Cheap Power Meters: Review & Comparison

How to reduce your electricity bills with a $20 power meter

11 min readMar 25, 2019

Today, humans use more energy than ever before, with the primary energy supply now exceeding twenty trillion watts — that’s about three thousand watts for every man, woman and child on Earth, 24 hours a day, 365 days a year, averaged across everywhere from the poverty-stricken plains of Sudan to the wealthiest suburbs of Los Angeles. A lot of this energy escapes from power plants, transmission lines and vehicle engines as waste heat, but whenever we reduce the electricity we use, we also reduce the amount of waste heat in the energy production process.

We’ve been encouraged to reduce our energy consumption, yet most of us have no way to easily measure our energy consumption in the first place! How can we tell whether something is using a lot of energy in our home? When we change something, how do we know if we’re saving energy and money?

It would help if our wall sockets all had built-in power meters, but as long as this isn’t a standard feature, we can at least buy power meters on Amazon. When I went to buy electric power meters, I couldn’t tell which units were any good, so I decided to buy three of the most inexpensive models, each $20 USD or less, and find out if any of them stood out as superior:

Spartan SP-PM120
RioRand (the white one, which looks curiously identical to the “Excelvan HP1101” in this review and the “Floureon” in this review)
TackLife EM02

Basic Features

The three units are all very similar. Each has one grounded power outlet, old-style LCD display, and almost identical published specs:

Spartan: 1800W max, “only suitable for 120V/60Hz”.
Rio Rand: 1800W max, 110–130 V, 60Hz, 0–15 A
TackLife: 1800W max, 100–150 V, 0–15 A

Unfortunately, none of the units allow 240V input. Therefore, you can’t use these in many countries outside North America, and you also can’t use them to measure large appliances like a stove, furnace or dryer that is attached to a higher-voltage line (if your house has one).

The Tack Life unit hogs the most space below itself, blocking access to the other plug when it is plugged into a standard wall outlet:

The Rio can be used in the top outlet with a slimline phone charger in the bottom, but it’s a tight fit. The Spartan wins this comparison: it is the most compact on the bottom and applies only slight pressure on the phone charger. All three units have similar widths and depths. The Rio is the tallest and the Tack is the shortest.

The Spartan’s LCD is a bit faint when viewed straight-on, and invisible from below. On the plus side, the display is crystal-clear when viewed from above, and the Spartan is the only unit that seems to have a backlight. The backlight turns on when you push any of the buttons.

The Spartan offers the highest precision when measuring small loads — 3 digits after the dot — while the other two only offer 1 digit. However, I discovered that you can’t rely on any of these units to detect very small loads correctly (as discussed below), and the Spartan shows voltage as an integer.

All three units are able to measure the power factor, which is a good sign. It suggests that the measured number of watts and kWh should be approximately accurate. (If a unit can’t measure the power factor, it can only measure the “apparent power” which, for some appliances, is greater than the real power.)

Power used by the power meters themselves

Since I have three units, I can use them to measure each other. My preferred unit for this experiment was the Spartan because it measures the Watts to 3 decimal places (on small loads). Although it has higher precision than the others, I discovered that is unable to measure very low power levels (below about 0.4W) and will show 0.000 W for such loads. The Tack Life unit seems to have the same problem. The Rio has less of a problem: I found a load that was almost exactly 0.3W, but the Rio said it was 0.2W.

So, measuring small loads requires a trick:

Using a splitter, I plug in a “side load”: something that uses a small but steady amount of power, such as a 2-Watt LED light bulb, or a night light that does not have a light sensor.
Estimate the average power used by the light bulb, by watching it for several seconds and recording a number in the middle of the range of numbers shown on the display.
Add the very low-power device, wait a few seconds, and record the new power level. Subtract the numbers to find out how much the power level changed.

I used this method with the Spartan to measure the power used by the other two units, then used the Rio and Tack units to measure the Spartan. Since the Rio and Tack units only offer one decimal place, I performed up to four measurements of the Spartan (with and without a side load), to be sure about the one decimal place I do have.

Unloaded results:

Spartan uses 0.4 W, or slightly higher
Rio Rand uses 0.41 W
Tack Life uses 0.27 W
Rio Rand and Tack Life use 0.68 W together (this measurement was just a sanity check)

When loaded with 10W LED bulb:

Spartan uses about 0.5W, or slightly lower
Rio Rand uses about 0.43 W
Tack Life uses about 0.27 W

The second test was more difficult because the light bulb’s power use seemed to slowly decrease over time. When I realized this, I kept the light on for several minutes, and after switching it between power meters I waited 30 seconds before taking readings. I also took reference measurements (light connected directly to first power meter, i.e. the Spartan when measuring the Rio or Tack) both before and after, then averaged them. Although the Spartan still offers 3 decimal places of precision for these measurements around 11W, the last digit always fluctuates wildly, and the next-to-last digit fluctuates slightly. I did not attempt to measure the power usage of these units under heavy load, because I don’t have a steady-but-heavy load to plug into them.

The units disagree about how much power the LED itself uses (in steady state): Rio says 10.5 W, Spartan says 10.10 W, Tack says 9.9 W.

Battery backup

The Tack and Rio units have built-in batteries so that they can remember the accumulated power usage (kWh) when you move them between different sockets. The Spartan is supposed to be able to do that, but it did not come with batteries even though Amazon and its manual both say it does; plus, it appears to require two button-batteries while the Tack only needs one.

The Rio is the only unit with a rechargeable battery. The battery is built-in and difficult to access, guarded by at least 3 deeply-inset screws. I don’t know whether the unit’s batteries are already charged; if not, the 0.41 W measured above may include charging current. Meanwhile, the Tack unit’s battery compartment is held on by one tiny, tight screw that was difficult to remove without stripping the head. The Spartan’s battery compartment doesn’t use screws; the cover is designed to be removed with fingers alone. Unfortunately, it doesn’t fit quite right and I couldn’t remove it without an additional wedge tool such as a knife.

Are the units in agreement?

I wanted to find out whether the units disagreed about how much power a given load is using. I cannot be sure which unit is the most accurate, since I don’t have a “reliable” reference measurement to compare with. What I can do is plug in the three meters in series with each other to get simultaneous measurements on all three at once. The plan, then, is to plug in several different loads and see what each meter claims about each load.

Experimental setup: I plug the Spartan into the wall, the Rio into the Spartan, and the Tack into the Rio, then measure the Watts, Current, Voltage and Power Factor. Since Tio and Tack use about 0.7W together, we can expect the Spartan to see a number about 0.7W higher than the Tack.

| Load      | Meter   | Watts | Amps   | Volts | Power Factor |
|-----------|---------|-------|--------|-------|--------------|
| Large fan | Spartan | 42.1  | 0.367  | 121   | 0.93         |
| (Low)     | RioRand | 43.7  | 0.369  | 121.0 | 0.96         |
|           | Tack    | 41.1  | 0.347  | 120.9 | 0.97         |
|-----------|---------|-------|--------|-------|--------------|
| Large fan | Spartan | 48.7  | 0.402  | 121   | 1.00         |
| (High)    | RioRand | 50.7  | 0.418  | 120.8 | 0.98         |
|           | Tack    | 47.9  | 0.394  | 120.8 | 1.00         |
|-----------|---------|-------|--------|-------|--------------|
| Blow dryer| Spartan | 397   | 3.346  | 119   | 1.00         |
| (Low)     | RioRand | 415.2 | 3.530  | 118.2 | 0.99         |
|           | Tack    | 393.1 | 3.333  | 118.1 | 1.00         |
|-----------|---------|-------|--------|-------|--------------|
| Blow dryer| Spartan | 1411  | 12.471 | 113   | 1.00         |
| (High)    | RioRand | 1453  | 13.13  | 110.9 | 0.99         |
|           | Tack    | 1374  | 12.41  | 110.6 | 1.00         |
|-----------|---------|-------|--------|-------|--------------|
| Massager  | Spartan | 15.2  | 0.207  | 122   | 0.61         |
| (Max)     | RioRand | 15.7  | 0.238  | 121.7 | 0.54         |
|           | Tack    | 14.4  | 0.215  | 121.6 | 0.55         |
|-----------|---------|-------|--------|-------|--------------|
| 32" TV    | Spartan | 36.36 | 0.459  | 123   | 0.64         |
| Panasonic | RioRand | 37.7  | 0.478  | 122.6 | 0.62         |
| 720p      | Tack    | 35.7  | 0.426  | 122.7 | 0.68         |

Notably, the RioRand systematically reports about 5% higher power and current than the other two units. The Spartan tends to report slightly higher power usage than the Tack, but remember, this was expected. I assume this means that the Rio’s numbers are biased high by about 5%.

I was surprised by the amount of power consumed by the fan, and that the lowest speed consumes a similar amount of energy as the highest speed. It might be that something is wrong with my fan — if you get a power meter, please check your own fan and let me know if you get different results. The small blow-dryer was also a surprise to me, as it uses more power than a typical microwave oven.

Power used by other appliances

Here are some things I learned with these meters:

Our inflatable hot tub (situated indoors) used an incredible 8 kWh per day (averaged over a 6-day period). In my location that’s about CAD$1.50 per day or about CAD$550 per year (including transmission charges and other fees). When I learned this, I took steps to add insulation on the top cover, which reduced power use to about 6 kWh per day ($1.14). I already knew better than to use bubbles in the hot tub; they would obviously cause energy usage to skyrocket due to rapid water evaporation. Since we live in Canada you might think it’s okay if the heat escapes into the house, but heating with electricity is much more costly than heating with natural gas (except when heating via heat pump.)

Our midrange gaming PC, including 27" monitor and peripherals, uses about 160W while idling, and about 260W fully loaded (playing DmC in stereoscopic 3D) which is about 5 cents per hour. If it is idling and the monitor is turned off, power drops to 120W or $0.55 per day. The rig uses 12W when powered off; I determined that this power is mostly used by the broadband router (about 5W) and the power bar itself (about 5W). You can save power by putting your PC to sleep, but I’ve noticed some PCs, well, they don’t wake up again. Hibernation is a nice alternative — your PC saves its memory to disk, then powers off completely. If you have Windows 10 but you don’t see the Hibernate option beside Shut Down in your Start Menu, read about enabling it.
My laptop uses about 19W while idling, and about 62W under load (playing Left 4 Dead 2 using built-in display).
Our BenQ MH630 projector (3000 lumens) uses 210W in “Economic” mode and 255W in “Normal” mode. This is amazing, since “Max wattage 210W” is printed on the unit. Power usage does not change between bright and dark images. Note: the lamp is up to 6300 hours now. It uses 7 W when powered off.
The 32" TV uses about 35 W for a nearly black screen, and about 40W for a white screen. Fun fact: this TV has a traditional backlight. OLED displays should use much less power when showing dark scenes.
My 150 L (5.3 ft³) top-opening freezer uses 96W when the motor is running, so I was surprised to see it using 2.3kWh over 24 hours. This meant that the duty cycle was 100% — it never stops cooling. Apparently this is what happens when it’s on the maximum cold setting. I’m turning it down now!

Summary: Pros & Cons

TackLife (left):

Pros: Battery-backed memory.
Cons: Bottom of the unit hogs far too much space. Fewer features than the others.

Rio Rand (middle):

Pros: Large three-region display. Auto-charging battery-backed memory. I paid the least for it (Amazon.ca).
Cons: Fewer features and generally less precision than Spartan.

Spartan (right):

Pros: High-precision measurement (except voltage). Hogs the least space under the outlet. Large three-region display includes a backlight and a histogram of power use over 7 days with resolution of 1kWh or higher (in addition to a high-resolution cumulative total). Biggest manual. Overload protection (not tested). Battery-backed memory if you have batteries (batteries are advertised as included, but missing.)
Cons: The display is unreadable when viewed from below. This unit also costs a little more on Amazon. Also, unlike the other two units, the kWh number on this unit resets every day, and I can’t figure out how to view the total kWh.

I’d say the Spartan is too “smart” for its own good: although the lower area shows a chart of energy usage over 7 days or 7 months, that’s a very rough depiction. By reading the manual I figured out how to view the exact number of kWh for each day — implying that you need to set up the clock before you start using it — or for entire months; in case you lost your manual, here’s a hint: press and hold the history button for 3 seconds.

Note: The units can also be configured to convert the number of kWh into dollars/cents and carbon emissions, but I didn’t bother to figure out how. I would say that none of the user interfaces was immediately intuitive, except that on the Rio there’s a “function” button that did what I expected (switching between different display modes). I thought that the way the TackLife switches between modes was the most intuitive, except that the button for switching modes is labeled “MODEL” instead of “MODE”.

All three units had a low price, but none of them works above 1800W or 150V.