I’m not an engineer. I’m the office administrator who managed purchasing for a 160-person manufacturing company—everything from toner to torque wrenches. About $800,000 a year across 60 vendors. I report to operations and finance, and I’ve learned more about industrial sensors than I ever planned to.
Here’s how I learned the hard way: In January 2025, I ordered what looked like a straightforward replacement—a Dwyer differential pressure switch for a dryer filter. The quote was $120. The total cost, after wrong deadband, overnight shipping, a second technician visit, and a rejected invoice? Around $700. And the switch was fine. The deadband was wrong.
If you’ve ever been asked to “just order the part,” this one’s for you.
It Looks Like a Simple Replacement
Surface-level, a replacement part order is easy: find the model number, find the cheapest price, hit submit. That’s how I operated in my first year. Like most buyers, I focused on the product name and the price and missed the specification sheet.
Most buyers focus on the model number and unit cost and completely miss the details that determine whether the part works in their system: range, deadband, output signal, electrical rating, port size, and media compatibility. The question everyone asks is “what’s your best price?” The question they should ask is “what’s included in that price?” (and sometimes, “what’s the deadband?”).
Example: a Dwyer static pressure sensor isn’t one product. It’s a family of products. You can choose a low-range unit for cleanroom monitoring or a higher-range unit for duct static pressure, and the output can be 4–20 mA, 0–10 VDC, or even a digital bus. If the old sensor output is 0–10 VDC and you order a 4–20 mA version because it was “in stock,” you’re not replacing a part—you’re starting a project.
The Real Reasons These Orders Go Wrong
Specs are a language, not a label
I remember ordering a Dwyer static pressure sensor with a range I thought was the same as the old one. Turned out the old one was range 0–5″ WC, and the replacement was 0–10″ WC. It didn’t give a meaningful reading until I noticed the numbers looked low. In my defense, the model numbers were identical except for one letter. (Note to self: check every character.)
With a Dwyer differential pressure switch, the spec that catches people is deadband. The machine needs a switch that trips at 1.1″ WC and resets at 0.6″ WC, but if yours has a 0.2″ deadband, it can cycle too fast and wear out the contact. Or it can keep a signal that tells the controller the filter is clean. The switch isn’t bad—the specification is wrong.
The part often isn’t the root cause
Another reason replacements fail: the component didn’t fail. The system did.
Take a simple continuity test. I keep multimeter probes in the maintenance drawer, and I use them more than I ever expected. When a pressure switch looks dead, it’s tempting to replace it. But if you put the multimeter probes across the contacts and check for continuity while applying pressure, you might find the switch is fine. The problem could be a blocked impulse line or a broken wire upstream.
Same with proximity switches. If maintenance says they need an inductive sensor bi4u, that model has a defined sensing distance and a specific housing style. But it only switches when it sees conductive metal at the right distance. If the target has a thin layer of paint or is vibrating outside the detection zone, a brand-new sensor won’t work either. The sensor was never the problem.
The “same as before” myth
There’s a common belief that the safest order is an exact model-number match. That was more reliable 20 years ago, when product lines stayed the same for decades. Today, manufacturers revise products, change wiring colors, and phase out parts. The “same as before” thinking comes from an era when a paper catalog was the only reference. That’s changed.
I’ve seen a manufacturer ship a “same model” sensor with a different connector style, and nobody noticed until the technician was at the panel with a connector that didn’t match. The part itself was fine. The mismatch was invisible in a purchase order.
Communication gaps and the importance of asking
I said “like the one on the dryer.” The supplier heard “same Dwyer differential pressure switch on the dryer” and sold me a clean-filter switch with a different set point. I didn’t realize until later that the dryer had two switches: one for high differential pressure and one for low-pressure interlock. They looked identical from the outside. (I really should have asked for a photo of the nameplate before ordering.)
These mistakes aren’t about being careless. They’re about using the wrong level of detail in the request.
What These Mistakes Actually Cost
Let’s talk about the real cost, because that’s what I’m accountable for at the end of the month. A $120 switch can turn into an $800 line item after:
- Wrong part return shipping and restocking
- Expedited shipping on the correct part
- A second technician visit
- Production downtime while the machine sits idle
- Accounting time on rejected invoices and budget variance calls
The unit price is the visible part. The hidden costs are mostly time and trust. A 4-hour line stop in a mid-size plant can easily cost more than the instrument itself—maybe a lot more, depending on what the line makes. I don’t have an exact number because we rarely calculate it correctly. That’s the point.
From my perspective, the single most dangerous word in industrial purchasing is “cheap.” The vendor who quoted $75 less on a pressure switch might be the same vendor who can’t identify the correct deadband, can’t invoice properly, and ships from a warehouse two states away because it wasn’t stock. The lowest quote is not the lowest total cost.
A Better Way (It’s Not About Magically Lower Prices)
I don’t say this because I work for a vendor. I work for a company that buys lots of parts from lots of vendors. What changed for me was learning to calculate total cost of ownership before comparing quotes.
Start with the manufacturer’s documentation. Dwyer publishes spec sheets, installation manuals, and model-number references on dwyer-inst.com (accessed January 2025). It takes 10 minutes and saves hours. For the Dwyer static pressure sensor or Dwyer differential pressure switch, write down every spec from the old device’s nameplate, not just the model number.
For troubleshooting, don’t replace parts to see if that fixes it. If you need to know how to troubleshoot a Rice Lake load cell, start with Rice Lake’s own troubleshooting guide. The most common issue is not a failed cell—it’s bad wiring, moisture in the junction box, or a damaged cable shield. Check input resistance, output resistance, and cable continuity before ordering a replacement. Keep a good multimeter and a set of multimeter probes handy—they’re a lot cheaper than a load cell.
And if you’re chasing a sensor problem, verify the basics: supply voltage, sensing distance, target material, and the actual wiring. With an inductive sensor bi4u, the sensing range is usually short. The target has to be a conductive metal, within range, and not masked by paint or oil. I’ve seen a dirty target cause a false “sensor failed” diagnosis more than once.
Finally, ask suppliers for the data sheet before you give them the PO. If they can’t send one, that’s a red flag. It almost always means the person on the other side is matching a description, not a spec.
The goal isn’t to become an expert. The goal is to stop turning every part replacement into a project.
This is the real cost problem. It’s not about the price on the page; it’s about the confidence that the part will work the first time. And that confidence comes from one thing: knowing exactly what you’re buying.
If you ask me, a $120 switch that works is cheap. A $90 switch that doesn’t is expensive, even before you count the “I told you so.”
