It was 2:47 on a Tuesday afternoon when the call came in. A food processing plant about 40 miles from my shop had a line down. Their Siemens S7-200 PLC was showing a solid fault light, and the maintenance manager on the other end of the phone sounded like he hadn't slept in two days.
"Can you tell me about the Siemens 200 PLC?" he asked. "Is it worth repairing at this age, or are we throwing money at a stone?"
I've been handling PLC repair and integration orders for eight years. In that time, I've personally made—and documented—23 significant mistakes. Combined, they cost roughly $40,000 in wasted budget. That's not a brag; it's context. Because the most expensive thing I've learned in this industry is that we replace far too many parts we've never actually diagnosed.
I grabbed my multimeter, threw it in the truck, and headed out.
Since he asked: the S7-200 is Siemens' compact controller line from the late '90s. It's been superseded by the S7-1200, but thousands are still running in the field. Datasheets and manuals are still freely available on Siemens' support portal. The honest answer to "is it worth repairing" depends entirely on what's wrong with it—which is exactly why the diagnostic step matters.
If you're new to this, a multimeter is your eyes for things you can't see. Voltage, current, resistance—it turns invisible electrical problems into readable numbers. For basic PLC troubleshooting, you need two settings: DC voltage for power supply checks and continuity for wiring checks. That's honestly enough for 80% of field diagnostics.
At the plant, I opened the dusty panel door and started with the supply voltage at the PLC terminals. The display read 19.8V DC. The S7-200's nameplate—and its datasheet, which I'd learned to read after my first year—calls for 24V DC. Under-voltage like that is enough to make a PLC act dead or behave erratically.
Then I re-tested with the I/O terminals disconnected. The voltage jumped to 23.9V. The supply itself was fine. Something on the load side was dragging it down.
I reconnected each input module one at a time, watching the multimeter display. Channel 4 dropped the voltage back below 20V instantly. I traced that channel to a solenoid valve on the filling line. Internally shorted, pulling about 2.1A on a circuit rated for 0.5A. That back-feed was dragging the PLC's power rail down and making the S7-200 look dead.
The fix was a $14 solenoid valve and 20 minutes of wire nuts. The plant was back online that afternoon.
What most people don't realize is that a fault light is a diagnostic prompt, not a death sentence. It's the PLC's way of saying "look at me." But when production is down, it's tempting to read it as "replace me."
The manager kept apologizing for "wasting my time" on a simple issue. I told him not to apologize. Because for me, learning this lesson cost $795 and an embarrassing conversation. He just got it for free.
Back in 2017—my first year—I was called to a metal fabrication shop with the same symptom: an S7-200 with a blinking fault light. I was young, eager, and profoundly overconfident. I didn't check the datasheet. I didn't even open my multimeter. I looked at that fault light, crossed my arms, and announced, "Yeah, that PLC's toast."
I ordered a replacement—a refurbished S7-200 CPU at $585, plus $210 for overnight shipping. It arrived at 9 AM. I swapped it in, powered it up, and the fault light blinked right back on.
I don't remember exactly where I checked the 24V supply after that, but I remember hoping nobody was watching. It was outputting 11.6V under load. A $110 power supply would have fixed the entire problem.
The client spent $795 on a PLC they didn't need and two extra days of downtime. I spent the next year double-checking every diagnosis and reading every datasheet before I opened my mouth. That's the real cost of skipping the first step: it takes longer to earn back trust than to lose it.
Once you see this pattern, you spot it everywhere.
A warehouse client called about a forklift truck battery charger that kept tripping breakers. The manufacturer's quote to replace it: $2,800. The maintenance manager was already filling out the purchase order when I asked, "Mind if I poke around with a multimeter first?" Twenty minutes later, I found a corroded input lug. High-resistance connection, pulling excessive current, tripping the breaker. A wire brush and some dielectric grease fixed it. That charger is still running today.
Or the Maytag stove control panel job. A friend's oven was flashing an F9 code. The appliance shop quoted $450 for a new control board. Same routine: multimeter out, traced the circuit, found a cracked solder joint on the relay. A $4 replacement relay and fifteen minutes of soldering—the oven's worked ever since.
I'm not a genius. I just learned the hard way that total cost of ownership includes the time you spend diagnosing before you replace. And that's the thing most buyers miss.
Most buyers focus on the price of the part. They completely miss the costs that come with skipping the diagnostic step:
Add it up and a "$585 replacement" becomes a $3,200 mistake very quickly. If you ask me, the least expensive tool in any technician's bag isn't the multimeter—it's the restraint to use it before you spend someone else's money.
After the 2017 embarrassment, I wrote our team's pre-replacement checklist. It's not sophisticated. It's the stuff you should do before you call a parts vendor:
Measure the supply voltage at the device terminals, under load. Check the datasheet for rated tolerances—Siemens publishes complete datasheets for every PLC model, including legacy S7-200s, free on their support portal. Inspect all connectors and lugs for corrosion or looseness. Isolate I/O circuits one by one to catch back-feeding loads. Only after all that, consider replacing the device.
I'm not 100% sure why we so often skip these steps when production is down. My best guess is that urgency short-circuits logic. But in the 18 months since we rolled out this checklist, we've caught 47 errors that would've led to unnecessary replacements. Forty-seven times a client didn't pay for hardware they didn't need.
That's about $40,000 in avoided costs. Which is, I'm well aware, almost exactly what my education cost me.
So to the manager who called about the S7-200: no, I don't think your controller is a brick. It was a victim of bad power and a shorted valve. The repair cost less than the battery pack in your cordless drill.
And to anyone just starting out: learn your multimeter before you learn your PLC catalogs. The cheapest part in any repair is the one you don't buy. The best repair is the one you don't need. And the most expensive sentence in this industry—the one I've now heard myself say—is "I'm pretty sure it's broken."
Check first. Replace second. Your budget—and your reputation—will thank you.