When you're on-site and a Siemens S7-1200 or S7-1500 won't power up, the schedule stops. I coordinate emergency service for an integrator in Monterrey. We get these calls at 3 PM for a line that needs to run by 6 AM the next day. You don't have time to read a manual. You need a checklist.
This is the sequence I use when triaging a 'No Power' or 'CPU won't start' call. There are five steps. The fourth one is a thing I see almost every new tech miss.
Is the CPU completely dead (no LEDs)? Or does it power on but flash a red 'SF' or 'BF'? This changes the first step. If it's totally dark, skip to Step 1. If it has a red LED, go straight to Step 3.
I know, checking if it's plugged in sounds insulting. But I've driven an hour to find a machine that was unplugged (which, honestly, felt excessive). The nuance is how you check it.
Use a multimeter to test the incoming voltage at the power supply's terminals. Don't trust the machine's panel lights. I've seen a 24V DC supply showing 19V on the terminals; the panel light was on, but it wasn't enough voltage for the CPU to boot reliably. According to Siemens documentation, the S7-1200 requires a 24 VDC supply with a range of 20.4 to 28.8 V. If you're below 20.4, the CPU will not start reliably, even if the power supply light is on.
Also, check if you're testing the car battery of the backup system? Not really. But the principle is the same—load testing. A voltage reading under no load means nothing. If you want to know the capacity, use a quick battery charger on the backup UPS, but for the PLC supply, measure at the terminals under load.
This is where the 'how to tell if a circuit breaker is bad' skill comes in. A tripped breaker is obvious—it's in the middle position. But a bad breaker? They fail closed (staying on) or open (appearing tripped).
I had a call in July 2024 where a line had intermittent power loss. The breaker looked fine. It was in the 'On' position. But using a multimeter, I found zero voltage on the load side. The internal contacts had failed open. The breaker was bad, but it didn't look tripped. So, step two: measure voltage on both the line and load side of the breaker feeding your PLC cabinet. If you have voltage in but not out, replace the breaker.
Also, check the 24V DC fuses. Use the ohm setting on your multimeter (continuity check) on the fuse itself (after disconnecting power). A blown fuse looks fine 90% of the time.
If the power is good (Step 1 & 2 pass) but the CPU shows a red 'SF' (System Fault) or 'BF' (Bus Fault), you need to look at the diagnostic buffer. You can't do this without TIA Portal software connected. This is where having a laptop ready is mandatory.
In TIA Portal, go online, right-click the CPU, and select 'Open Online Diagnostics.' The diagnostic buffer will tell you exactly why the CPU stopped. It's usually one of three things:
The message is usually precise. It will say 'Hardware component removed at rack 0, slot 2' for example. This saves you hours of guessing.
Here's the thing I never expected to be a top cause of issues. The S7-1500 modules connect via a U-connector at the back. If you plug a new I/O module in, but forget to insert the U-connector between the modules, the CPU will see a bus error and may refuse to start. Or, the connector can become dislodged during vibration.
I saw this on a conveyor line three months ago. The tech had swapped an output module. The CPU kept flashing red. He checked the fuse, the voltage, the program—everything. The surprise wasn't the code. It was that the U-connector was sitting in the bottom of the cabinet. It had never been installed after the swap. The CPU couldn't see the new module, so it faulted on start.
Physically open the cabinet. Look at the rail. Are all the connectors seated? Push them together. This is a five-second check that solves a surprising number of cases.
Some Siemens PLCs (especially S7-1200 and S7-1500) use a SIMATIC Memory Card. If the card is corrupt, missing, or has the wrong firmware version, the CPU will not start. The LEDs often flash in a specific pattern: 'STOP' LED flashing slowly means 'no memory card.' Check this.
If you have a card, try removing it and starting the CPU without it. It will come up in STOP mode and with a fault, but if it powers on, you know the card is the problem. Then, you need to reformat the card and reload the firmware and program.
Looking back, I should have learned step 4 earlier. At the time, I assumed it was always a software or power issue. But in a dusty, vibrating cabinet on a factory floor in Mexico, the physical connection is the most likely to fail. Save yourself the 80-dollar rush order for a new power supply. Check the bus connector first.