pcb resistors

[trevwhite]

Hi, as any one used a pcb track as a current sense resistor? I found a current sensor ic that recommended a track pattern that gave 150mOhm resistance with 5% tolerance. I believe the tolerance was for manufacturing rather than temperature changes.

I was wondering if such 'pcb resistors' remain within tolerance w.r.t. temperature?

Thanks

Trev

[PChi]

I have worked on product that used tinned copper wire not the PC track. It's a cheap idea. The temperature coefficient of change of resistance with temperature for copper is high. The data that I have is alpha = 0.00428 which I think means that it changes by 0.4% / degree. Also over load trashes the track hence the PCB.

[Psi]

You can also get hall effect current IC's where you have a PCB trace running under the chip and the sensor inside picks up the current through the IC package. I don't think the track width matters much with those.

It's not very accurate though.

[trevwhite]

Hi

Thanks for the reply. It doesn't sound the best idea really. Maybe for detecting some current it would be okay.

Trev

[Psi]

Yeah, it's no good if you need accuracy. 

[ejeffrey]

Track resistors can be useful for things like current sharing resistors but they aren't very good for measurement shunts unless you really don't care about accuracy.

[Zero999]

I agree, it's no good for accuracy but the question is why are you measuring the current in the first place?

Is it for overload protection? If so, a PCB trace may be ideal, as copper has a positive temperature coefficient so the reading will increase with the temperature, which might be good if the sense trace is near the device being protected.