I just got my first standby call out at work! A circuit breaker in the cloud experiment had tripped out causing them to lose power. My boss for the evening (I’m still very much on standby duty training…) Michael quickly found the problem and fixed it – but it set me thinking about how useful it would be to have ‘sticking plasters’ that measure temperature of electrical equipment (such as circuit breakers).
The problem was caused by an overload on the circuit. Simply put, too much stuff had been plugged in and turned on. It happens all the time, in domestic and industrial locations. After a while (between 0.001s and an hour or two) the circuit breaker decides it’s had enough of the overload and trips – cutting off whatever is connected downstream. In the meantime the circuit breaker (and the cables on either side of it) heat up, as a function of the current passing through the resistance of the copper cables. Most circuit breakers can get quite hot (>90 degrees C) without flinching, though occasionally you get one which is sensitive, or mostly just broken.
After fixing the problem we did a quick current measurement using a clamp meter (basically a multimeter with a large split-core Current Transformer stuck on the end) and found that everything was within the normal parameters for the circuit breaker. However, I think the call out service is often used because of this type of problem – okay it was only my first ever call out, but I’ve heard of it happening before.
So I thought of two things that would be useful in diagnosing and monitoring this type of problem. The first is an infra-red thermometer, the type with a laser targeting spot. Sure this sounds expensive, thoughI think Maplin sell them for less than £50 these days (don’t ask about the accuracy and calibration!), but would be really useful for measuring the temperature of equipment after a trip. However, it isn’t perfect as normally the standby service arrive some time after the trip happens, according to the following sequence:
- Circuit goes into overload
- Circuit breaker trips
- Somebody (or something… there’s a lot of electronic status monitoring already deployed) notices
- The person/thing that’s noticed calls the control room
- The control room call the standby service
- The standby person leaves their house and heads to the fault location
- The standby person find the fault (this could take a while… especially in complicated buildings)
- Fault gets fixed (hopefully!)
So there’s probably up to an hour of cool-down time in there, maybe more – all of which would render taking the temperature on arrival a little previous. It might be nice to take the temperature once it’s been fixed, to see if it starts to increase rapidly again (an indication that there is a persistent fault) but this can be done more scientifically using a clamp meter.
I think it would be cool if someone (maybe me?) could make a sticky probe to clamp on the device which would enable it to be remotely monitored over a short period of time. This could be either a split core CT for current monitoring (more accurate and expensive) or just a simple thermal sensor (probably a thermistor, or perhaps a more expensive thermocouple), or maybe even both – to see which works best.
The device would need the following characteristics:
- Small enough to fit into cramped existing switchboards
- Radio strong enough to penetrate the switchboard casing
- Batteries and/or power harvesting so that signals can be transmitted for a reasonable period of time, between 24 hours and 7 days.
- The sensor (temperature and/or current)
- Some form of enumeration – it would be great if there was a sticker you could just peel off that would title the device with the time of activation
- A way of getting the data back to the IP layer, so a radio-IP gateway. This is probably the trickiest bit, especially with a low cost solution in mind. It could take the form of a wifi RF package within the device, or a lower level, significantly cheaper radio format back to a base node with a wired connection.
- A website/database for tracking the data collected, and signalling when it’s time to go back and collect the magic sticking plaster.
Best of all would be if the device was self powering, so could be left in place indefinitely – each deployment helping to improve the reliability of the electrical network in a lasting and meaningful way. With a few devices in place, the biggest benefit would be from a ‘pre-trip’ temperature/current threshold – and if this feedback could be provided in a meaningful way to the user, they might even be able to load-shed (i.e. turn things off) before the trip actually happens.
Maybe one day…