Blog

The Digital Agency for International Development

Solar Diaries - Day 7 & 8 - Deltas

By Alan Jackson on 02 October 2010

retrospectiveWhen we do retrospectives at the end of our coding iterations we look at what went went well (the "plusses") and what we'd like to change for the next iteration (the "deltas"). It's a much more positive approach to think of deltas rather than "what went badly" which can turn in to an excuse to winge.

In that spirit I thought I'd talk about some of the deltas of the solar power experiment.... nah, forget it, I'll just winge!

Last week I was travelling about a lot so didn't really work off the solar supply. Which meant it had at least a good 4 sunny days of charging before I started day 7. At the start of the day the battery voltage was 13.2 Volts, the highest it's ever been. We ended the day at 12.7V. Day 8 started at 12.7V ending at 12.1V. Both days were very cloudy, so you'd expect the load to take more out of the batteries than the PV panels replaced.

When I run my typical setup, an IBM x41 tablet (running Ubuntu, of course), the iiyama 24" monitor in eco mode and charging my mobile phone off the laptop USB port, I'm taking just under 5 Amps (about 60 Watts). The battery, if full, should be able to supply 5 Amps for about 20 hours, or about 2.5 days. The recommendation (considering the UK's propensity for cloudy days) is to have 4 days of reserve.

The first wingy-deltas are that the battery is under-dimensioned for my load. It's been pretty sunny so far and I still ran out of power a few days ago. I suspect that the PV panel is also only half as big as it needs to be to replenish my load, even at this relatively sunny time of year.

When I run the 24" monitor and laptop from the 12V supply I get buzzy stripes over the screen - they're not so visible when looking at things that are mostly white, but on darker backgrounds it's a real issue. If I unplug the laptop from the 12V power supply the lines disappear. My laptop has only got an analogue SVGA connection and I suspect I'm getting some kind of "earth-loop" effect picking up mains hum. The whole solar power system is currently un-earthed (is this wise?). Or maybe it's something to do with the laptop power supply dirtying up the signal for the monitor, although I swapped the supply for the Maplin's generic DC-to-DC converter/regulator and the problem was identical. I could  partially test this hypothesis by plugging the monitor into the mains and seeing what happens. I also seem to remember at "Small Is..." we were getting the same issue which went away when we used a digital signal cable instead of analogue, an option I don't have with my laptop.

When the IBM laptop is charging (ie drawing a large current) and I put the PixelQi/Samsung netbook on to charge as well then the IBM starts making a terrible buzzing sound and the battery state starts falling rapidly - in about 5 minutes it will go from nearly full to half empty. I don't know if this is causing permanent damage or what's going on. The laptop has also buzzed sometimes when the monitor is on full brightness. This might have something to do with trying to pull too much current out of the battery when it's in a low state of charge. We're speculating that the laptop power supply might be going through some rapid cycle of trying to draw current from the battery, the battery voltage dropping and then the power supply shutting down abou 20 times a second. That's just a wild guess.

The battery has never yet been fully charged - the solar charger has not yet shut down charging because the battery was full. Perhaps I should have started this experiment by fully charging the battery before putting any load on it, but I was too eager to start.

I need to permanently fix the PV panel to something so I don't have to keep taking in at night and putting it out during the day... like some bad cat. I also need to drill a hole in the side of the house to run the PV cables through so I don't have to keep running them out window.

I'm on mains power today as I'm pair-working with Nate from the Skiff. I haven't yet got a battery solution for working from the Skiff. Chris has warned me that charging a lithium-ion battery can be tricky and if you get it wrong they explode. However Ni-MH batteries are much more forgiving and similarly light in comparison to lead-acid. I'm wondering if it's better to get a bigger/newer lithium-ion for the PixelQi/Samsung to get a full day out of it and get an Ni-MH battery just for the pair-working iiyama screen (about 15 Ah of load) or get a battery that will cope with my normal daily load (about 40 Ah of load).