Let's Make Robots!

A watched meter never boils

I would love to be able to have continuous readouts from my electricity meter in my home. Such a tedious job is time consuming and tedious. So I need a robot to do that for me.

I've seen projects use digital cameras and optical character recognition software, but to me that seems such overkill. Also, I do not have any cameras to spare or the skills to program OCR into my software. I do however have in my lab:
- IR emitting LED
- IR photo transistor
- piece of heat shrink tube
- LM358 opamp
- potmeters
- lots of AA batteries and holders
and the usual experimental equipment including a brand new oscilloscope, one like this!

The wheel has a silver rim with one black spot in it. That rim is my target. The black spot should be detectable. Even from a distance: I cannot open the meter. It is sealed by the power company. I need to observe it through the little window.

This is the 3 mm IR LED. The potmeter regulates the current. It proved crucial. Too much light and the detector is blinded. Too little and there is more noise than signal. A current meter showed that 8 mA is just fine for my purposes.

The detector needs to be very directionally sensitive. A lens would be perfect, but complicated to set up. Luckily a photo transistor is so sensitive that you can afford to stuff it in the end of a tube. My transistor and tube are 5 mm wide. The tube is 20 mm long. They are not in this photo.

So now I know what it takes for the optical set-up to work. See the video or a demonstration.


Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

Just thinking that a laser pointer might be a better light source than you IR LED. Most of those cheap red laser pointers emmit enough IR to be read by the IR photo transistors. Especially the 850nm phototransistors.

Because the laser is more accurate you should be able to get a cleaner signal. Many of those cheap lasers have an adjustable focus. If you unfocus the laser slightly to get a small spot rather than a tiny point this may help with the alignment of the phototransistor.

When I looked at Gareth's spoked laser interrupto, I couldn't help but wonder that myself. I haven't tried it yet. Mainly because my current setup is giving me excellent results now.

However, the real challenge lies in my gas meter. It has no disk, just blck wheels with white numbers rolling away upwards. I already discovered that the number 1 gives the lowest reflection, but the tiniest adjustment of the optical setup is enough to disturb the entire 0-9 readout. Perhaps I should fire my one and only laser into my gas meter. No?


No what could possibly have gone wrong?!

Looking at the meter, it's a south-holland one from the year 1964. I estimate they do 187.5 revolutions for every kWh :)

Have you tried running the tape to hold the reader horizontally? That is the direction you will have to go when it is complete so the power company can read it without upsetting your device. And also would black tape shield from light better and improve reading?

Also is your meter indoors? Here they have regulations where the meter must be accessible at all times and if it is locked must be standard key. If they have to estimate it can be costly.

I had to read your question four times before I understood the misunderstanding on your end. This is only my experimental setup. I already built a plywood lid that will fill up the entire window. Thus obscuring the inner workings altogether. I had to go back to this tape and scraps method when it did not work. The tube and the potmeter on the LED turned out to be crucial. Now I hav perfected the optical setup, I will go back to a nic lid again. It can be removed without upsetting the fine tuned focus of my IR set-up.

About our power companies: they do not require outdoors access. In recent years it has become common for customers to read their own meters and just send the data to the company, via snail-mail or web. Only when you raise an (surely automated) eyebrow or two, will the company bother to send an actual meter reader. I guess the unions lost this one...

Failing to send in the readings in time will cause an estimation of your meter based on previous years. They are likely to over estimate, but next year's reading will always compensate for that.

This is in the Netherlands btw.

They will overestimate and charge you another 18% on top of that for the btw ;)

Nice usefull project, any chance you will attempt to slow the f*ck*r down also? I'm very interested :D

That sounds a good system for the power company. Big business is always finding ways to cut number of staff. Soon there will be no jobs for anyone. I wonder how long before they start to do that here. But perhaps smart metering will be implemented in all households before then. So they can start charging ridiculous amounts for peak usage.

We have a state government sponsored program here if you pay $50. An electrician comes out and installs a wirelesss power monitor and some cfl lamps. Pretty good deal. It really doesn't require much skill as the transmitter just clamps over the meter cable unintrusively.

energy monitor


Have you determined yet how many revolutions equals a kwh or watt hour yet?

I see your going non-invasive too rik.

I would be so much nicer and easier to break the "do not break seal" and get inside .......... (you know to paint some stripes and twiddle some pots ;-)

My meter has a boring tiny notch cut out of it for each rev .- fat chance of detecting

With a sensitive hall effect sensor you might be able to detect that notch Gareth.

:D Nerds never sleep :D