with your project backup, because every offset you compute for the system words depends on it. If the project file is gone, read the settings from the panel's own setup screens rather than guessing the base.