It's a webserver powered by potatoes. Think back to the potato powered LCD clocks in the 80s - same thing.
Well, there was this conversation we had with a friend in the pub last month which got a little out of hand...
There's a website with details of its construction, or you can visit the server itself (warning: may not work for various reasons).
We're using a low power 386 motherboard. We don't have a hard drive or floppy drive, but have the operating system kernel, a simple webserver and a couple of webpages stored on ROMs. There are more details (and pictures) in the gallery.
We are using a minimal build of an old Linux kernel (1.x). The webserver software is custom built for our purposes (only understands HTTP GET requests, for example).
We know, and you're only the 800th person to tell us. As noted above, the webpages are on a ROM. We noticed the typos after we burned the ROM image, and we haven't got around to (read: can't be bothered to) make a new one.
We're using standard white baking potatoes at present, but we have experimented with other varieties. King Edwards tend to be too starchy for our purposes, but waxy potatoes such as Charlotte (or any other salad potato) work well due to the higher acidity. On the other hand, they are smaller and so don't last as long (dry out more quickly, electrolyte becomes exhausted). More salty potatoes, such as Cheshire new potatoes, are a possibility.
Depends on the variety of potato, the number of potatoes in use and the load. Using about a dozen baking potatoes (for cost reasons) we've managed to get uptimes in the region of twenty to thirty hours. This improves if we wrap the potatoes in clingfilm to prevent them from drying out, but we've not got past the thirty-four hour mark with the setup we're using at present (but we're still feeling our way around - more experimentation needs to be done).
Again, this varies considerably. Using the zinc/copper electrodes that we have at present, we get a voltage of about 0.8V (+-0.1V) and a maximum sustainable currrent of about 15mA. We can draw this current for about 15 hours before we notice an appreciable drop, so a back of an envelope calculation of total useful energy would be in the region of 650J.
Possibly, but we haven't really tried yet. If we used electrodes with a larger surface area, we should be able to increase the maximum current we could draw. Another possibility is that we could rupture more of the cell membranes in the potato to improve the flow of ions (and probably reduce the internal resistance of the electrical cell). Cooking the potato briefly in a microwave oven might do the trick, if it doesn't dry the potato out too much. It has also been suggested that we could mince the potatoes in a blender and decant the goo into plastic cups, but this feels a bit like cheating...
What about them? We doubt that they would improve the system, and we're a little ambivalent about the whole genetically modified issue. There's a lot of hype and sensationalism coming from both sides of the argument, and we think that an informed public debate is only going to happen when the media stops believing (and printing) everything people tell them and starts placing responsible journalism above circulation-boosting scaremongering. (soapbox over)
Not really. Spent potatoes probably contain quite a bit of copper and/or zinc, and so could be slightly toxic. If nothing else, they'd taste foul.
No. We're serving rather more people than we expected (over ten thousand a day at peak), and the server is just not able to handle that number of connections. Also, even though we try to reduce the time the server is down to change potatoes to a minimum, the server sometimes runs out at night when non of us are around. If you can't connect between about 8pm and 10am GMT, this is probably why.
Almost certainly. Citrus fruit, particularly lemons or grapefruit, would do the job well (the original potato clocks on which we modelled the spudserver were also marketed as lemon-powered clocks). Any fruit should do as long as it is quite moist and slightly acidic. Brussel sprouts would probably be too dry, for example. The main reason we chose potatoes is because they are innocuous and cheap.
Erm. Pass. Apart from being a potential waste of good beer, we're not sure that it would be acidic enough. Wine might work, but the build up of tartrates and tannic compounds on the electrodes could affect the endurance of the cell. Besides which, beer was involved in the (drunken) bet which lead to the creation of the spudserver, and see where that's got us!
Almost certainly not. As noted elsewhere, the spent potatoes could contain significant amounts of toxic heavy metals, so disposal is a potential problem (heavy metals might be leached out into groundwater if buried). We're still trying to work out the best way of disposing of the used potatoes safely - the best suggestion we've had so far was to auction them off on eBay as spudserver memorabilia!
Not really. Fleischmann and Pons had actually done some research, while we're far more likely to admit that this is all a big hoax. Besides which, we don't see this affecting world palladium (or potato) prices.
Only if the person to whom we were describing it was easily amused. We built the spudserver just to show that it could be done, not because we thought it was important.