Troubleshooting
Unified Comfort Panel Maintenance Mode Reboot Loop
Short answer
A completed image transfer or factory reset proves nothing on its own: the panel is recovered only when it boots to normal runtime and repeats that after a second restart. Read the address shown in the maintenance dialog, work on the one interface, X1 or X2, that the procedure names, and run the vendor's reset for the exact displayed state. If maintenance mode still returns, stop repeating and replace the panel.
On this page
A Unified Comfort panel that comes back in maintenance mode after every restart has not been recovered, whatever the transfer tool reported. The image transfer and the factory reset are actions; the acceptance test is the settled screen after a full restart, and it has to pass twice. Work the fault in this order: read the address the maintenance dialog shows, keep the X1 and X2 interfaces separate, run the reset procedure that matches the displayed state — Siemens document 109821880 covers the state named maintenance mode with IP 0.0.0.0 — and judge nothing until the panel has restarted twice. If maintenance mode still returns after one complete, uninterrupted pass through the documented procedure, the software recovery is exhausted: record the evidence and replace the panel instead of running the same transfer a third time.
What maintenance mode is, and what ends it #
Maintenance mode is a service state. It exposes recovery and setup operations in place of the normal runtime, and a panel sitting in it cannot operate the machine. There are two documented ways in. The hardware way: switch the panel off, press the Maintenance button — you find it in the opening between the two interfaces X1 and X2 — then switch on and keep the button pressed until the boot splash screen appears. The software way: in the device Control Panel, open System Properties, select Reboot, and press the Reboot in maintenance mode button.
The way out is not a dialog choice: the panel leaves maintenance mode only by rebooting into runtime, which is why every judgment in this fault waits for the restart to finish. One safety rule before any of these sequences: if this panel is the only operator interface for a running line, bring the machine to a safe state before you power the panel down.
Read the dialog before you run anything #
Write down what the dialog shows before you change anything: the mode text, the IP address displayed, and which port the cable from the PC is in. The address decides the procedure. The published reset for Unified Comfort and Basic panels is written specifically for maintenance mode with IP 0.0.0.0 — document 109821880 — meaning the panel has come up without a usable address, and that document's sequence is the one to run.
A panel showing a real, configured address is in a different documented state; forcing the 0.0.0.0 sequence onto it by habit adds variables instead of removing them. One procedure per attempt, run end to end, and no steps borrowed from other panel families. The screen you photographed after the last failed attempt is not trivia: the address on it is the state the next procedure has to match.
Keep X1 and X2 separate through every attempt #
An address change on X1 is an X1 result only. It says nothing about X2's address, its link, or its reachability, and an OS transfer that then fails through X1 is a separate result again — record address changes, transfer attempts and reboot outcomes as three different statuses, never as one success. ProSave's MAC-address search and an IP connection are different access paths: a MAC search that comes up empty does not block the IP-based documented procedure, so run the documented path rather than retrying the search. Put the engineering PC on the port the procedure names and address it for that port.
Check the PC itself too: a second active adapter can own the route to the panel, which is exactly the failure described in the article on why a PC fails to reach an HMI when Wi-Fi owns its address.
Worked example: why a healthy panel still refuses the second port #
Take an illustrative setup: the PC's adapter at 192.168.0.10 with mask 255.255.255.0, panel X1 set to 192.168.0.50 with the same mask, panel X2 still without a usable address. The PC sits on network 192.168.0.10 AND 255.255.255.0 = 192.168.0.0, and X1 sits on 192.168.0.50 AND 255.255.255.0 = 192.168.0.0 — the same network, so the IP connection to X1 works.
Now suppose X2 carries 192.168.1.5: 192.168.1.5 AND 255.255.255.0 = 192.168.1.0, a different network from the PC's 192.168.0.0. Nothing from this adapter reaches X2 — no MAC search response, no ping, no transfer — however successfully X1's address was just changed. These are example numbers, not prescribed values; the real inputs are the address displayed for the port you are plugged into and the addressing the procedure document gives.
The reset pass that matches the IP 0.0.0.0 state #
Run the pass that matches the 0.0.0.0 state like this:
- Switch the panel off.
- Press and hold the Maintenance button in the opening between X1 and X2.
- Switch the panel on while holding the button, and keep holding until the boot splash screen appears.
- Let the panel settle in maintenance mode with IP 0.0.0.0 — the state document 109821880 is written for.
- Connect the PC to the interface the document names and set the PC addressing exactly as the document gives it.
- Check the image file against the exact panel before transferring; the guide on getting the right firmware file for a legacy HMI panel walks through matching the file to the device.
- Run the factory reset and the image transfer as the document sequences them, and let every step finish — no power cut, no cable swap, no closed tool mid-transfer.
- Restart and judge only the settled screen.
An interruption inside step 7 is enough on its own to put you back at the start of this list.
Acceptance checks after every restart #
After every restart, check four things:
- Final screen: normal runtime, not the maintenance dialog. A brief stage label while the panel boots decides nothing; the settled screen is the result.
- Address: the panel shows the address configured for the interface your project connection uses — not an address assumed from the other port.
- Reachability: the PC talks to the panel at that displayed address on the port you are plugged into.
- Repeat: from runtime, restart once more and confirm it returns to runtime. One clean pass is not recovery.
All four passing means the panel is recovered and ready for the project reload. The restart rule is the same one that closes out other boot faults — the write-up of a fatal E6 boot loop on a PanelView Plus 7 ends the same way, on the settled screen. If the fourth check fails and maintenance mode is back, the next section applies.
When the loop survives a complete reset #
If one complete pass — button start, matched state, correct port, uninterrupted reset and image transfer, two full restarts — still lands in maintenance mode with the same displayed state every time, stop. A third identical attempt cannot change an outcome the first complete pass already settled. What you have instead is the evidence pack: the address shown in the dialog, the port the PC used, the transfer tool's completion result, and the final screen after each restart.
Hand that to the panel vendor's service channel, because a loop that survives a verified-complete documented reset points at the panel itself, and no repetition of the transfer will move it. If the repair is not worth it, the direct replacement path is an industrial panel PC in the same opening, running the visualization as a PC-based HMI.
Questions asked
During the reboot a brief message reading Manufacturing mode appears — is that the failure?
No. It is a stage the panel shows on the way through its boot sequence, not a result, and none of the documented procedures turns on that label. Let the restart finish and read the settled screen: normal runtime means the panel recovered, the maintenance dialog again means it did not. Photograph the final screen with its address either way, because that is the evidence the next step uses.
Does the factory reset wipe the project on the panel?
Yes. A reset to factory settings returns the device to its delivered state, so the runtime project, the panel's network settings and anything else stored on it are gone. Keep a current project archive on the engineering station before you start, so the reload after a successful recovery is a transfer, not a rebuild.
Why can ProSave not find the panel by MAC address?
MAC-based discovery works at the link level: it only sees devices on the same link as the PC's adapter. A panel port on another link, or sitting without a usable address, will not answer, which is why the documented recovery for this state addresses the panel by IP instead. Work the documented IP path; an empty MAC search on one port says nothing about the other.
Is the reset the same on a Unified Basic panel?
The published reset document covers Unified Comfort and Basic panels together, but the entry differs: Basic panel hardware carries a physical reset button on the back of the device for a reset to factory defaults, while Comfort panels use the Maintenance button in the opening between X1 and X2. Confirm the button's exact location on your model before the first attempt.