## Using traces as resistors

Posted on Tuesday, March 4th, 2014 in project logs by DP

NsN writes:

A while ago I needed some small value resistors as current shunts and I started wondering about the feasibility of using traces as resistors.
There are many theoretical reasons as to why copper traces make bad shunt resistors, but I found surprisingly little practical data.

The traces are (from left to right):
– 20 mil wide, 1221 mil long, should be ~30mOhms
– 6 mil wide, 611 mil long, ~50mOhms
– 12 mil wide 1221 mil long, 50 mOhms
– 6 mil wide 1221 mil long, 100 mOhms

Via the forum.

This entry was posted on Tuesday, March 4th, 2014 at 1:00 pm and is filed under project logs. You can follow any responses to this entry through the RSS 2.0 feed. You can skip to the end and leave a response. Pinging is currently not allowed.

### 4 Responses to “Using traces as resistors”

1. Chip says:

The thickness of the copper on the board is something you don’t have complete control over. However, you might be able to ‘trim’ the resistance after board manufacture to achieve some degree of accuracy. Or, it may be easier to merely have a wide range calibration for the circuit to adapt it to the variable resistance of the trace. Either way, mass production isn’t easily feasible as each board will need to be calibrated individually. If you do manage to have a reliable range of resistance, variations in different board lots will still make future manufacture require calibration.

and the calibration quickly goes out the window when you take into account that the resistance of copper change about 1% for every 2.5’C and the current will heat it

• NsN says:

That’s another aspect I wanted to try out. The current heats the copper, which changes it’s resistance which in turn changes the current.

It should be easy to calculate and model, but It’s always more fun to try something out for yourself ;)

One factor often used on “trace width calculators” is an accepted temperature increase of 10°C, which would mean a 4% change in resistance. I think the general variance in operating temperatures might be a bigger problem.

One reason I started thinking about this, was that I’m playing around with a small adjustable power supply, and the 30mOhm resistors were pretty much the only component I would need to order. If I could substitute these with a simple trace for the first prototypes, it would save me some \$20USD shipping costs on a 30 cent component. (This isn’t for anything commercial, just hobby stuff)

• NsN says:

That’s actually one of the things I wanted to explore. You can calculate the theoretical resistance relatively straightforward. But I was wondering how accurate the calculation would be in a real scenario. Etched traces have a trapezoid form and I doubt that the traces are absolutely of uniform width.

However, it also depends a lot on your needs. I know of one commercially produced brushless motor controller that uses simple traces instead of shunt resistors. I you just need to know whether the current is 0, 1/2 or full strength, a simple trace would be sufficient.