IMDTouchKnowledge BaseShop main store

Troubleshooting

Why a Micro800 Program Runs but the C-More Panel Reads Nothing

HMI and touchscreens Published 6 min read

Short answer

Work four layers in order: USB enumeration under CCW, the program's task assignment and Run mode, the IEC 61131-3 TIME conversion, and tag scope. Keep TIME inside the logic, convert DINT milliseconds once with ANY_TO_TIME at the preset, and make every panel-facing tag global. A clean download proves nothing until each layer checks out.

On this page
  1. Work the stack in this order
  2. First: make the USB link solid
  3. Second: prove the program is scanning
  4. Third: convert TIME once, at the preset
  5. Fourth: set tag scope before the panel project
  6. Worked example: a 2.5 second bump timer
  7. The decision for each case
  8. Questions asked

A Micro850 that downloads clean while a C-More panel reads nothing is almost always four separate faults stacked in order: the USB link never enumerated, the program is not assigned to a task or the controller is not in Run, a TIME value is being treated as a number, or the tag is program-local and invisible to the panel. Work them in that order, because each layer masks the one below — do not touch tag scope while the USB link drops mid-download, and do not blame the panel driver until the value is visibly changing in the CCW variable monitor. The one standard in play is IEC 61131-3: TIME is a duration, not an integer, and it never crosses to DINT without an explicit conversion.

Work the stack in this order #

  1. USB enumeration — CCW sees the controller and holds the connection through a whole download.
  2. Execution — the program is assigned to a task, the controller is in Run, and no major fault is present.
  3. Data types — every panel-facing tag is DINT, INT, REAL or BOOL, and TIME stays inside the logic.
  4. Scope — every panel-facing tag is global, or addressed fully qualified.

Skipping ahead is what turns this into a three-day fault. A tag-scope fix made over an intermittent link proves nothing, and a type fix made on a program that is not scanning cannot be observed.

The Micro800 USB port is a programming port. CCW does not talk to it directly; it goes through the RSLinx Classic layer installed alongside it, which normally creates the USB driver in the driver list by itself the moment you plug the controller in. When that does not happen, work the causes in order of cost:

  • Device Manager shows an unknown device with the controller powered and cabled — the driver never bound. Let Windows search for it once, and if the wizard finds nothing, browse manually to the driver location from the software install.
  • The device appears, then drops mid-download — cable or hub. Use a direct port, not a hub, and turn off USB selective suspend on that root hub.
  • Browse works once and fails after a CCW restart — stale communication service state. Restart the service or reboot before you reinstall anything.
  • The device never appears at all — three causes turn up most often: antivirus intercepting ccwbasicd.exe, a stale RSLinx USB driver, or a firmware/CCW version mismatch. Add the exclusion, refresh the driver, and keep CCW at least as new as the controller firmware, because a controller flashed newer than the CCW version will not enumerate.

The decision: a CCW reinstall costs about an hour and is the last thing to try, not the first. Driver binding and the physical path fix the large majority of these links.

Second: prove the program is scanning #

A clean download proves compilation, not execution. Check these in order:

  • Controller mode — after a download the controller can sit in Program mode. Put it in Run explicitly and confirm the RUN indicator.
  • Task assignment — in the Micro800 project tree, a program (ladder, FBD or ST) must be assigned to a task. An unassigned program compiles and downloads but is never scanned. This is the most common cause of a perfect rung that stays dead.
  • Fault status — read it in CCW. A major fault halts scanning after a successful download, typically from a runtime conversion or an array index error, which is exactly what bad TIME handling produces.
  • Array bounds — every indexed access must stay inside the declared dimension. Out-of-range indexing on this platform is a runtime fault, not a compile error.

One safety note: before you switch the controller to Run, make sure the machine is in a safe state, because a program that starts scanning can command motion nobody is expecting.

The decision: outputs dead and timers flat means task assignment first, mode second, fault status third — all before you open the panel project.

Third: convert TIME once, at the preset #

IEC 61131-3 defines TIME as a duration type with its own literal syntax — T#500ms is half a second — and requires explicit conversion between TIME and the integer types. The Micro800 compiler refuses implicit assignment between them, and a timer's PT input takes TIME, not a millisecond count. The working pattern keeps the panel-facing value as a DINT in milliseconds and converts once, immediately before the preset:

VAR
    StartBump   : BOOL := FALSE;  (* run command *)
    BumpTimeMS  : DINT := 2500;   (* written by the panel, ms *)
    BumpPreset  : TIME;           (* internal only *)
    BumpElapsed : DINT;           (* read by the panel, ms *)
END_VAR

BumpPreset := ANY_TO_TIME(BumpTimeMS);   (* DINT ms -> TIME *)
TON_1(IN := StartBump, PT := BumpPreset);
BumpElapsed := ANY_TO_DINT(TON_1.ET);    (* TIME -> DINT ms *)

The rules that follow from it:

  • Never map a TIME tag to the panel. Expose DINT milliseconds on both the setpoint side and the elapsed side.
  • One conversion in each direction, on a rung or a line you can see — not buried in a UDFB, where a runtime fault gives you no context.
  • The ANY_TO_* converters accept a source of a different type. ANY_TO_TIME gets you into TIME from a DINT; it does not launder TIME into TIME. If the compiler rejects the instruction, the input type is wrong, not the instruction.

Fourth: set tag scope before the panel project #

Micro800 variables exist in two scopes, and this choice decides whether the panel can see the tag at all:

Scope Declared in How the panel addresses it
Global Controller-level global variables Plain name: BumpTimeMS
Program-local Inside one program Fully qualified: MainProgram.BumpTimeMS

Promote every panel-facing tag to global scope; it removes an entire class of addressing failures and makes the tag list self-documenting. If a tag must stay program-local, address it qualified, with the program name and case exactly as the project tree shows them. Check the variable's external-access attribute — a tag without it is not served to a client even when the name is right. And a mistyped address fails silently: the panel object shows a blank or placeholder, not an error, so copy names out of CCW instead of retyping them.

The panel reaches a Micro850 over Ethernet by symbolic tag names or by numbered registers. Symbolic access is where scope, spelling and external access rule; the register path is where word order and layout rule instead, the same trap you close when you swap words on a C-More 32-bit Modbus tag. Either way the data type must line up between the controller declaration and the HMI tag definition, the same boundary discipline as displaying a PLC real value over Modbus on an HMI.

Worked example: a 2.5 second bump timer #

The operator wants a 2.5 s bump, the panel entry is in seconds, and the controller tag is in milliseconds, so scale once: 2.5 s × 1000 ms/s = 2500 ms. BumpTimeMS receives 2500, and ANY_TO_TIME(2500) yields T#2500ms on the preset. When StartBump goes true, TON_1.ET counts from T#0ms toward T#2500ms, and ANY_TO_DINT(TON_1.ET) puts 0, 1, 2 … 2500 on the panel as plain milliseconds.

The test that proves the whole stack: watch BumpElapsed in the CCW variable monitor while the C-More numeric display runs. Both should step to 2500 together. If the monitor counts and the panel stays blank, the fault is scope, spelling or external access. If neither counts, the fault is execution or the link — go back to layer one.

The decision for each case #

  • No device in the CCW browse: fix the driver binding, the antivirus exclusion for ccwbasicd.exe, or the CCW/firmware version match. Do not reinstall CCW first.
  • Download clean, outputs dead: assign the program to a task, then put the controller in Run and read the fault status.
  • Value changes in the variable monitor but the panel is blank: scope, spelling, or the external-access attribute.
  • Compiler rejects the conversion: the source is already TIME — move the conversion to where the DINT lives.
  • Timer preset or display reads nonsense: a TIME tag is mapped to the panel. Replace it with DINT milliseconds and one ANY_TO_TIME at the preset, one ANY_TO_DINT on the elapsed side.

Questions asked

Why does ANY_TO_TIME reject a TIME input?

The ANY_TO_* converters exist to cross types, not to repeat one. ANY_TO_TIME turns a DINT into a duration; ANY_TO_DINT turns a duration back into a number. Feeding TIME into ANY_TO_TIME is the mirror error of feeding a DINT into ANY_TO_DINT: the source type is already the target. When the compiler rejects the instruction, the value at its input is already TIME, so move the conversion to the point where the DINT millisecond value actually lives.

Does a successful download mean the program is running?

No. A download proves the project compiled and transferred, not that it scans. The program must be assigned to a task, and the controller must be in Run mode with no major fault. Verify the mode after any power cycle before you decide the logic failed: if the controller shows Program, set Run and confirm the RUN indicator is lit.

Can the C-More panel read a program-local Micro800 tag?

Yes, if you address it in fully qualified form, with the program name and tag name exactly as the project tree shows them, including case, and the variable's external-access attribute permits it. The failure is silent: a wrong name gives a blank object or a placeholder, never an error. Copy the string out of CCW rather than retyping it, and remember that renaming the program later breaks the qualified reference the same way.

What data type should the C-More tag use for the timer value?

DINT, with the HMI tag set to the same signed 32-bit width as the controller declaration. Micro800 does not silently promote types, so mixing DINT, INT and REAL across the boundary shows up as garbage values, not error messages. Do width changes in logic at one visible point, the same discipline as the TIME conversions, and clamp the panel's numeric entry to the range the process can use, for example 0 to 10000 ms, so a mistyped value never reaches a timer preset.

Keep reading

  • Troubleshooting

    Why an HMI Writes Integers but Not B3 Bits Over DF1

    When an HMI writes T4, C5 and N7 data but B3 bits never change over DF1, the serial link is proven and the failure sits in the HMI software's binary-file translation. Check the engineering software version before baud, cable or PLC settings: a documented case was fixed by upgrading EBplus from 2.10 to 5.07.

    HMI and touchscreens

  • How-to

    Displaying a Do-More String on a C-More EA9 Panel

    Use the C-More EA9 Dynamic Text object bound to the Do-More short-string tag, and give it a trigger tag that is on whenever the text should update: ST1, the always-on Do-More system bit, works immediately, while a page C-Bit set on at page open and off at page close in the Event Manager ties the display to the open screen.

    HMI and touchscreens

  • Troubleshooting

    Stop a PanelView Timed Write Corrupting SLC Clock Registers

    The PanelView's time-status write transmits a two-digit year, so S:37 drops from 2006 to 6 after every update. Point the time and date status tags at an integer file and let ladder logic add 16#2000 to the year before moving the set into S:37 through S:42, or clear the tag entries on the Status Tags tab of the Terminal Setup dialog to stop the write completely.

    HMI and touchscreens

Why an HMI sends email in simulation but not from the panel