Clear HMI USB log storage from the panel and archive first

Can an HMI panel clear its own USB log storage?

Most panels that log to USB include a maintenance or setup screen for storage. One widely documented path is to press and hold the upper-left of the touch screen until test mode opens, press the memory button, press clear, then pick the storage device.

Hardware manuals from several panel makers document this kind of screen: setup pages that back up project and log data to an SD card or a USB pen drive, and a clear action for the selected memory. Button names vary between panels; the function is the same. Find the same screen in your panel's manual before you use it, so you know which devices it lists and what it deletes. A clear is destructive and immediate: it empties the device you select and nothing else. If your manual shows no storage screen, the fallback is to pull the stick and format it in a PC, which takes the panel out of logging for the few minutes the drive is out.

Will the panel overwrite the oldest log files when the drive fills?

Usually no: unless your logging configuration has a documented overwrite-oldest or file-count rotation option, expect the panel to stop writing new records or raise a storage error once the drive reaches capacity.

FIFO means first in, first out: when capacity is reached, the oldest file leaves so the newest one fits. A USB destination behaves that way only when the logging setup says so. One major panel line documents an overwrite-oldest mode tied to a maximum file count: once the count is reached, the next file starts and the oldest one is overwritten. With no rotation configured, the drive simply fills, and then the logger has nowhere to put new records, so writes stop or an error appears. Treat FIFO as unavailable until your panel's manual says otherwise. Look in the project software's data logging configuration for names like overwrite, rotation, number of files, retention or stop on full. If the drive is already full, confirm the oldest-file rollover first so you know whether files were rotating before you clear anything.

What do the symptoms tell you before you clear anything?

Match the symptom to the cause before you touch the clear button: no new records with a full drive points at capacity, new records with old files disappearing points at active rotation, and free space with no writes points at a device fault.

• No new records and the drive reads full. Capacity is the cause. Run the archive-and-clear below.
• New records arrive while older files disappear. Rotation is already active. Do not clear mid-cycle; check the configured file count and retention in the logging setup first.
• Free space remains but no records arrive. This is a device or write fault, not a capacity problem. Test the stick in a PC as described in USB drive not recognized by an HMI panel: test the drive, and check that the logging setup still points at the right storage device.
• A storage error survives a verified clear. Stop clearing. The fault is in the stick, the port or the panel, and the last section below gives the order to check.

Decide the case, then act: capacity gets the archive-and-clear, rotation gets a configuration check, and a device fault gets the drive test before anything is deleted.

How do you archive and verify the log files before clearing?

Copy the files you need to two places and check the copies open with the right date range before you press clear, because the clear action deletes the selected device's contents and there is no undo.

Pause logging first with the project's documented controls, so the panel is not writing while you work; some service menus offer a stop-storing action for exactly this. Then remove the stick the documented way, and if your panel has no eject action, add one as shown in remove a USB data-log stick safely: add an eject button, so a mid-write pull cannot damage the file. Copy the whole log folder to a PC and to a second location such as the plant network share. Opening the file list does not prove the copies are usable: open the first and the last file in the archive and check the record range covers the retention your quality system requires. Before the clear, confirm what the screen will delete. Some panels keep project backups or recipes on the same drive; move those files too, or clear a device that holds logs only.

How do you clear the USB storage from the panel?

Clear from the panel's documented memory screen with logging paused: on the common path that means hold the upper-left of the touch screen until test mode opens, press the memory button, press clear, select the USB storage device, then follow the confirmation prompts.

1. Pause logging with the project's documented controls.
2. Enter the maintenance mode your manual documents; on the common family, press and hold the upper-left of the touch screen until test mode opens.
3. Open the memory screen and read the device name and capacity it shows.
4. Press clear, select the USB storage device, and follow the on-screen confirmation.
5. Exit maintenance mode and turn logging back on.

Several makers put backup to an SD card or a USB pen drive and the clear action on the same setup screens, so your manual may show a setup-screen path instead of a test-mode path; both are documented routes. If the device name and capacity are not shown, or the target could be more than one device, cancel and read the manual again. A broad format aimed at the wrong device erases project backups and recipes along with the logs.

How do you confirm logging restarted after the clear?

Watch for one new record with a current timestamp: a log written every 5 seconds should show a fresh record within about 15 seconds of restart, and a once-per-minute log within about 2 minutes.

Then check the file list: a new dated file should appear, and the free-space reading should match a drive that holds only fresh records. For panels that start a new file each day, how an HMI trend log file rolls over at midnight on USB shows what the file list should look like after a clear. If no record arrives within three logging intervals, stop and go to the last section; do not clear again, because a second clear deletes the evidence the fault needs.

What does a worked sizing example look like?

Size the drive as retention days times measured megabytes per day, then read the result against a standard stick size: 84 retention days at 6 MB per day needs 504 MB, and an 8 GB drive sits about 6 percent full at cycle end.

Scenario: a plant must keep 12 weeks of quality records, and the panel writes one CSV file per shift, three shifts a day, each about 2 MB.

• Daily growth: 3 shifts × 2 MB = 6 MB per day.
• Retention: 12 weeks × 7 days = 84 days.
• Required capacity: 84 days × 6 MB per day = 504 MB.
• Drive fitted: an 8 GB stick, which is 8000 MB at the decimal scale drive makers use. Fill after one full cycle: 504 / 8000 = 0.063, so about 6 percent.
• Margin: 8000 / 504 = 15.9, roughly sixteen cycles of room for a skipped clear, larger trend blocks and file-system overhead.

Measure the megabytes per day from the panel's file list or the folder properties on a PC after one week of real production, because trend-heavy screens can double the figure. Check the file layout too: if your logger writes one continuous file instead of daily files, FAT32 caps a single file at 4,294,967,295 bytes, so split the log by day or by shift before the ceiling.

What if writes still fail after a verified clear?

Treat a storage error that survives a verified clear as a drive, port or panel fault, and work that order: test the stick in a PC, fit a second industrial-grade stick, then check the panel's storage selection before calling support.

• Test the stick in a PC. Copy a large file to it and back. A stick that reads fine can still fail at the small repeated writes logging produces; move to an industrial-grade stick.
• Check the file system. Most panels read FAT32. A stick formatted NTFS or exFAT can appear on a PC and still be unreadable or unwritable in the panel; reformat FAT32 once the archive is verified.
• Fit a second stick from a different batch. If two known-good sticks both fail, the fault is in the port or the panel.
• Call the panel maker's support channel with the model, software version, the logging configuration from the project, the storage screen reading, the exact error text, and the result of the two-stick test.

Recommendation: archive, verify, then clear from the panel's memory screen, and make that the last manual clear you need. Measure your megabytes per day, multiply by your retention days (84 days at 6 MB per day needs 504 MB, so an 8 GB stick leaves about sixteen times the margin), and set the documented overwrite-oldest option in the logging setup if your panel offers one. Keep the clear procedure from the manual posted at the panel, because the next full drive will belong to a different shift.

Questions