Low-Power Control Systems

I needed a heavy-duty current sink to test and calibrate current sensors for the Nagios Renewable Energy Monitor system.

Something able to sink the ‘normal’ load currents of up to 20 Amps and maximum load currents of 30 Amps I am designing for.

I had a heater element in a damaged 12V coffee filter machine that might be up to the job. But when I opened up the filter machine housing, I was compelled to try to repair the forlorn device.

After all, it would make as good a current sink if it was brewing coffee as if it were heating the workshop.

The filter machine came my way after being stored in a waterproof container into which rain had seeped, creating a water-tight bath with the filter machine submerged at the bottom of it.

I should have taken a photo of the filter machine when I first fished it out. A vivid orange deluge poured out of it and, once dry, splotches of fine orange rust coated most of the housing and the power lead. The rust powder had worked its way into every crevice, making its cleaning a very manual chore.

It was most fetching, but not something you’d want to involve in high-current electrical testing, let alone food preparation.

More importantly, the internals of the 12V ‘car cigarette lighter’ plug at the end of its lead were corroded into uselessness. As was the on-off switch on the front of the filter machine. And who knew what had gone on in the heater element beneath the warmer plate and in the themostat I found there?

I replaced the 12v power plug from my junk box and the switch with one from Maplin. Hurrah for the convenience of standard switch sizes – now that all switches are produced in one Chinese factory.

After lots of drying – and even more cleaning – a multimeter test of the thermostat showed it was conducting so I took a bet that the heating element might also be electrically intact…

Trepidatiously, I fired up the machine. The 12V power plug melted a few moments later but not before I heard the first faint gurglings of warming water.

Clearly, this machine still had potential for its designed purpose. But I questioned its usability, except as a test load. The problem is that neither a melting cigarette lighter plug nor the happy gurgle of warming water are satisfactory indicators of operation. Except when metered up for testing, it would be hard to know what the filter machine was doing to my batteries at any given moment. And it was impossible to know what it was doing to my batteries over a period of time.

After all, when wind generators and solar panels provide your power, you need a very good sense of what is consuming power at any given moment, as well as how long it has been consuming power. That intuitive sense is even more important for monitoring power-hogs like a coffee filter machine.

At which point I started to pimp it up.

It needed some indication of ‘power-on’ and a separate indication of ‘power-being-consumed’. A couple of small green and red leds from the junk box (actually from a Maplin ‘joy-bag’), a couple of 3.3k resistors and a couple of carefully positioned and carefully drilled holes in the filter’s housing and we were ready for testing.

The lower green led to the right of the switch shows when power is connected, while the upper red led to the right of the switch shows when the heater element is powered, which indicates when the filter machine is draining power heavily, as well as when it is brewing. The red led should switch on and off as the thermostat reacts to the heater element temperature.

And that’s where it all went wrong.

After swapping the plug for a 16Amp cigarette plug, I tried again, this time with a 30 Amp ammeter hacked into the supply line.

Check that meter reading in the lower lefthand corner. 18Amps! That’s a lot of current.

Note the leds – they did their job perfectly. But after the thermostat reached its 62 deg C and switched the element off for the second time it never switched on again. The leds were telling me the story – power was there but power was no longer getting through the thermostat to the heater element.

It appears that the 18 amps consumed by the heater element overwhelmed the thermostat. Now, that may be because the thermostat has been under water and was slightly corroded or it may be that the heater element is partly corroded and consuming far more current than designed for. After all, 18 amps is a heck of a lot to ask a cigarette lighter plug to supply. It’s hard to believe the machine was ever designed to do that to normal cigarette lighter plug.

A replacement thermostat restored operation and gives me a reliable high-current load to test and calibrate current sensors.

The machine even brews coffee, though I think the high current consumption indicates that water has penetrated the heater element, corroded it and it is shorting it to the metal sheathe that houses the element.

The machine is now helping me design an AttoPilot+Arduino-based voltage and current sensor suitable for interfacing with the Nagios Renewable Energy Monitor.

Note those 20 Amp and 30 Amp-max appliance load currents mentioned at the start of this piece are loads created by CFLs, radios, laptops, and low-power servers. They are different – and lower – than the dump-load currents seen at the battery bank terminals when the dump controller kicks in.