The Short Answers First
If you typed "sick" into a medical supply search bar hoping for a fever thermometer, you're in the wrong aisle. SICK is a German industrial sensor manufacturer — photoelectric sensors, encoders, safety systems. Nothing to do with feeling unwell. (The brand name collision has confused our purchasing inbox at least once a quarter since I started in 2017.)
Here's what I actually know after eight years of specifying and ordering this equipment — and after roughly $4,200 in personal ordering mistakes:
- SICK wire draw encoders are excellent for linear measurement over long distances, but they'll bite you on installation orientation and cable routing if you don't plan ahead.
- Fluke 116 vs 117: The 116 is built for HVAC work (temperature, microamps for flame sensors). The 117 is a general-purpose electrician's meter with non-contact voltage detection. Pick wrong, and you'll be borrowing someone else's meter by week two.
- Thermal cameras and glass: No, FLIR thermal cameras cannot see through standard glass. Not a limitation you can firmware-update your way around. Physics says no.
Everything below is the supporting evidence — the projects, the dollar amounts, and the checklists I now keep pinned to my desk.
Who Am I and Why Should You Care
I handle industrial measurement and sensor orders for a systems integration team. Over eight years, I've processed maybe 600 orders — or rather, closer to 550 once you exclude returns and re-orders. I've personally made and documented enough significant mistakes that our team now runs a pre-purchase checklist I wrote. We've caught 47 potential errors using it in the past 18 months.
Not bragging. Just context.
The wire draw encoder mistake happened in March 2022. We needed linear position feedback on a 12-meter conveyor retraction system. I ordered a SICK wire draw encoder with a 10-meter measuring cable because the project spec said "12m travel." Wired it up, powered it on, and the readings drifted past 10 meters like they were on vacation. The encoder itself was flawless — I just didn't account for the cable length needed for mounting and routing. That error cost $890 in expedited re-order plus a 6-day production delay.
Looking back, I should have read the full installation diagram instead of just the specs table. At the time, I assumed '10m measuring range' meant 10m of usable travel. It doesn't. It means 10m of cable. There's a difference.
The SICK Wire Draw Encoder Problem Nobody Warns You About
Here's the counterintuitive part: SICK wire draw encoders are genuinely reliable — that's why I keep specifying them. The drum mechanism, the spring tension, the IP67 sealing. All solid. But the failure mode isn't in the encoder. It's in how you mount it.
If you mount the unit at an angle where the cable doesn't pay out straight, you'll get measurement drift within weeks. Not because the encoder is bad, but because the cable develops memory curls. We had a unit on a vertical lift application where the cable ran at about 15 degrees off-axis. Twelve months in, readings started wandering ±3mm. On a 4-meter travel, that's fine. On a 200mm positioning window, that's a reject.
Lesson: mount in line with the travel path. If you can't, use a guide pulley. The $120 pulley saves you a $900 encoder replacement plus downtime.
Even after choosing the right model number, I kept second-guessing. What if the cable tension wasn't right? What if the ambient temperature in that warehouse (unheated, this was January) threw off the readings? The three weeks until the first calibration check were stressful. It passed. But I've had ones that didn't.
Fluke 116 vs 117: What the Spec Sheet Doesn't Tell You
Why does this matter? Because I've watched two junior techs order the wrong one in the same quarter.
The Fluke 116 is purpose-built for HVAC diagnostics. It does temperature (with the included thermocouple), microamps (critical for testing flame sensors on furnaces), and standard electrical measurements. If you're doing boiler commissioning or rooftop unit troubleshooting, this is your meter.
The Fluke 117 is marketed as an "electrician's multimeter." Non-contact voltage detection, auto-volt mode, low-input impedance to eliminate ghost voltages. It's designed for commercial electrical work — checking panels, tracing circuits, verifying dead lines before maintenance.
Here's the trap: the 117 has temperature capability if you buy the optional probe. But it doesn't come with one. I once ordered three 117s for a team that primarily did HVAC service calls. When the first tech went to read a supply air temperature, he looked at the meter, looked at the missing thermocouple port cover, and called me. Not his fault. Mine. Those three meters sat in the cabinet while we ordered 116s anyway. Wasted about $450 on the 117s — they eventually found a home with the electrical crew, but that was six months later.
If you're doing a mix of both: buy the 116 for HVAC work and have a dedicated electrical meter. Don't try to make one meter do everything — you'll end up with a $200 compromise.
Thermal Cameras Through Glass: The Experiment That Cost Me $1,200
I didn't fully understand the value of detailed specifications until a $1,200 order came back completely wrong. Not wrong from the vendor — wrong from me.
We had a client who wanted to inspect electrical panels through a machine enclosure window. Tempered glass, 8mm thick. I specified a FLIR thermal camera thinking "thermal imaging sees heat, heat passes through glass eventually, right?"
No.
Standard thermal cameras operate in the long-wave infrared spectrum (roughly 8–14 micrometers). Common glass is opaque in that band. The camera literally cannot see through it. What you get is a very accurate reading of the glass surface temperature — which tells you nothing about the panel inside. (Note to self: add "check material transmission" to the site survey template.)
The fixes exist — you can use a special IR-transmissive window (germanium or certain zinc selenide formulations), but those cost more than the camera itself in some configurations. Or you cut an inspection port, which defeats the purpose of the sealed enclosure. We ended up running the inspection when the line was down and the enclosure was open.
After the third rejection in Q1 2024 — a client asking why their thermal survey couldn't see through a glass viewing port — I created our pre-check list. It now has 22 items. The glass one is item #7.
What I'd Do Differently (And What I Still Won't Claim to Know)
Looking back, I should have paid for the SICK encoder cable guide pulley from the start. I should have ordered the Fluke 116s instead of trying to save $150 with the 117s. I should have read the thermal camera's spectral range specification before promising a through-glass inspection.
If I could redo those decisions with what I know now, I'd have saved about $2,500 out of the $4,200. The other $1,700? That's the cost of learning edge cases nobody documents.
Here's my honest boundary: I know SICK sensors well enough to specify them correctly 95% of the time. I know Fluke meters — both models — inside and out from an application standpoint. But I'm not a thermal physics expert. If your application involves unusual materials, high-temperature surfaces, or through-barrier measurements, talk to someone who does optical transmission specs for a living. That's not me.
I'd rather work with a specialist who knows their limits than a generalist who overpromises. And I'd rather tell you "this isn't my strength, here's who does it better" than sell you a camera that won't work through your existing enclosure window.
That's the difference between someone who reads spec sheets and someone who's actually paid for getting it wrong.
"The vendor who said 'this isn't our strength — here's who does it better' earned my trust for everything else."
Still learning. Still making checklists. Still occasionally googling "sick thermometer" and remembering why our purchasing inbox gets so many confused medical inquiries.
