DL05 Lockup After Power Blink Is a Firmware Brownout Bug

DL05 CPUs on remote oilfield pumping-unit panels locked up after short power blinks, and the only recovery was a site visit to remove power completely and reapply it. The cause is the CPU's handling of a brownout, not the panel wiring or the mode switch. AutomationDirect issued corrected CPU firmware for the DL05 (V4.24) and the DL06 (V1.20) that addresses the "Shutdown on Power Brownout" behavior. The steps below restore a dead unit tonight, list the fixes that waste time, and then apply the permanent repair with a check after each step.

Match the failure signature before you change anything

Two distinct end states occur after a blink, and both are the same fault. In one, the CPU indicates RUN but does nothing (locked up). In the other, the CPU drops out of RUN and stays out after power returns. A unit in either state does not recover on its own.

Observation	What it tells you	Action
Failure follows a power blink or glitch	Brownout handling in the CPU	Continue with this procedure
Fails with the switch in RUN and also in TERM	Mode switch position is not a factor	Do not spend time on switch settings
CPU shows RUN but outputs and logic are frozen	Locked-up state	Full power cycle, then plan firmware update
CPU is in STOP after power returns with the switch in RUN	Dropped-out state	Full power cycle, then plan firmware update
Neighbor units on the same system never fail	Site power quality and unit-to-unit variation both affect exposure	Update all units, worst sites first
Fails with no power event at all	Different problem	Stop here; check logic, watchdog and hardware separately

Check before moving on: you can tie each failure to a logged or witnessed power dip. A fleet-level symptom counter helps. One 40-unit system was reporting roughly 3 to 5 units per week needing a manual reset before the fix existed.

Restore a locked unit with a full power cycle

This is the temporary restore only. Remove all power to the CPU supply, wait for it to fully de-energize, then reapply. Toggling the Run/Terminal/Stop switch does not clear the lockup.

1. Make the pumping unit safe for a restart. The process comes back up under whatever start-up logic is in the program.
2. Open the panel disconnect or CPU supply breaker and hold it open until the CPU LEDs are fully dark.
3. Reapply power and watch for the CPU to enter RUN.
4. Confirm outputs respond and the program is intact before leaving the site.

Check: CPU reaches RUN and the unit runs its normal cycle. Log the unit as a repeat offender and continue with the firmware steps; a power cycle only resets the symptom.

Skip the fixes that do not clear a brownout lockup

These are the usual quick fixes. Each one has a specific reason it fails or costs too much.

• Off-delay timer on the incoming power. Most blinks are shorter than the timer's hold time, so the timer never drops out while the CPU still resets and hangs. The timer variant that works needs a PLC output to force the timer to drop out during a brownout, which is unusable when the CPU is the thing that locked up.
• Mode switch in RUN or TERM. The lockup occurs in both positions.
• Moving to the DL06. It has the same problem on its original firmware. Its 24 VDC auxiliary supply does not cure it.
• 24 VDC version of the DL05 without battery backup. The DC-powered unit still fails unless a battery holds the supply through the dip.
• UPS. A UPS with a clean output waveform (check it with a scope; output quality varies between brands) does mask the blinks. It is a valid stopgap for a critical panel. It is a poor permanent fix on a low-cost controller panel, because the UPS can cost as much as the PLC.
• Replacing the platform. Swapping to a different PLC family removes the symptom, but the firmware update below removes the cause on hardware you already own.

Check: none of these is on your work order as the permanent fix. A DC supply with a battery or a UPS is acceptable only as a documented temporary measure on a unit awaiting its firmware update.

Read the CPU firmware revision on every unit

The fixed revisions are V4.24 for the DL05 CPU and V1.20 for the DL06 CPU. Units running older revisions are exposed. Read the installed revision from the programming software's PLC information or diagnostic display, or from the CPU's label if it is marked. Confirm the exact location in the CPU user manual for your hardware.

Mechanism, briefly: on a marginal supply dip, the input supply rail falls below the CPU's brownout threshold and then recovers before a full reset has completed. A CPU whose brownout handling is faulty resumes in an undefined state. It either runs with frozen scan or stays in STOP. Timers and external delays on the AC feed cannot address that, because the defect is in how the CPU sequences its own power-down and restart.

Check: build a table of every DL05 and DL06 with location, serial number and current revision. Rank by how often the site sees power blinks and how far the site is from the shop.

Flash DL05 to V4.24 or DL06 to V1.20

Download the firmware only from AutomationDirect's official firmware download area for the DL05 and DL06 and use the update instructions that come with the package. Do not use the firmware for one CPU family on the other.

1. Upload and save the running program, including any retentive or V-memory values you rely on, before the update.
2. Bench-update one spare unit first and confirm it accepts the new revision and runs a test program.
3. Update the worst-affected field unit, then read the revision back and compare it to the target.
4. Download the saved program if the update procedure clears memory, or compare it to the saved copy if it does not.
5. Return the switch to RUN and confirm the CPU reaches RUN unattended after a normal power-up.

Check: the revision read back shows V4.24 (DL05) or V1.20 (DL06), the program matches the saved copy, and the process starts correctly on a normal power-up.

Separate the surge damage from the lockup

Fried DL05 CPUs on the same systems are a second, separate problem. One 40-unit system that had been running under six months lost five DL05 units to power surges, while earlier DL250/450-based systems lost units only to direct lightning hits. The firmware update addresses brownout behavior. It does not add surge protection.

• Add surge protection on the AC feed at each panel and bond it to a good ground.
• Check panel grounding on remote units, since long lines feeding a pump site pick up surges.
• Record whether a failure is a hang (recoverable by power cycle) or a dead unit (no LEDs, no recovery). Only hangs belong to the firmware fault.

Check: dead-unit failures are tracked in a separate count from lockups, so you can see whether the firmware update reduces hangs while surge counts move independently.

Prove the fix with a controlled power-blink test

Test the updated unit on the bench with a load-disconnected panel or a spare CPU before you trust it on a wellhead. Use a switched supply or a relay that interrupts the CPU power for varying short intervals, including interruptions too brief for an external timer to notice.

1. Load a test program that toggles an output and increments a counter so a frozen scan is obvious.
2. Put the mode switch in RUN and apply a series of short interruptions of different lengths, with the CPU running in between.
3. After each interruption, confirm the CPU returns to RUN without any manual action and the counter resumes.
4. Repeat with the pre-update unit if you still have one, to reproduce the baseline failure.
5. Install in the field and log every reset request over the following weeks against the pre-update figure (3 to 5 manual resets per week across 40 units on the affected system).

Pass criteria: no interruption leaves the CPU locked or in STOP, outputs come up in a known state, and the pumping-unit restart logic behaves as designed after each power return. Confirm the panel still starts the unit safely, because an unattended restart is exactly what the fix now delivers.

FAQ