What Really Defines a “Good” Resistor in Precision Current Sensing Applications?

When we talk about a “good” resistor, most of us probably start with tolerance and power rating.

But in real current sensing designs, especially at higher currents, I’ve found those specs don’t always tell the whole story.

For example:

• A low tolerance part can still drift under load if the TCR isn’t well controlled.
• Thermal gradients across the resistor body can shift readings more than expected.
• Long-term stability under continuous current stress sometimes matters more than initial accuracy.
• At low current ranges, even small thermal EMF effects can start to show up.

On paper, many parts look perfectly fine. In actual hardware, the behavior can be quite different once temperature rise and layout effects come into play.

So I’m curious how others approach this:

• When selecting resistors for precision current sensing, what’s your top priority?
• Do you focus more on TCR, construction type, or long-term drift?
• Have you ever had a “good spec” resistor that didn’t behave well in the real system?

Interested to hear real-world experiences from others working on similar designs.