In January 2025, I was standing in a warehouse parking lot next to a newly delivered GE transformer, holding a list of small problems that had nothing to do with the transformer. The office manager was asking about a dead Asko dishwasher control panel. The IT guy was holding a beeping APC battery backup and surge protector. And I was about to wire the cooling fan circuit with the wrong component. It was going to be one of those weeks.
I've worked as an electrical controls engineer for about nine years, mostly with small and mid-sized industrial clients. I've personally made and documented 12 significant mistakes totaling roughly $40,000 in wasted budget. This project added three more lessons to that pile. If you're planning a similar GE transformer upgrade, you're probably thinking about the big pieces—the transformer, the relay, the protection functions. I'm here to tell you that the small pieces can bite just as hard.
The project: A GE transformer and a Multilin 850
The client was a family-owned food processor. Their old transformer was failing, and they needed capacity for a new freezer line. We spec'd a GE Prolec pad-mounted transformer, 750 kVA, with a GE Multilin 850 transformer protection relay. The 850 was the right choice because it gave us GE Multilin 850 transformer differential 87T plus everything else in one box.
I could have used a simpler overcurrent relay. But their old site had a history of cable faults that damaged the transformer, and they wanted more robust protection. The GE Multilin 850 transformer protection functions we configured included:
- 87T transformer differential, percentage restrained with harmonic blocking
- 50/51 phase and ground instantaneous and time overcurrent
- 50N/51N neutral overcurrent
- 27/59 undervoltage and overvoltage
- Temperature monitoring via RTD inputs
That's a lot of protection for a 750 kVA transformer. But redundant protection is not the same as excessive protection. It's the difference between a minor cable fault and a $30,000 replacement.
The dishwasher side trip
While the transformer was on order, the office manager asked for a “quick favor.” The break room Asko dishwasher was dead. The display was dark, and the unit wouldn't respond at all.
I pulled the control panel away from the door and found the problem almost immediately: water staining on the main board near a connector. The board wasn't the root cause—moisture was. I ordered an Asko dishwasher control panel replacement, sealed that connector with a moisture-resistant coating, and had the dishwasher running in about two hours.
Cost: about $180 for the panel, plus my time. That's nothing compared to the transformer line item. But for this client, it was exactly the kind of small issue that makes a large project feel less scary. I'm not in the business of judging requests by their dollar value. Small doesn't mean unimportant. It means trust.
The APC battery backup and surge protector
The same week, the IT guy pulled me into the server room. The rack was beeping. The old APC battery backup and surge protector had been installed in 2018. It wouldn't hold a charge.
I checked the battery first. It had swollen, which is a clear “replace me” signal. I swapped in a new APC battery backup and surge protector with similar output capacity, then wrote the install date on a label.
Here's what I didn't fully appreciate until then: A UPS combines a battery backup and surge protector in one unit, but those two halves age differently. The battery can die in three to five years. The surge suppression components have a finite life too. If you don't date the unit and schedule a check, you're running on luck. I should have caught this months earlier. That one's on me.
The difference between a contactor and a relay
Now the part that should have been simple.
The GE transformer's cooling fan control circuit was straightforward. A thermostat closes a contact. That contact is supposed to pick up a contactor. The contactor switches the fan motor.
I was in a hurry. I saw “relay output” in the wiring diagram and thought, “A relay is a contactor, right? Contacts are contacts.”
No. Wrong.
A relay is designed for control circuits. It can switch a small load like a signal, a pilot light, or the coil of a larger contactor. A contactor is designed to make and break power circuits, especially inductive loads like motors, with heavier contacts and arc suppression. Using a relay to switch a fan motor directly is a textbook mistake. The relay contacts can handle the running current, but they can't handle the locked-rotor inrush. They either weld closed or fail open.
Mine welded closed. The fan stayed on after the temperature dropped. I caught it because I was still on-site, reviewing the GE Multilin 850 settings. I opened the panel, saw the heat marks on the relay, and knew exactly what I'd done.
The replacement part was a definite-purpose contactor rated for the motor inrush. It cost $47 and took one day to get from a local supplier. The embarrassment cost more.
That's the difference between a contactor and a relay in practice: relay for control, contactor for power. Use both. Don't substitute one for the other.
The GE Multilin 850's output contacts are also pilot-duty contacts. They're meant to energize a contactor coil or send a trip signal to a breaker, not to switch a motor directly. The instruction manual says this. I didn't read it closely enough.
The 87T scare
Then came the relay setup. We had the GE Multilin 850 configured with transformer differential 87T. When we energized the transformer for the first time, the 850 displayed a differential current.
My first instinct was to disable the 87T element and “deal with it later.” That would have been irresponsible. Differential protection compares current entering the transformer with current leaving it. If the difference isn't zero, either the transformer is faulted or the secondary wiring is wrong. In my case, the wiring was wrong.
We checked the CT circuits and found one set of CT secondaries with reversed polarity. The 87T was doing exactly what it should have: it detected an imbalance that shouldn't exist. We corrected the CT leads, and the differential current dropped to near zero. IEEE C37.91, the standard guide for protecting power transformers, has a reason it spends pages on CT polarity. This is it.
The 850's 87T element also includes harmonic restraint, which stops it from tripping on magnetizing inrush. That's one of the GE Multilin 850 transformer protection functions that prevents nuisance trips. But no amount of relay intelligence can fix a CT polarity mistake.
If you're commissioning a transformer with 87T protection, check CT polarity before energization. It's a ten-minute test that saves you from a very confusing first startup.
What I'd do differently
Looking back, I should have done several things in a different order:
- Read the control component ratings before wiring, especially the relay vs contactor decision.
- Checked CT polarity for the 87T element before the first energization.
- Put UPS battery checks on the maintenance calendar instead of waiting for the beeping.
- Pulled the dishwasher control panel apart before ordering the replacement board. The root cause was moisture, and the board was only the symptom.
I didn't do all of that the first time. But that's why the checklist exists now. The list has grown because of projects like this one—the ones that expose what you didn't know before it cost you time, money, or credibility.
A GE transformer is a solid piece of equipment. A GE Multilin 850 gives you real protection if you set it up correctly. But the transformer and the relay won't save you from your own wiring mistakes. The side requests—the Asko dishwasher control panel, the APC battery backup and surge protector, the relay that should have been a contactor—those are the details that build or break trust.
Trust me on this one. I've got the burned contacts to prove it.
Leave a Reply