Which shunt resistor placement is best?

I came across this picture in a LinkedIn post recently.

I understand the reason behind the top two pics. The tracks are too thin to handle the current. However, I do not understand the bottom two pictures.

I want to know the following:

1. Why is the bad way, considered wrong?
2. Practically what effects will the bad way cause in the measurements?
3. Why is the good way, considered correct?

Image source: LinkedIn - "BacktoBasics: Current Shunt/Sense Resistor PCB Layouts" by Amaldev Venugopal

Probably to distribute the current over all three shunt resistors more equally. Keep in mind that those little traces connecting the pads of the shunt resistors also have resistance.

In practice, the layout choice might have a lesser impact than in the one-shunt-case above. But if you can, it's a good idea to keep things symmetrical most of the time.

You're right about the small track on the top "wrong" layout being an issue. But beyond current handling capabilities of the track, the current that is forced through there will cause some inaccuracies in the measurement due to the IR drop over the track. The kelvin connections small tracks on the inside of the current sense pads are routed that way to minimize any parasitic resistance being added to the V=IR equation. Parasitic resistances come from the component pads, solder, traces. Placing the SENSE2 escape track there adds the track resistance to the V=IR equation.

For the bottom "wrong" one: There is one input and output of the 3 parallel resistors. The input is split into three currents and then summed back into the total current at the output node. This node should be the best spot for your measurement. It wouldnt matter in an ideal world where traces have 0 resistance but they do.

For illustration: 1A through 3 parallel 1ohm resistors, you'd expect 333.33mV across them.

Measuring at the summing node (where the load exits): very close to 333.33mV

Moving the load to the top of the parallel resistors (like the "wrong" layout): less close to 333.33mV

As you can see, the measurements are different but I would hardly classify these as "right" and "wrong". It's a matter of how accurate you need your measurement to be.