Andy's bright ideas.

We are investigating having solar panels to generate electricity fitted on the roof. I am administrating the project at present, and I am intending to get quotations from three companies, and then pick what seems to be the best value for money. That is likely to be the lowest cost, but only if other important factors are comparable.

For example, comparable means that the peak power specified is the same, or very close. Because the maximum size of installation eligible for the domestic feedin tariff is 4kW peak, we want an installation as close as possible to that figure. Spirit Solar, the first to quote, offer 3.92kW.

There are several factors I think relevant, and I'd be grateful if anyone reading this could tell me of anything I have overlooked.

First of all, adding solar panels must not damage the house in any way. Spirit Solar said they would survey the roof, and recommend work to strengthen it if necessary before fitting the panels. Such strengthening work would be guaranteed for 50 years.

They also fit the panels spaced from the tiles, on brackets screwed to the rafters, fitting between one tile and the next so that one of the tiles just has to be ground down to allow it to fit flush. That would seem to me to be as unlikely to cause damage as reasonably possible, but what to *you* think?

Secondly, the installation must make financial sense. I have been very unsure on this point in the past, but I have a draft spreadsheet, soon to be updated with real figures, to compare the return on investment of buying a solar installation, or keeping the money that would have been spent in that, in the best ISA rate of interest currently available. I intend to update the spreadsheet and make it visible on Google docs, but in broad terms:

1. I assumed that interest rates rose a little in two years time, and that the increase of the price of electricity rose faster than the rate of inflation figure. (It does seem that prices are rising much faster than the figure given as the rate of inflation would lead one to expect - at least if one trusted government statistics.)
2. I assumed that money received for the feedin of electricity would be saved in an ISA every year at the same rate as the money saved in the notional comparison ISA.
3. I used the government approved figures, widely claimed to be slightly on the safe side (good!) to calculate the possible feedin tariff income.
4. I assumed that it will be necessary to fit a replacement inverter during the life of the installation. The one that Spirit
   Solar fit is guaranteed for 10 years, so one might hope that two of them would last for a total of 25 years.

On sample figures that I believe to be close to reality, we would expect to have equal sums of money in savings by the middle of year 11 from the date of installation. I would have expected to work for at least 8 years from now in any case, and having the savings available to draw on is less important while working than after (and if) I retire. Since nice Mr Brown has taxed away a proportion of the pension I might have had, working for at least 15 years if it is possible would seem to be required.

The feed in tariff is set to continue for 25 years from the date of installation, and that, coincidentally, is approximately our statistically expected lifespan.

It is possible that we may wish to move house before the breakeven date, though that could be so disruptive of my work (for anyone who doesn't know, I work as a freelance electronics designer) that it might be very tricky. If we did move house, it is possible, but not certain, that we should get enough extra for the house to make it worth having fitted the solar installation. What do others think?

Then there is the question of how much extra the house insurance would cost. I haven't got a figure for that yet, and I suppose it is barely possible that there would be no increase in cost, but that could increase the payback time or even make it not viable. I need to investigate.

On one hand, the government may try to rescind the commitment to keep paying the feedin tariff, or change its inflation linking to make it less viable, in future years. I don't think that is very likely, but they might. On the other hand, the present energy policy seems set to give us power cuts, in which case there might be an official change, or a hack, to make the solar energy power the house even when the grid is switched off in an area. That will be of *some* help, though the times it is most likely that there will be long power cuts is in winter. Last winter there was a 3 week period of little or no energy generated by the windmills, just at the time when there is least daylight for the solar panels.

There is also the possibility that, if enough people install solar power, then power cuts during the day may be less likely. This is a benefit we may not ever be able to assess, but if it did make a difference it could lessen the economic slump that such power cuts would cause.

Can anyone think of any other questions I should ask myself before taking this costly step? I have asked for three quotations, (Spirit Solar, recommended by moneysavingexpert.com, SolarMerge, who did a solar power installation for Geekette8, which she said she is pleased with, both for how the job was done and its functionality, and Solartricity, who fitted a solar power installation for my half brother. He also seemed happy with how it works, and with the efficiency with which the fitting was carried out.

I spent today at the Texas (formerly Unitrode) power supply design seminar. The UK one this year was held in the motorcycle museum near Birmingham. It was at least a floor below the slightest phone signal, or glimmer of daylight.

From my point of view it was the best such seminar I have attended in at least the last ten years, because they covered two topics that I want to use in my work soon, and others that gave me useful extra understanding and analysis and modelling techniques that are likely to be useful frequently.

I'll try to avoid making everybody's eyes glaze over, but the part of my work I most enjoy is where the laws of physics meet the product. In my line of work, that normally means designing switched mode power supplies. Energy efficiency has always interested me, and a major design aim for switched mode is efficiency.

One useful feature of some types of smps is that they can provide more than one output, by making a transformer with extra windings on it. For example, the power supplies used in most desktop computers provide +5V, +12V, and -12V. Hear the wisdom of ancient Plautus! It is well known that if you draw lots of power from the main output of a power supply and none from the other outputs, those outputs normally provide much more voltage than expected. A 12V output could reach over 20V. It is easy enough to measure to what extent that happens, and then determine if remedial action is needed, once the power supply is designed and assembled. However, it is helpful to be able to calculate what will happen in advance, and, perhaps, calculate that simply adding a small LED to indicate that the 12V supply is present will draw enough power to prevent the voltage reaching an excessive level.

The first session covered a way to model transformers having multiple windings, and calculate this aspect of what a power supply will do before one is built and tested.

All the other sessions were at least interesting and thought provoking, but the case study of a solar powered lighting application demonstrated just how much better power supply design can be if, instead of analogue circuitry such as has been used for decades, you use a digital controller and make it all work using software. I expect to start designing something with a Piccolo DSP very soon. Some aspects of it may initially be difficult and confusing, but once I have learned what is needed I will be able to design some items a useful amount better. Cheaper and more energy efficient - which my customers always need. I await an evaluation kit.