http://www.tomshardware.com/us/200 [...] index.html

In Part 3 we focus on the hardware for the solar arrays used to generate power for our PC. We fretted over every half-watt to obtain the longest possible up time even on cloudy or rainy days.

What is DYI?

Do You Itself?

What a wonderful article! I learned so many things like that the sun rises in the east and sets in the west. Or that thicker cables have less resistence than thinner ones. Amazingly, the authors also tell us that green/yellow cables are more "efficient" than blue cables. These examples about sum up the fine quality of the writing even though, some subtleties might have been lost in translation.

I also fail to understand the point of the undertaking. It would have been a lot easier, cheaper, faster and logical to just use some notebook - may be with a red keyboard for efficiency...

Not to mention, who cares about the power savings when you're paying 4 grand for a low end PC?

Pointless.

Agreed! The whole thing is dimwitted and quarterbaked. Let's see:

With an assumed average consumption of 90W for the computer and the display, the solar maximum output is just about 4 times the load. Considering that it you never get the maximum output in wintertime and that there will be only about 8 hours of usable daylight (if at all), the system couldn't stand a single day in January. No way.
Running on battery alone, the system wouldn't survive much more than a day. What's the point of a battery at all - especially considering that a battery has a limited lifetime (5 years or ~$100/year or about $0.27/day or about the same as the potential cost of buying the energy from the power utility).

It is a great idea to use powerefficient equipment including computers and using clean, green energy is fine, too. But using solar energy in high latitudes for 24/7 computer operation makes as much sense as using hydro-electric energy in the Sahara for a space heater.

Ditto. Nevermind lack of research into DIY solar in general. A couple of issues of Home Power magazine would have done wonders for their setup. Turntable tracking is a joke. If you're going to add complexity and failure points of solar tracking at least do two axis perpendicular. How many days until that OSB, rope, and hand drill setup dies? And how many days did it take to build a system that is inefficient and prone to failure? Drop tracking altogether and just add another cell. And drop the battery and direct 12v power supply. Run grid-tied with a laptop if you want an efficient solar PC.

Don't forget to add something that will shovel the snow off the cells in the winter.

In some places you can sell your power back to the power company... that would be cheeper then the battaries.
Most nights the system would be running idle. So it wouldn't consume as much power.

how many solar panels would you need to run a high end overclocked gamming system.
Most people reading this would have high end systems.

my 2¢

Lets not be too mean here...

There is translation so the green/yellow cables are more "efficient" than blue cables does have a picture clearly showing the larger wire is whats efficient.....

My question is....Looking at the live stats the pc seems to go down allot. Is it running out of power at night and shutting down?

Worst case you still have a super efficient(not for games...but still ok for most other things....) PC to use with the power utility....

Well, the green/yellow cables were just an example for the overall "journalistic" quality of the article. I could have chosen amongst at least 20 other examples.

Of course, the system goes out a lot. The battery and the solar panels are too small to support 24/7 even in September. Or may be the wooden panel frames are just shaking in the wind. In other words, your own observations are proof that the autors did not (!) succeed "in building a desktop PC system that runs completely and exclusively off solar power".

In the worst - and obviously existing case - the article totally missed - amongst others - its title ("DYI Solar-Powered PC Solar Components" ) - even though, come to think about it, I have no idea what the title really means.
And I would challenge the claims about super efficiency. They are not substantiated; any similarly powerful notebook (like C2D with a mobile graphics card) would be running with less than 40W of average power consumption and it would be at least equally powerful and probably a lot cheaper. The rediculous explicit claim of "world record" may justify a little sarcasm. It also just distracts from the importance of power efficiency and usage of renewable energies.

I wonder if wind would have done better?

You would need to live in a windy area to begin with and a giant wind generator will probably be more of an eyesore than the rigging for solar panels.

The article is a decent concept with not much practicality. Even the author stated that.

Something more practical would be to set up a solar cell array to supplement power to a house. Thereby reducing dependency on your local electricity provider. But that's beyond THG's scope.

Project lifespan??

I am sorry to say, but I do not think Your experiment will last very long

First thing:

Your power budget is 534 KWh annualy if the computer is supposed to run 24/7.

1 KWp (One Kilowatt of rated solar panel power) will in Denmark yield 800 KWh per Year. You live further south, so I will credit You another 200 KWh. If You have the same Solar input all Year round, (which is impossible unless You live in Africa near the equator), You will still need at least 600 Watts of Solar Panels.

Next thing, Your Storage Battery.

I am really looking forward to see to how long You take to run the battery completely down.

Lead Acid batteries made for Solar Power or Marine use, can withstand a discharge down to 50% of full charge without damaging the battery. Some batteries even alows You to go to 30% SOC on occations, but not always.
If You discharge below 50% SOC (state of charge) You will shorten the lifespan of the battery. Discharge it completely and You will cut the lifespan in half in one go.

So, If You want Your $500+ battery to live much longer, You need one more battery equal to the one already obtained. Next, remember to set the charge controller to Solid GEL, if this option can be selected. Solid GEL Batteries cannot be overcharged like ordinary lead acid batteries.
I trust You obtain the consumption through the charge controller. This way You may save Your battery's life because a decent solar controller cuts off consumption before the battey discharges too much. Regrettably I am not able to find Your charge controller's data.

Please don't be offended. I am sure You tried to think of everything, but You are obviously new to solar power.

Best of luck.

ChrisD

I think some people r missing the point. a computer that runs 24/7 completely off the grid. where you only option might be a gas generator. in remote locations has huge merit. although i think a laptop with solar might be a better choice. i am thinking 3rd world.

The battery has been hovering around the 12V mark for the past few *days*; it looks like as soon as there is enough energy to run the computer, it's turned on without regard to the battery's state of charge (whereas every effort should be made to prevent the battery from going below 50% SOC). The $500 battery will die from sulfation if it never gets recharged--as it looks like is quite possible with the usage pattern.

Above mentioned practical flaws aside, as a concept article (however poorly executed), I think it served its purpose. We can all think of practical improvements, from adding another panel or two while forgoing tracking, to increasing battery capacity; but if you needed to run a computer in a remote location where power was an issue, yes you would use a laptop. While parts of the article (perhaps as translated) are downright silly, it is still thought provoking, and I enjoyed reading it.

I agree with you, jtt283, the article, project, and concept are downright silly. Nothing, but really nothing makes any sense. However, I enjoyed reading it. I laughted a lot.
Yet, I am not quite sure if THG was the right place for publication. May be "Mad" would have been better.

More seriously: Are there any conceivable reasons for running a desktop (not server) 24/7 without putting it to sleep mode when unused?

Like pretty much everyone else is saying, good idea, bad implementation.

I would have definitely used a laptop, but, if I was going to do a desktop system, I would have definitely used an underclocked processor, and 2.5" hard drive. Since you wouldn't reasonably be using this system for anything too extreme, an underclock and undervolt would be reasonable, and lower power consumption a LOT.

2.5" hard drives aren't exactly hard to come by either, and it's not like they'd be storing a ton of data, so I think one of those should have been used as well. The argument that they wanted to use off the shelf parts doesn't work, because the PSU wasn't either.

As for the battery issues, you guys are all pretty much right in saying that that battery isn't gonna last long at all, a little more research in that department would have went a long way.

I think the concept of a computer running of nothing but sunlight is pretty cool, but it's completely unreasonable to ever use in any circumstance (at least until solar cells get WAY more efficent and way cheaper.)

Sapit, I leave my computer on 24/7 simply for convenience, and in case anyone wants to leave me messages via MSN. It's hardly costing me any money and it's near silent, so why not? I do underclock it overnight and when I'm gonna be gone all day, at 400mhz and 0.9v the cpu pulls about 4w full load, the hard drives shut off after a certain amount of time, monitor shuts off, etc, so it's really not pulling much power. This may also sound strange, but I can't sleep without the ever so faint hum of my computer running, without it on it's just too quiet and I can't sleep. Doesn't make much sense but yeah...

400.....what cpu is it....i cant get down that low....

think 1200 is the lowest i can go....never tried....but 800(QDR) fsb is as low as the board goes to my knowledge....

so 6 x 200 = 1200

from the looks of the system thay need 2 more solar panles and maybe 2 batts and that should be able to keep the system going for longer

good try tho

my server gives me my hum to make me sleep (2 to 6 of the disks spinning away )

-------------------------------------
UK THG web site sucks as thay are not linking to forums i only found this by some how been able to use the USA web site

None of the UK main page stuff links to forums thay have an comment system that no one uses

All main page stuff should go into its own thread ( like bit tech does http://www.bit-tech.net/ that web site works very well)

You guys don't get it. You should be laughing at yourselves. As clearly stated throughout the whole series, this was a test, a concept, a what if scenario approached from the POV of the average Joe faced with what is available out there. How would we ever know if this is or isn't a viable solution if no one has the balls to see for sure.

I can see the the limitations now and the complexity and difficulty in getting it all put together. I can now see that its more expensive than I would like it to be and that it would require more money to make it stable and longer lasting.

But I would never have been able to reach those conclusions without these two guys going out on a limb and building it for us. So ya'll can just STFU, unless you have something better to show us.