Hacking a Kill-a-Watt for very low power measurements

[NiHaoMike]

Measuring standby power - Page 1

One of the more intriguing suggestions, which I have yet to try is to modify a Kill-A-Watt by removing the 2mOhm (or so) current shunt and replacing it with a 200 mOhm (or so) current shunt. Device should read approximiately 100x more sensitive than it is. Obviously, you don't want to plug something that draws 15 A into it after modification, but it should be good for generating low power measurements. Probably very inaccurate, but better than nothing.

I modified one of mine to do just that, but I also added a pair of inverse diodes to provide protection against inrush and a switch so it can be used in "normal" mode. (There are lots of gotchas to adding a switch if you want anything resembling an accurate measurement!)

What I found is that most switching wall warts draw very little power - usually 0.3W or less. (I found that a recent Samsung phone charger used so little in standby that even a 100x more sensitive Kill-a-Watt would not register it - that's less than 10mW!) Linear wall warts vary all over the place, from a few watts for the inefficient ones to around 0.3W for the good ones, just like many switching wall warts.

BTW, don't expect that mod to be particularly useful for actual savings since 1W left on 24/7 is 0.72kWh/month - way down into the noise for even a very efficient household. (In an alternative energy system, just use a multimeter.) It's purely for bragging rights!

[Daox]

Cool project Mike. Thanks for sharing. I've long wondered what the new switching power supplies pull because you know they use something. This clears it up.

[NiHaoMike]

I should add that I have went through several dozen wall warts and of them, the least efficient switching unit used about 0.5W in standby. The best linear was the 0.3W I found earlier - a little worse than many switchers which averaged 0.2W. Note that while the presence of a high efficiency symbol ("III", "IV", or "V" inside a circle and/or Energy Star logo) guarantees that the PSU is efficient, the switchers without them did almost as well. Linears not rated for efficiency are indeed likely to be very inefficient, but a few of the smaller ones were under 1W. The inefficient linears easily give themselves away by noticeably warming up with no load.

Of course, the devices that the wall warts plug into vary widely in standby power use, with many not having a standby mode at all. Those that do generally use little extra power of their own. (For the purpose to measuring power use of devices that use wall warts, an inline adapter to hook up some multimeters is far more accurate than measuring the input of the wall wart.)

The indicators on power strips use about 0.2W to 0.3W. (I suggest leaving most surge protectors plugged in since having more surge protectors connected enhances the surge protection of the entire circuit they're on.)

[Ryland]

I just use my multi-meter on the amp setting, fractions of an amp are no problem and I don't have to hack kill-a-watt.

[NiHaoMike]

With AC, you can't just multiply current and voltage to get power unless you're doing a vector multiply, which is impossible on a standard multimeter. Additionally, with nonlinear loads, a multimeter must have DSP in order to measure properly.

[wuck]

Originally Posted by NiHaoMike

Measuring standby power - Page 1

I modified one of mine to do just that, but I also added a pair of inverse diodes to provide protection against inrush and a switch so it can be used in "normal" mode. (There are lots of gotchas to adding a switch if you want anything resembling an accurate measurement!)

OK Mike, I'll bite. Considering you're switching a 2 mOhm (.002 Ohm) shunt in & out, What kind of switch did you add and were you able to test the calibration afterwards? I saw "The Electrician" in the other forum was able to test his with a nice scope/app.

I don't own a kill-a-watt, but I purchased a WattsUp? similar device many years ago. It's hard at work logging my new refrigerator at the moment. I did open it up several years ago, but can't remember what I saw in there.

Pat

[NiHaoMike]

The trick is switching the current being measured and the sense point separately. It's a 15A DPDT switch that bypasses the additional shunt in one position and leaves it in circuit in the other. The other side of the switch changes whether the sense amplifier senses the original shunt or the added shunt. That prevents it from sensing the voltage drop across the switch and added wiring, which can easily be more than the value of the original shunt.

[ICanHas]

Additionally, with nonlinear loads, a multimeter must have DSP in order to measure properly.

This is absolutely incorrect. Additional digital blocks are not required.
DSP is just *a* way of measuring true watt, not a requirement.

Instantaneous voltage times instantaneous current is always instantaneous power.
the integration needs to occur on the multiplied product.

Power measurement can be entirely passive such as a multi-range iron vane panel meter or entirely electronic four quadrant analog multiplier that produces voltage output directly relating to real-time true watts. An entirely analog filter can be fitted to produce a continuous real time integration.

Real time true watts measurements pre-dates any of computer chip driven stuff and these technologies still the cutting edge in many aspects and vastly superior to adequate, but functional low sampling rate multi-pass sampling method used in most China sourced measurement tools do not capture every curvature in every wave cycle.

[NiHaoMike]

Analog multimeters have relegated themselves to niche applications. Almost all of them used nowadays are digital. There are still lots of analog watt hour meters in use, but they're not exactly portable instruments.

[ICanHas]

Entirely electronic four quadrant multiplier will let a digital(or analog) multimeter measure real watts without any additional digital circuitry or software embedded firmware magic (cheap for the OEM)

The multimeter only acts as a panel meter readout. The four quadrant multiplier analog wattage converter front end can be built into a multimeter sized enclosure.

Digital processing does not always rule (Or, what makes an analog engineer laugh?) | Analog Devices

I think the above user has some sort of bias towards the use of firmware based soft computing embedded systems to the point of proclaiming the use of DSP is a requirement rather than telling the truth that it is just one of the ways.