Mitsubishi PLC Training Taught Me About Outputs. A $16 Relay Taught Me About Procurement

It wasn't a Mitsubishi PLC that failed. It was a $16.20 relay, and it was my fault.

I manage procurement for a 36-person control panel shop. We spend roughly $1.1 million a year on automation components. I've kept every order in a cost tracking spreadsheet for over six years. The invoice for that relay is still in the system. The cost of the mistake is in a different column.

A while back, our lead engineer suggested I sit in on Mitsubishi PLC training at our distributor. I said no at first. I buy parts, I don't write ladder logic. He insisted. Good thing he did. The trainer said something that stuck with me:

A PLC output is a switch, not a power source. If you connect an inductive load like a relay coil, use a suppressor or pick a relay with one built in.

I heard those words, and then I proceeded to ignore them when it mattered.

The Relay That Looked Fine

Last year we had an order for a series of pump panels. The panel design was straightforward. It used a Mitsubishi PLC, a 24 VDC DPDT relay in an 8-pin socket, and a contactor for the motor circuit. Nothing exotic. The relay was a name-brand component, but a supplier offered a private-label equivalent at a lower price.

I want to say the quote was around 30% lower, but don't quote me on the exact percentage. The point is it was enough for me to look at the datasheet.

I did look at it. I compared coil voltage: 24 VDC, same. Contact arrangement: DPDT, same. Contact rating: 10 A, same. The relay looked like a drop-in replacement. I approved it for all 48 relays on the order.

Here's the thing: the differences that matter are rarely in the first five rows of a datasheet. They're further down, in the footnotes.

The original relay had built-in coil suppression. The private-label relay did not. The datasheet didn't hide it. It was listed under coil specs, right where it always is. I just wasn't reading that far.

Two Dead Outputs and One Expensive Lesson

Because the relay coil was wired directly to a transistor output on the Mitsubishi PLC, every time that output shut off, the coil without suppression sent a voltage spike back into the card. It didn't fail on the bench. It failed in the field, after hours of cycling.

First report: one output channel stopped working. Our engineer replaced the PLC output card, watched it for a day, and thought it was a one-off. It wasn't. Two more sites reported the same issue in the next week. That's when we compared the original relay spec to the private-label spec line by line.

By the time we finished, we'd replaced all 48 relays, paid for two emergency service visits, sent replacement output cards, and put a technician on overtime to re-test the panels. Total impact was a little over $5,700. Maybe $5,200? No, $5,700. I have the charge code in the spreadsheet.

The part savings on the original order were less than $500.

The frustrating part is that this wasn't an unusual component. It was a 24 VDC relay. I had bought hundreds of similar relays over the years. My confidence is what skipped the footnote.

I'm not saying the supplier tricked us. Their datasheet was accurate. I compared a small slice of the specification and called it done. That's a sourcing failure, not a vendor failure.

The Checklist I Use Now

The fix wasn't banning private-label relays. It was making our review process slower on purpose before the PO, not in the field.

Relay specifications now get read like a contract. Not a suggestion. When I review one, I check:

• Coil voltage and coil type: AC or DC, plus pickup and dropout values if available
• Coil suppression: none, built-in diode, or RC network
• Contact arrangement and pinout, not just DPDT
• Contact rating at the actual load type. 24 VDC and 230 VAC are not interchangeable
• Minimum switching load or dry circuit limitation if the relay handles low current signals

I also ask for a sample and put it in a test panel before we buy in volume. If a component is a real drop-in replacement, the test takes an hour and costs nothing. If it isn't, the test just saved the entire project.

That same checklist habit changed our contactor sourcing too. A contactor is not a big relay; it has its own rules. But the discipline is identical. We don't accept a quote that says just 24 V anymore. It must say 24 VDC or 24 VAC, and for motor loads it must include the AC-3 current rating, not a general current number. One near-miss later, I can tell you that's not over-engineering. It's procurement prevention.

OEM vs Private Label Relays Is the Wrong Question

If you search for relay OEM vs private label, you'll get a lot of opinions. My opinion after this project is simple: the question is not whether the relay is private label. The question is whether the exact relay you're being quoted does what the drawing requires.

Private-label relays can be perfectly fine. At least, that's been my experience when they're specified by people who read the full datasheet and not the cross-reference guide. But equivalent is a sales word, not a specification. OEM and private label tell you who branded it. They don't tell you if the coil suppression exists, what contact material was used, or how it behaves at low current.

The Mitsubishi PLC training I almost skipped taught me the theory. The relay taught me the practice. It took one order, 48 relays, and more money than I care to repeat to understand that 10 minutes of specification checking at the beginning is cheaper than any rush fix at the end.

Don't let a supplier's cross-reference sheet do your thinking. Check the current datasheet, verify a sample, then buy. If you're in a hurry, that's exactly when the $16 relay becomes a $5,700 problem.