Industrial Instrumentation: A Procurement Guy's Guide to Picking the Right Tool (And Avoiding My Mistakes)

There's No "Best" Instrument — Only the Right One for Your Scenario

In my first year handling instrumentation orders (that was 2017), I made a $3,400 mistake that still stings. I ordered 45 differential pressure gauges for a filtration upgrade. Specs looked perfect on paper. Three weeks later, installers called me from the site: the gauges couldn't handle the static pressure. Wrong application entirely. We returned 42 of them (three were already installed and ruined). Restocking fees plus the rush reorder cost us $2,900 in pure waste. Not to mention the 4-day production delay.

That's when I started keeping a running document of every instrument mistake I've made or witnessed. It's now 14 pages long. What I learned: there is no universally "best" instrument — only the right instrument for your specific scenario.

Here's how I now categorize industrial measurement problems, with the actual lessons baked in.

The Four Scenarios I See Most Often

Based on our order history since 2017 — roughly 2,300 instrument orders — most measurement needs fall into one of these buckets:

1. You need to monitor a process pressure (general industrial)
2. You need to measure differential pressure (filtration, flow, level)
3. You need vacuum measurement (negative pressure, suction systems)
4. You need position or level feedback (tank level, linear displacement)

And then there's the outlier: thermal imaging questions, which come up more than you'd think.

Scenario A: "I just need a pressure gauge"

If you're monitoring basic process pressure — air, water, hydraulic fluid, steam — this should be the simplest order. It rarely is.

My third-worst mistake: ordering a gauge with a 2.5" dial for a panel where operators read it from 6 feet away. Customer complained the numbers were "unreadable." Fair point. We swapped all 12 units at our cost.

What I've learned to specify:

• Dial size: 2.5" for close-range panel display. 4.5" for anything over 3 feet. 6" for control room visibility.
• Connection: Bottom mount vs. back mount. Get this wrong and you're re-plumbing.
• Wetted materials: If you're measuring anything corrosive and don't know the wetted parts, stop and ask. I've seen 316SS gauges destroyed in 48 hours by an acidic process stream.
• Range: The rule I use — operating pressure should fall in the middle third of the gauge range. If your process runs at 90 PSI, don't spec a 100 PSI gauge. Spec a 300 PSI gauge. The 100 PSI unit will cycle near the top of its range and die early.

Ashcroft's general line covers most of these needs. But here's what I tell my team: the brand is less important than the spec match. A perfectly specified no-name gauge beats a badly specified name-brand one every time.

Scenario B: "It's a differential pressure application"

This is where I see the most errors. Differential pressure (DP) is not just "two pressure gauges glued together." The 2017 filtration disaster taught me that.

When you actually need a differential pressure gauge:

• Filter monitoring (clogging indicator)
• Flow measurement (through orifice plates)
• Liquid level in closed tanks (hydrostatic)
• Clean room pressure differentials

My second big lesson: static pressure matters as much as differential pressure. A DP gauge rated for 100" H2O differential may still fail if your system static pressure is 400 PSI. You need to check both ratings.

For Ashcroft specifically — their differential pressure instruments have published static pressure limits. Read them. I didn't, once. That $3,400 mistake? Static pressure over-range. The gauges were permanently deformed after 6 hours.

Also worth noting: for filtration applications, a differential pressure gauge with a switch output is usually what you actually want. That way you get an alarm when the filter clogs, not just a dial reading that nobody watches.

Scenario C: "I need to measure vacuum"

Here's where a lot of buyers get tripped up. There are multiple "truths" about vacuum measurement:

Truth 1: "Vacuum" means different things. Rough vacuum (760 to 1 Torr), medium vacuum (1 to 10⁻³ Torr), high vacuum (< 10⁻³ Torr). These are not the same instruments.

Truth 2: A standard pressure gauge rated for "vacuum to 30 PSI" is not a vacuum gauge. It's a compound gauge that happens to read vacuum. For actual vacuum process control, you need a dedicated instrument.

We had a customer order what they called "ashcroft vacuum pressure gauges" for a resin degassing system. They ordered compound gauges. The process ran at 50 mTorr. Compound gauges are useless below about 20 Torr. They found out after the batch failed.

What I ask now before any vacuum order:

1. What's your target vacuum level? (Get a number, not "high vacuum")
2. Is the process gas clean or contaminated? (Contamination kills certain sensor types)
3. Do you need a gauge to look at, or a signal for your control system? (These are different products)

If you can't answer #1 with a specific number, you're not ready to order.

Scenario D: "I need position or level feedback"

This is a different beast entirely. Pressure and vacuum are about fluids. This is about tracking where something is.

For liquid level in tanks: Level sensors come in many types — float, ultrasonic, radar, capacitive, hydrostatic. Each has failure modes I've personally witnessed:

• Ultrasonic: failed when foam formed on the liquid surface
• Capacitive: drifted when humidity changed
• Float: stuck when the product crystallized on the stem

If your process is "clean liquid, stable temperature, no foam" — a simple float switch works. If any of those conditions are false, you need to spend more.

For linear position measurement: This is where wire draw absolute encoders come in. I'll be honest — I'm not an encoder specialist. What I can tell you from a procurement perspective:

• They're used for measuring linear travel (like a piston position or a gate opening)
• "Absolute" means the position is known immediately on power-up — not "incremental" which requires homing
• The"wire draw" refers to a cable that extends and retracts

What I've seen go wrong: spec'ing an encoder for the wrong stroke length. Always add 10-15% to your actual measured travel. Cables don't like being fully extended at the limits.

This gets into motion control territory, which isn't my expertise. If you're integrating an encoder into a servo system, talk to the controls engineer before ordering. I've seen $1,200 encoders destroyed because the cable was routed wrong.

The Thermal Camera Question I Get Every Quarter

"Can thermal cameras see through glass?"

I hear this at least four times a year from maintenance teams. The answer is no, not typical glass. Standard glass — like window glass — blocks long-wave infrared radiation. That's how thermal cameras see.

FLIR (the brand most people ask about) publishes this clearly: you cannot use a standard thermal camera through a glass window. The glass itself appears as a uniform temperature (usually reflecting the camera's own temperature), and you can't see what's behind it.

Exceptions: certain specialized IR-transmissive materials like germanium or zinc selenide. These are used in industrial thermal windows. But they're not "glass" in the normal sense.

What this means practically: if you need to monitor a process inside a glass-lined reactor or through a sight glass, standard thermal imaging won't work. You need either a specialized IR window or a different measurement approach entirely.

My mixed feelings on thermal cameras for industrial use: they're incredible for finding hot spots, insulation gaps, and electrical issues. But they're often oversold for process monitoring. If you need continuous temperature measurement of a specific point, a thermocouple or RTD is cheaper and more reliable. Thermal cameras shine for surveys and diagnostics, not for permanent installation.

How to Figure Out Which Scenario You're In

Three questions. Answer them before you order anything:

1. What am I actually measuring?

• Pressure (positive, gauge, or compound)
• Differential pressure
• Vacuum (and at what level)
• Liquid level
• Linear position
• Temperature (surface, spot, or area)

2. What's the environment?

• Temperature range (ambient and process)
• Corrosive or clean
• Vibration present?
• Indoor or outdoor

3. What happens after the measurement?

• Human reads a dial → basic gauge is fine
• Signal to control system → need transmitter or switch output
• Alarm condition → need switch, not just gauge
• Data logging → need digital output

I've created a one-page pre-order checklist with these questions. Since implementing it in Q1 2024, we've caught 31 specification errors before they became purchase orders. Estimated savings: somewhere around $18,000. I wish I had hard data on the exact number — I started tracking late — but even the conservative estimate makes the checklist worth the five minutes it takes to fill out.

The single biggest lesson from seven years of instrument orders: five minutes of verification before ordering beats five weeks of returns and rework. Every time. No exceptions.

I still kick myself for that 2017 filtration mistake. If I'd just asked one question — "What's the static pressure?" — I would have saved $2,900 and a lot of embarrassment. Don't learn that lesson the way I did.