current sense resistors

mercrazy

Newly Enlightened
Joined
Jul 28, 2014
Messages
160
what's the difference in a current sense resistor and regular thick film? do current sense carry power just as good? are they ok to use to set current on pwm chip circuits if LED current passes through them and as long as current stays below their rating? thanks
 

DIWdiver

Flashlight Enthusiast
Joined
Jan 27, 2010
Messages
2,725
Location
Connecticut, USA
There can be differences, but often they are exactly the same thing. It's just 'current sense' resistors are really low values. As far as I know there isn't any generally understood difference between a current sense resistor and any other low-value resistor of the same power rating.

You should probably stop reading here.

Sorry, but I just have to keep going...

That said, I can think of two features you might encounter which are targeted primarily at current sense applications:
1. Four terminals. For high precision, and to mitigate parasitic resistances, two terminals are used to carry the current, two others are used to sense the voltage. This is sometimes called a Kelvin connection. These are unusual, but available.
2. Low thermal-emf. Any junction of two dissimilar metals is a themocouple. Every thermocouple generates a voltage dependent on temperature. A surface mount resistor would have at least these, probably more, at each end: copper trace to solder; solder to nickel plating on resistor; nickel to resistive element. Since these are symmetrical at the two ends, you'd think that they would cancel. And they would, if everything were at the same temperature. But in a hot resistor there are inevitably thermal gradients, which cause the resistor to generate thermal EMF (voltage). In high current, low voltage applications, this can be significant. I found in my 13.5A driver, a 0.005 ohm resistor generated enough thermal EMF to alter the current setting by 0.2A or more. This was most disturbing when I commanded a change from 10A to 0, and it went to zero then over the next 20 seconds or so, rose back up to 200 mA, then gradually went back to zero.

Both of these are rarely required, and substantially expensive due to low production volumes, as well as more complex design.
 

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
With today's small surface mount resistors, it's not usually a concern, but with physically larger resistors, inductance can be something to watch out for.
 

bshanahan14rulz

Flashlight Enthusiast
Joined
Jan 29, 2009
Messages
2,819
Location
Tennessee
Another thing is tolerance. It could be that resistors with a tighter tolerance rating may be advantageous in a situation where you need very precise voltage reading.
 

SemiMan

Banned
Joined
Jan 13, 2005
Messages
3,899
I am calling bs on the thermal EMF of the resistor being the cause of your 200mA shift. The output resistance of that thermal emf would be quite high. In conjunction with the 0.005 load resistance the effects would be minimal. Thermal emf usually comes into play with low voltage high impedance circuits.
 

DIWdiver

Flashlight Enthusiast
Joined
Jan 27, 2010
Messages
2,725
Location
Connecticut, USA
You're right. I went back and checked my notes, and the worst case peaked at a bit below 100mA, not 200mA. In this circuit, 100mA in the LED takes 483uV in the sense resistor. And when the LED is off, the sense resistor sees a fairly high-impedance load. Since the EMF is in series with the 0.005 ohm resistance, it is not loaded by it.

I observed the offset to decay over a period of tens of seconds after switching the LED off. In one case, the offset appeared several seconds after switchoff, peaked, then settled gradually. For a long time I blamed the op-amp. I contacted the manufacturer, who suggested maybe I wired my op-amp inputs backward (yeah, right). I changed to a completely different amplifier. Same problem. Then I ran across this video: http://www.digikey.com/videos/en/v/Thermal-EMF-Error-in-Current-Sense-Resistors/1680490880001 Notice that one of the resistors shoots off the scale at 60uV just from the demonstrator touching the circuit board near the resistor. The two competitors' resistors that don't shoot off the graph are low-EMF resistors.

There's no suggestion in the video, or in other literature I've seen from Vishay, of how bad the EMF values can be, but I found this data sheet that claims 50uV/C is low! http://www.ohmite.com/cat/res_mcs.pdf

After seeing this, I found that by touching the board, I could change the offset, then watch it settle back down gradually after I remove my fingers. When I changed to a low-thermal-EMF resistor, the problem vanished.

I do have a -10mA (-50uV) bias built in to overcome any possible offset in the op-amp, and this seems to also cover what low themal EMF remains.
 

SemiMan

Banned
Joined
Jan 13, 2005
Messages
3,899
Check out Wsl series from Vishay. Makes this a moot point mainly from a design standpoint. Good layout resolves it as well. Should rarely be an issue.
 

DIWdiver

Flashlight Enthusiast
Joined
Jan 27, 2010
Messages
2,725
Location
Connecticut, USA
The WSL series is the one I'm using now. They aren't cheap though.

Layout issues are a completely different animal. Also critically important in that type of design, however. It never occurred to me that a trace 1/8th of an inch long, carrying 15 mA would be enough to be a problem (though not a huge one).
 
Top