EEVblog - uCurrent article

[Bob Engelhardt]

I hope that this is the right forum - it isn't actually about an EEVblog, but the article link is there.  And I hope that it hasn't been discussed before, but I couldn't find it.

I think that that switch SW1A is superfluous.  It's common is connected to IN+ for all settings of the switch.  IOW, a wire from IN+ to R12 would serve the same function.

It seems to me that it would be better not running the sense signal through switch contacts and a single pole switch would be cheaper.

Am I missing something?

Bob

[grenert]

If you didn't have that switch there, and all three lines were just connected to R12, then all three resistors R1, R2 and R8 would always be in the circuit, regardless of the setting of SW1B, effectively turning everything into a mA setting (since the resistance of R1 will dominate in parallel).

[amspire]

I think that that switch SW1A is superfluous.  It's common is connected to IN+ for all settings of the switch.  IOW, a wire from IN+ to R12 would serve the same function.

It seems to me that it would be better not running the sense signal through switch contacts and a single pole switch would be cheaper.

Am I missing something?

Bob

Yes, you are missing something.

A switch can easily have a contact resistance of 20 milliOhms If you try and switch in a 10 milliOhm current sense resistor with a 20 milliOhm contact resistance switch, then you could end up with no accuracy at all. The uCurrent could read high by 200% with your suggested wiring as you would be measuring the voltage across the sense resistor plus the SW1B contact resistance. Even the best quality switches only offer about 1 milliOhm contact resistance and that would give 10% accuracy on the maximum current range.

Sw1A ensures that the uCurrent sees only the voltage across the current sense resistor in use, and so the effect of the switch SW1B contact resistance is eliminated.

Richard.

[ejeffrey]

Nope.  The point is that no current flows through SW1A, so the contact resistance is unimportant.  The full sense current flows through SW1B, and thus it has some voltage drop due to contact resistance.  If you omitted SW1A, the sense lead would see the voltage drop across SW1B in series with the sense resistor.  It would cause a serious error on the milliamps range, a non-negligible error on the microamps range, and no error on the nanoamps.

[Bob Engelhardt]

If you didn't have that switch there, and all three lines were just connected to R12, then all three resistors R1, R2 and R8 would always be in the circuit, regardless of the setting of SW1B, effectively turning everything into a mA setting (since the resistance of R1 will dominate in parallel).

Actually, no.  The 3 shunt resistors would still be switched by SW1B.

[Bob Engelhardt]

Yes, you are missing something.

...you would be measuring the voltage across the sense resistor plus the SW1B contact resistance. ....

Sw1A ensures that the uCurrent sees only the voltage across the current sense resistor in use, ...

Riiigght!    Thanks, Bob