NEW!!! Owl + USB Connectivity CM160 Energy Saving Wireless Power Electricity Smart Monitor

A simple clip-on power meter with a handy remote display lets us know how much electricity we are using, how much it is costing, and how much CO2 is being produced. The desirable option of USB connection to the computer allows for usage profiles to be saved and compared, and the supplied software helps us in this task. The transmitter can accommodate up to three clip-on current sensors, allowing use with three-phase electricity.

The wireless range is good and the display on the hand held unit is big and easy to read. Setting up the receiver/handset is simple, and several of the default values were already correct for UK making life even easier.

However, I was less than convinced about its accuracy. So I made up a short extension lead with the live wire brought out in a loop for the meter to clamp around, and I tried some simple tests.
1. A fitting containing four 15Watt low-energy fluorescent bulbs - a nominal 60Watts: the meter suggested they were taking 115Watts, almost double the rated power!
2. A freezer rated at 100Watts: it showed 55Watts. Hmm.
3. A plasma screen rated at 300Watts displaying a static test pattern: it flipped between two readings, either 215Watts or 430Watts! What?!
4. A kettle rated at 2000Watts: it went from 2100Watts to 1900Watts, cold to boil, exactly as expected.

I have just compared the readings on my electricity meter with those of the CM160 for 24hour's usage, and the CM160 was 28% high. All of my lighting is low energy fluorescents, so perhaps they are the main reason for the excess power reading. But I don't know for certain, and I guess it might be something to do with power factors. I will be borrowing a True RMS Ammeter in a couple of days so I can see how the actual current compares with the suggested current measured by the CM160 clamp.

I guess it is better to be pessimistic and show we are using too much. But I want to see the real answers. I've not bothered with looking at the software in detail yet, not until I know if I can trust the CM160. More later.

Addendum 1. 01/10/2010. Software.

The software is basic and simple. I found the way of pushing into the graph to get more detail counter-intuitive to start with, but soon got used to it, as in that respect it is like some CAD programs. One can print the graphs, but they are almost too simplistic, and while one can have four panes, there are only two flavours of graph. I guess where the multiple panes comes into their own is if one has more than the single sensor plugged into the transmitter.

If a permanent and on-going record of the consumption is required, one needs to remember to connect the Owl to the PC at least once a month, otherwise the internal store in the Owl eventually wraps around and writes over the old data.

Addendum 2. 01/10/2010. Accuracy.

I've been able to make some accurate measurements. The results might be interesting to some, very boring for most, prepare to switch off, zzzzz... :^)

Overall Accuracy.
For an eight day average on the whole house the OWL CM160 reads just 1.5% high, much better than my first measurement. The first brief 24-hour spell did not have us doing any significant cooking (electric hob & oven) or tumble drying or washing machine, but had lots of fluorescent light (LEF) in the usage. I've done an audit and my spreadsheet estimates that perhaps 20% of our annual electricity consumption goes on lighting.
The absolute accuracy is consistent and within that of the Fluke meter at +/- 0.5% for Resistive loads, such as a filament bulb or a kettle.
eg 1: the Owl showed a 60W filament bulb used 58W, and the current through the Fluke also corresponded to 58W.
eg 2: the Owl showed a 2000W kettle (rolling boil) used 1908W and it was 1902W with the Fluke.
The mains measured at 234V during these tests.

Power Factor and LEF?
It is a different story with LEF bulbs however. One cannot even select particular brands as being not compliant with CE standards. Some 15W LEFs measured almost correct at about 18W with both measurements, others (different style but same brand!) were way wrong at up to 30W on Owl and 16W on Fluke! So I put a 1.0 Ohm power resistor in the neutral leg of my test fixture and connected a battery powered oscilloscope across it (ie this laptop with a USB Picoscope 3204). The current waveform of the well behaved 'good' bulbs, while quite lumpy, looked almost like it was trying to be the sine-wave one would expect from a resistive load. But that of the 'bad' bulbs was incredibly spiky, mostly off and then with high peak on-current for a very short time. I noticed that a mobile phone did not work so well near these 'bad' bulbs, which is another indication they do not meet CE standards, despite CE being stamped on the bulbs.

Freezer motor.
Yes, the 100W motor was only using 55W, and yet it is not on for much of the time, and the freezer is brilliant. Buy an Electrolux!

Plasma.
This uses a steady 215W with four small kinks in an otherwise clean sine-wave. But then it does have many filters and a dedicated circuit for power-factor correction. I don't know why the Owl occasionally doubles the value, since the Fluke did not, and nor did the oscilloscope show anything strange.

The Verdict?
You can believe the Owl CM160. It is near enough accurate to be eminently fit for purpose. But beware, some manufacturers of electrical equipment are selling non-CE-compliant stuff.

Addendum 3. 22/10/2010. Thermometer

There is a simple digital thermometer incorporated in the receiver, and I thought I had better check it as well. I already have two independent digital temperature measurement devices from Oregon Scientific as well as two old-fashioned Factories Act thermometers (RS Components 301-757). The Oregon base unit and its external MSF receiver and thermometer and hygrometer combo unit both indicated 22.0°C, and the Factories Act thermometers were both on the 22°C line, when the Owl indicated 22.8°C. The Owl's accuracy is reasonable for non-scientific use, but I would expect a modern digital unit to be more precise than this.

Having recently asked the people at Owl, they told me it is not intended as an accurate instrument, but merely as a `helpful guide to usage' and they actually suggested 5% as the accuracy limits!

Addendum 4. 22/10/2010. Software

I have a problem with the software, not apparent when I first tried it after only a couple of days of use. If I push down into a month bar I find that not all the days in a month are accessible, only the middle six.

I've reported this to Owl, and they are looking into it.

29/10/2010
I originally tried it on a 3GHz XP machine. It is now working exactly the same on this 1.8GHz W2KP laptop, just the same, although W2KP is not listed on the install CD label. :^)

Addendum 5. 29/10/2010. Battery Life.

The Receiver unit has run out of battery today after just seven weeks of use; the 4.75V of three new AA cells was down to 2.45V only. It still remembered all the info however, and the only detail that has been missed is a few hours overnight, and the clock was three hours slow but still trying to tick on. It is happy to run on the USB with no batteries in, so I've left it plugged into a permanently powered USB socket to see if the fresh batteries last more than seven weeks.

I am surprised at how quickly this has run down. I have an Oregon Scientific weather station that uses much the same 2.4GHz technology and those batteries last a couple of years in each unit.

The Sensor unit still has ample battery showing at 4.55V for the three AA cells.

Addendum 6. 18/12/2010 Battery Life - again.

The receiver batteries still run down even though it has been plugged into a permanently powered USB hub! Not as rapidly as before, but after another seven weeks the three Duracells are now down to 3.9V, ie about half discharged. I would have expected no discharge if the 'free' USB power was being used for everything.

The Sensor unit still has ample battery showing at 4.45V for the three AA cells.

Addendum 7. 18/12/2010 Software.

The store and display software has behaved itself for the last couple of months, with no missing days. So I have to assume there was a problem I did not notice with the earlier locations of the transmitter and receiver, and the receiver just ignored data it did not receive properly.

The USB driver is another matter though. It does not work properly if the laptop goes into Standby with the Owl connected, and an Unresolved Interrupt conflict forces a cold boot on Restart. However, if I unplug the Owl just for the five seconds during which the W2K laptop goes into Standby, then everything is fine on the instant Restart. My two other permanently connected USB devices cause no such problems. Maybe this is sloppy programming, and maybe the more fault tolerant XP or W7 will cope.

Addendum 8. 08/09/2011. First anniversary, update.

Both units are still working satisfactorily after a year in continuous service, with the Sensor battery measuring at 4.00V for the original three Duracell AA cells.

I've not bothered to use the USB data logging software for several months now, ever since it caused the boot problems on my ancient laptop. It is now un-installed. Read more ›

We bought this because with 4 teenagers in the house, our energy bill was spiralling out of control. We wanted to get everyone involved in trying to reduce costs and thought that this might help. Well it most certainly does! Took only 10 minutes or so to set up and could instantly see the cost of what was on in the house. We left it running for a week and carried on as usual, noticing when the monitor was particularly high (electric shower, double oven, tumble dryer). We then went around the house switching off all the TVs on standby, lights in unoccupied rooms, computers not being used etc to get the baseline cost as low as possible. As we switched things back on we watched the monitor to see which appliances made the cost jump. We've been able to work out the cost of leaving lights on or a computer all day when no-one is even in the house. After the second week we downloaded the data to the PC and bingo second week's cost was quite a bit less than the first! All the kids have been suggesting ways of reducing costs so it's been really helpful.

I set up the monitor with enthusiasm. The instructions were reasonably clear and the remote monitor behaved just fine. Then came the moment of truth: I can't run this monitor (or any other brand) on my system. The sender unit requires separate input mains leads to the consumer unit: my thick, shielded mains cable in an older house just won't work with this device. The instructions do warn that the Owl won't work on shielded input cables, but that's a bit late. This should be spelt out clearly in the product description.
I could have sought a refund but instead passed on the machine to a family member as a Christmas gift.