Can I measure the watts used by my PC with a DMM?

[moeburn]

Hello!

I want to see just how much power (in watts) my desktop computer is actually using, but I don't want to buy a Kill-A-Watt or Watts Up or an expensive UPS. I know that I can calculate wattage from current and voltage, but can I measure the current on 110v with my cheap $15 DMM?

I'll tell you what I can read on the DMM, without completely understanding what it means: It's 600V CAT III, 1000V CAT II, it has 3 ports, but I've only ever used the ground and "volts/ohms/mA" ports, because I usually use it for small DC projects.

It says that for THOSE two ports, it is "Fused, MAX 200mA, 1000v{line/dash symbol}, 750v{squiggly line symbol}"

But 200mA is too small, so that's no good. But then there's ground and the OTHER port, which is called "10A DC". Well its 10 amps, so that's plenty, although it says "UNFUSED" and it says its for DC.

So am I poop outta luck, and I gotta buy either a better DMM or a Kill-a-Watt?

 

[(*steve*)]

but can I measure the current on 110v with my cheap $15 DMM?

The short answer to that is YES.

BUT

1) a cheap multimeter is going to make it quite a dangerous thing to do.
2) the resulting measurements of voltage and current are likely NOT RMS
3) even if they are RMS measurements, simply multiplying them together won't give you the power in Watts.

The problem with AC is that if current and voltage are out of phase (i.e. maximum current gets drawn at a time other than at maximum voltage) then you need to take this phase shift into account (multiply by the cosine of the phase difference) in order to calculate *real* power.

In addition, even though a SMPS may draw maximum current at or near the peak of the voltage, it may only draw it as a brief high current burst. This needs to be converted into an RMS value (ad that's non-trivial).

Where there are lots of devices drawing power out of phase with the mains, the mains waveform can even be altered. This makes the calculation of the RMS voltage a little more tricky. Fortunately this is typically less of a problem that for the current measurement.

A device designed for the task is going to do most of these corrections for you.

However, beware that they can't tell if the device is consuming or generating power (not that this is likely to be a problem in your case).

 

[moeburn]

The short answer to that is YES.

BUT

1) a cheap multimeter is going to make it quite a dangerous thing to do.
2) the resulting measurements of voltage and current are likely NOT RMS
3) even if they are RMS measurements, simply multiplying them together won't give you the power in Watts.

The problem with AC is that if current and voltage are out of phase (i.e. maximum current gets drawn at a time other than at maximum voltage) then you need to take this phase shift into account (multiply by the cosine of the phase difference) in order to calculate *real* power.

In addition, even though a SMPS may draw maximum current at or near the peak of the voltage, it may only draw it as a brief high current burst. This needs to be converted into an RMS value (ad that's non-trivial).

Where there are lots of devices drawing power out of phase with the mains, the mains waveform can even be altered. This makes the calculation of the RMS voltage a little more tricky. Fortunately this is typically less of a problem that for the current measurement.

A device designed for the task is going to do most of these corrections for you.

However, beware that they can't tell if the device is consuming or generating power (not that this is likely to be a problem in your case).

Thanks for that very informative reply! I guess I'm poop outta luck then. I wish I lived in one of those cities that let you rent a Kill-A-Watt from the local library, but I'm in Toronto. Hell, even Ottawa has Kill-a-Watts for rent in their libraries.

 

[moeburn]

Quick question, what if the computer was plugged into a portable car 12vDC to 110v AC-inverter, and the multimeter was set to measure 10A DC, in series with the 12V line coming out of the car's accessory jack?

 

[(*steve*)]

Again, you have the problem of it not being a steady load, and also some real vs apparent power issues.

One option is to have the computer the *only* thing turned on in your house (it means turning off fridges etc and isolating things like aplarm systems and other things that may be wired directly into the power -- maybe you can turn off circuit breakers).

Then, you may be able to use your power meter to make an estimate.

Some have a light that flashes some number of times per kWh, others have a disk that rotates (really old ones), whilst on others you may have to waut for 1/10 of a kWh to flick over on a display.

With any of these methods you can time how long it takes to consume some known amount of power and then convert that to a rating in watts.

For example, if you note that 0.1 kWh is consumed in 38 minutes, then you can calculate that in an hour (60/38) * 0.1 = 0.158 kWh. So the device uses an average of 0.158kW, or 158 W.

This method also has the advantage of taking an average over time. It works well with fridges and air conditioners, etc. which cycle on and off.

 

[moeburn]

Again, you have the problem of it not being a steady load, and also some real vs apparent power issues.

One option is to have the computer the *only* thing turned on in your house (it means turning off fridges etc and isolating things like aplarm systems and other things that may be wired directly into the power -- maybe you can turn off circuit breakers).

Then, you may be able to use your power meter to make an estimate.

Some have a light that flashes some number of times per kWh, others have a disk that rotates (really old ones), whilst on others you may have to waut for 1/10 of a kWh to flick over on a display.

With any of these methods you can time how long it takes to consume some known amount of power and then convert that to a rating in watts.

For example, if you note that 0.1 kWh is consumed in 38 minutes, then you can calculate that in an hour (60/38) * 0.1 = 0.158 kWh. So the device uses an average of 0.158kW, or 158 W.

This method also has the advantage of taking an average over time. It works well with fridges and air conditioners, etc. which cycle on and off.

I'm not worried about it not being a steady load, in the sense that I know the PC will draw more power when it is doing more calculations, etc, if that's what you mean.

I actually have a little wireless energy meter supplied by Toronto Hydro, that uses this thing that mounts onto your power meter with a little light sensor arm that goes over the meter's flashing light, and then transmits that data to the wireless energy meter. But it only has a resolution of 0.1kW, not good enough for what I wanted to measure. I wanted to see how much less wattage Windows 8 was using over Windows 7, since it is keeping my CPU 10 degrees cooler (and also specifically designed to use less power), but I'd need precision down to at least 10w for that.