9

Looking for a shunt resistor for measuring power usage and was wondering if anyone had any experience with these.

It should be capable of handling 110V/20A in-home circuit.

EDIT: Energy monitors are the rage these days and I was thinking about putting a simple circuit together to try out. This would be used to measure a typical appliance or electronic device such as stereo or television. Similar to a Kill-a-watt or multimeter. I haven't really thought about accuracy as much as getting something to work well. I haven't thought this all the way through just thought maybe some here may have already "been there done that".

jdiaz
  • 1,933
  • 4
  • 21
  • 20

7 Answers7

4

The key is finding one with low enough resistance that it doesn't have to dissipate much power.

Two of these in parallel would do the job: http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=TMC5-.10-ND

With two in parallel, the resistance will be 0.05 ohms, so at 20 A, you'll have to dissipate 20 * 20 * 0.05 = 20 W. They're rated for 5 W continuous, or 10 W for the two of them. As long as you don't have to maintain 20 A, they'd work fine.

I'll see if I can dig up something better.

--

Edit: here's a better solution.

Try one of these: http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=MR50.01FTR-ND

The resistance is 0.01 ohms, so at 20 A, you'll dissipate 20 * 20 * 0.01 = 4 W. The part is rated for 5 W continuous, so you're safe.

pingswept
  • 12,581
  • 4
  • 46
  • 64
  • 2
    The problem with using a 0.01 Ohm resistor is that you will get only 9mV for a 100W load. This may be quite difficult to measure without some good amplifier in front of the ADC (or a nice sigma-delta ADC with low reference voltage). – jpc Apr 27 '10 at 23:56
  • That's a good point. I'd assume that you need an instrumentation amplifier across the shunt resistor. – pingswept Apr 28 '10 at 01:58
  • 2
    Low shunt resistors always require instrumentation amplifiers (or current sense monitors). You can get cheap chips that amplify signals 50x or even 100x and work for a broad voltage range. You are likely doing signal condition before you put it in a ADC anyway. Instrumentation amps can achieve 1000x of 10000x gains, but that isn't of much use at mV's. – Hans Aug 12 '11 at 09:38
4

You haven't stated the accuracy you require or the range of currents you want to want to measure.

If you are looking for a 5% measurement you may want to look at a current transformer or a hall effect device. The nice thing about these is they give you isolation from the line. Take a look at the various power meter application notes on the TI, Microchip and ADI sites. They will probably have some specific recommendations.

For a resistor be careful of the power dissipation. You need to derate for dissipation and you may need to look at the temperature coefficient for accuracy. Also some power resistors have very specific mounting recommendations -- height off the PCB, the amount of power being dissipated by adjacent components. If you need an accurate measurement you will want a resistor with Kelvin connections (four terminal). If you need a wide range of currents you may also need an instrumentation amplifier.

Daniel Grillo
  • 7,659
  • 18
  • 51
  • 69
jluciani
  • 11,646
  • 1
  • 34
  • 54
  • I've looked at CT's and hall effect sensors but I wanted to look into this method more. – jdiaz Apr 27 '10 at 23:46
2

Ah, you actually want to build a power meter-- take a look at the STPM01. It's an IC designed to be the core of a residential power meter.

http://www.st.com/stonline/products/families/analog_and_mixed_signal/interface_ics/related_info/stpm01.htm

Digikey has them in stock for $9.94 in qty. 1. You'd need to connect it to a microcontroller of some sort via SPI.

I suspect that there are also other ICs like this available-- maybe check Analog Devices, Linear Technology, and TI.

pingswept
  • 12,581
  • 4
  • 46
  • 64
2

For higher current ranges, such as the one you're looking at, it is probably better to use an inductive pickup. This allows you to measure high currents accurately, without having to dissipate lots of heat from shunt resistors.

An example would be the Digi-Key 398-1081-ND, which gives a DC voltage out proportional to the AC current in the system.

Jesse
  • 866
  • 1
  • 12
  • 14
1

The ACS714 from Allegro comes in a common SOIC package and has an internal shunt resistor, so no worries about parasitic resistances in measurement.

ACS714 PCB

The ACS714 measures bidirectional currents up to 30A, and outputs 66 to 185 mV/A.

stevenvh
  • 145,145
  • 21
  • 455
  • 667
1

If you're just measuring AC, use a current-sensing transformer, such as this EPCOS one from Digikey (or flip through some others). It offers much lower resistance than a shunt resistor (0.8 mΩ nominally) and provides the added benefit of isolating your sensing circuitry from the mains line.

Hall-effect current sensors also offer similar performance, such as these from Allegro, but work with DC as well as AC. Digi-Key carries most of their product offering as well.

Nick T
  • 12,360
  • 2
  • 44
  • 71
0

Shunt resistors supposed to be used for DC, for AC suppose to use amper clamps which is the same current transformer.

auto25242
  • 81
  • 1