A Three Winding Power Transformer, a Friday Deadline, and the $4,800 That Saved the Project

The Friday Folder

On January 26, 2024, at 4:47 PM—I remember the time because I was already packing up—our project manager dropped a spec folder on my desk. We'd just won a controls upgrade for a regional food processing plant. The client wanted their legacy control system retired before the spring production cycle. Deadline: February 28.

I handle quality and brand compliance at Delta. I review everything that leaves our building before it reaches a customer—roughly 200+ projects a year. Transformer specs land in my inbox constantly, and this one showed up at the end of a Friday.

I should've waited until Monday. Instead, I opened it.

What the Spec Sheet Left Out

The BOM looked clean at first. A current step down transformer for a 4–20 mA instrument loop. A 3 phase buck transformer to knock the incoming 480V down to 240V for a motor control panel. A replacement power supply transformer for the 24V DC rail. Nothing unusual—except the plant's drawings contradicted themselves in three places.

I called the site electrician Saturday morning. That's where the real picture came together. The customer also needed a low voltage to high voltage converter for a legacy test bench and—this is the part nobody had flagged—a three winding power transformer to feed both the new panel and a future expansion they hadn't told us about yet.

The original BOM assumed the old three-winding unit could be reused. Nobody had verified its nameplate. When the electrician finally got a photo of it, the ratings were gone—worn off sometime in the last decade.

So we needed a three-winding replacement, plus a step up down converter to handle their test bench voltage swings. That pushed the transformer scope from three units to five.

Simple, right? It wasn't.

Four Quotes, Two Philosophies

We had four pre-qualified suppliers. I sent out updated specs on Monday.

Two quotes came back inside 48 hours. The other two took four days and kept asking clarifying questions about the winding configuration.

Here's where I'll be blunt: the fast responses were cheaper. By a lot. Roughly 22% less on the full package, with a lead time quoted as "3–4 weeks, roughly."

The slower suppliers gave us a 10-day lead time—guaranteed—for the same order.

The numbers said go with the cheap option. My gut said something felt off about the responsiveness. I kept asking myself: is 22% worth potentially missing February 28?

Worst case: roughly $4,200 in rush fees plus a client who never calls us again. Best case: $3,100 saved and a marginally slower ship date.

The expected value math favored cheap. The downside felt catastrophic.

The 11:00 PM Email

We went with the cheaper supplier. Partly on price, partly on the optimistic assumption that "3–4 weeks, roughly" would land on the early side.

Two weeks later—February 12, 11:04 PM—I got an email. Their copper supplier had pushed a shipment. Our three-winding unit would be delayed. New ship date: March 4.

We'd miss the deadline by four days.

For context: the client's spring production cycle starts March 1. Every day of downtime past that runs them about $6,000 in lost throughput. Missing by four days isn't a four-day problem—it's a $24,000 problem, plus the trust hit.

I spent the next morning on the phone with the other supplier—the one we'd passed over. They had our specs on file. They could start winding immediately. Lead time: 9 days, guaranteed. Price: $4,180 more than what we'd signed with the cheap vendor.

I didn't need to run the math twice. We cancelled the first PO, ate the cancellation fee, and placed the second order.

If I'm being honest, the $4,180 hurt. Budgets are real. But the alternative was a conversation with the client that I did not want to have.

What Actually Shipped

The second supplier delivered on February 21. Two days ahead of their own quote. Every unit passed incoming inspection—winding resistance, turns ratio, insulation, all within tolerance. IEC 60076 lays out the standard testing regime for power transformers, and every one of those tests came back clean. I signed off on the whole batch in under an hour because there was nothing to argue about.

The client cut over to the new system on February 26. Two days of buffer. Production started on schedule March 1.

Total cost of the certainty premium: $4,180. Plus a $600 cancellation fee on the first PO. Call it $4,800 against a project that would have cost us roughly $24,000 in direct losses—plus a relationship that's hard to price.

I'll take that trade every time.

What I Learned About "Probably On Time"

People think rush premiums exist because suppliers charge more when they can. That's backwards. Suppliers who consistently hit dates can charge more because they've built the systems that hit dates. The causation runs the other way.

The cheap quote isn't cheap because they're efficient. It's cheap because they're passing the uncertainty onto you.

Three things I do differently now:

1. I ask suppliers what happens on their end when a shipment slips. Not what happens to us—what happens to them. If the answer is vague, they've never planned for it.
2. I budget the certainty premium upfront. For any project with a hard external deadline, I assume we'll pay 10–15% more for guaranteed timing, and I price it into the quote.
3. I don't treat "3–4 weeks, roughly" as a date. It's a wish. Get a real date or walk.

One more thing—if your project involves a three winding power transformer, the winding configuration alone makes lead times less predictable. Copper availability, core lamination schedules, tap changer assembly—any of these can slip. That's not a reason to avoid three-winding designs. It's a reason to pay for suppliers who've already built schedule buffers into their own operations.

The Part I'd Do Differently

I should've run the reverse scenario on day one. Not "what do we save if this works?" but "what do we lose if it doesn't?"

The first number was $3,100. The second was $24,000. One of those is a rounding error on the project P&L. The other isn't.

If you're speccing a low voltage to high voltage converter, a step up down converter, or any transformer with a long assembly lead time—here's the takeaway: the quote that saves you 20% isn't the cheapest one. The cheapest one is the quote that delivers on the date you actually need it.

Everything else is just an option on a headache.