An HMI showing a black screen does not necessarily mean the panel is broken. A supply voltage drop, a backlight problem, hanging during start-up and the wrong start screen can all look similar to the operator. Correct diagnosis begins by separating the symptoms rather than replacing parts at random. The first aim is not to get the screen working at any cost; it is to protect personnel safety, the state of the machine and the application data on the panel.
01First check: what state was the machine left in?
When the screen is unusable, it is established whether the machine is continuing its automatic cycle, which commands the operator can no longer see, and how a safe stop will be carried out. It should not be assumed that power-cycling the HMI will also restart the PLC. Similarly, a dark screen does not mean that the machine is de-energised.
Measurement, disassembly or wiring work inside the control panel is carried out by authorised personnel under appropriate energy isolation. Checks that require work on the mains side are not included in the operator checklist. The model's user manual and the plant's permits to work always take precedence.
02Place the symptom in one of four categories
| Observed condition | Possible area of investigation | Useful record |
|---|---|---|
| No light or sign of start-up at all | Power supply, connections, input protection | The last event before power was applied |
| Restarts at the start-up logo | Supply stability, storage, application | Restart interval and video |
| Screen lit but no picture | Start screen, software or display path | Last loaded project and version |
| Picture present, values not updating | PLC communication, addressing, network | Connection error and timestamp |
This distinction is more valuable than a one-line service request saying “the panel will not start”. For example, process values remaining fixed while a live clock keeps running is a different situation from the whole device freezing. Likewise, the Ethernet link light being on does not prove that application communication is healthy.
03Do not assess the power supply only at no load
Some panels draw more current at start-up than in stable operation. A voltage that looks normal at the power supply output can drop by the time it reaches the panel through a long or loose connection. Authorised personnel assess the measurement at the device input and under the condition in which the fault occurs; if logging is possible, short dips are examined.
A single voltage range cannot be given here for all panels. The rated value, permitted tolerance, polarity, supply isolation and the manufacturer's connection instructions are checked model by model. Trying a different adapter or power supply simply because “it says 24 V” can cause new damage.
04Reloading should not be the first step
The panel may hold recipes, user settings, operating counters or project data that exists only on the device. A factory reset, formatting or loading firmware at random can erase these. First, the date of the existing backups, whether an editable project exists and the panel's full model code are established.
A downloadable runtime file and an editable source project may not be the same thing. The licence, communication driver and panel operating system version are also part of compatibility. If a new backup can be taken, the current state is preserved first; any changes are planned after that. Whether the file obtained can actually be restored is a separate matter to test.
05A simple diagnostic scenario
Suppose an operator screen returns to the start-up logo every 20–30 seconds. Before concluding that “the software is corrupted”, it is recorded whether any other loads switch on at the same moment as the restart. If the fault occurs only when a valve group energises, a shared supply or connection problem is a more meaningful starting hypothesis.
This is a diagnostic example, not a fault verdict in itself. The same symptom can also appear with a storage fault. A good investigation selects the measurement that will confirm the hypothesis and records the finding. A cause that has not been measured may have been masked by chance through a part replacement.
06Information package to send to the service team
- Full model, serial number and a photo of the device label.
- A short video of the start-up behaviour; LED and audible indications.
- Information on any power cut, liquid exposure or project change before the fault.
- PLC model, communication type and the date of the last working project/backup.
- Recipes and user parameters that must be preserved.
07What should be verified after the repair?
A picture appearing is not sufficient for acceptance. Touch detection, communication, recipe read/write, user permissions and a controlled restart must be tested. It is checked that critical commands affect the correct machine and that loss of connection is clearly shown to the operator. The device must be verified not only on the bench but also under real connection and operating conditions.
08Example verification scenario
Suppose a panel shows a picture but does not respond to touch. First, it is recorded whether the screen is actually refreshing and whether the problem affects all areas. Signs of a protective film, contamination or physical pressure can be inspected from the outside. The diagnostic method supported by the model is then assessed by authorised personnel. A communication fault and a touch fault on the same device are not lumped together under one heading. If replacing the panel is being considered, resolution, project compatibility and connection interfaces are checked together.
Technical references and further reading
For product- and version-specific values, the relevant manufacturer's current documentation should be taken as the reference. The calculations and scenarios above are engineering examples for illustrative purposes.
