Why we obsess over dielectric testing for our IR sensors
We test every single IR sensor lamp that leaves our shop. Every one. We run them through withstand voltage and insulation resistance checks because, in wafer fab, a tiny electrical leak is a nightmare. It doesn’t just break the sensor—it can trash your entire batch of silicon.
Finding the breaking point
Here’s how we do it: we push these components well past their normal operating voltage. We’re basically looking for the breaking point. By stressing the insulation, we make sure the material can handle random voltage spikes without arcing. If a lamp fails, it usually means there’s a microscopic crack in the quartz or some junk in the seal. We don’t “fix” those. We toss them in the scrap bin immediately.
Why “sampling” doesn’t cut it
Some shops just test a few samples from a batch. That’s not enough when you’re dealing with high-end semiconductor gear. A sensor might look fine during a basic continuity check, but then it hits the thermal cycling of a real production line and just gives up. We measure leakage current down to the micro-amp. If the resistance dips even a little below our spec, the unit is out. It saves you from those annoying “ghost” signals or ground loops that make your temperature readings jump around for no reason.
The balancing act
Now, there’s a trade-off. Rigid testing is great for safety, but too much stress can wear out the components before they even reach you. We’ve spent a lot of time finding the sweet spot—voltage high enough to catch the defects, but not so high that it ages the insulation prematurely. But keep in mind: we can give you a perfect sensor, but you’ve still got to have your machine’s grounding wired up right. If your chassis ground is floating, even the best sensor in the world can’t stop electrical noise from hitting your wafer. We handle the strict QC on the insulation so the sensor isn’t the thing that fails in your vacuum chamber. We’ll give you the data; you get the peace of mind.