I don't really understand why I didn't try this before. Its simple. its recommended and it works. Have I been too focused on flexibility why simplicity could be the answer?

I was speaking to Chris from Euro-Line, an importer of consumer solar products and he highlighted a document that I'd already seen. I took another look and staring me in the face was a recommended and tested solution using equipment that I already have. Its the same setup that I tested with the P3 panel. Just plug the panel into the Tekkeon MP3400 and wait for enough sun. You might remember back in the early posts that this is how I found out that Li-Ion charging solutions where not so efficient and how it set me on the path to research a more flexible solution.

With the '12v' 25w Sunlinq panel I have I assumed that a 12V output wouldn't drive a 19v input and after looking at the diagram again I though 'why are they recommending this solution? It shouldn't work.'

Image taken from GlobalSolar.com PDF here.

Looking more closely at the specs of the panel, its clear now why it works. The 12 panel isn't strictly 12V. The voltage varies according to the load and in fact with an open circuit the voltage is way up over 20 volts. However, with a load of around 800mA, the charging current for the power bank, the voltage sits nicely at around 20V. Tada!

With a 25W panel, 800ma at 19V is reached at around 60% sun power. On a clear summer day here, the sun is over 60% power for around 5 hours between 11 and 4pm. The charger needs 4 hours to load up 56W of energy.

Now here's an idea. Between 12 and 2, the panel is producing 40% more energy than the Li-Ion battery is taking. Can I mop that up with a lead-acid battery?

Testing continues...

In a recent comment here, someone asked why the Lead-Acid battery was needed. Its probably not too clear in the video why I use it so I reproduce my answer (which comes from the best of my knowledge!) here.

There are two main problems with charging Li-Ion batteries from Solar panels.

Firstly, Li-Ion batteries (in notebooks and battery bank) charge using a constant current (stream) of power. For common notebook batteries and battery banks such as the Tekkeon MP3400, this is around 1A. A lot of the 12V portable solar panels only reach this power at high sun levels meaning you can only use them for a few hours mid-day. In fact a 12W panel might not be enough to even start the charging process. Secondly, if you have a huge panel that could deliver, say, twice as much power as needed, its not used. Only the power needed is taken. The rest is wasted.

These two problems can be overcome at the expense of weight with a lead-acid battery.
L-A batteries are more flexible. You can charge them with a trickle and also with a higher charge rate. They are much more suited to pairing with a solar panel. The problem with this solution is weight. Small 12v L-A batteries are over 2KG in weight!

What's needed is a flexible Li-Ion battery charging solution. Currently there are no products on the market that can archive this but I'm searching hard!

In summary there are 2 solutions.
1 - Get a panel powerful enough to charge a Li-ion battery at 70% of its rated output. For example, a 25W panel and the Tekkeon MP3400 Li-ion battery. This will give you about 4 hours of charge time on a sunny summer day.  (Mid-Europe) This should be enough to completely fill up the Li-ion battery.
2 - Go for a heavier solution with a L-A battery and give yourself more charging flexibility.

The  latest weather report is still bad. No chance of leaving in the next 10 days unless the weather report is wildly wrong.

Its unbelievable! We  had the hottest, driest  April ever and since then the clouds and rain just won't go away and summer storms are threatening almost every day. Looking at the sun graph for the last three days I doubt I could have got get even a few hours of computing time out of it.

It gives me more time to set up more tests and products of course, but I'm really keen to get going ASAP, especially as I have a holiday booked with my wife on the 19th July. The last possible leaving date for the tour would be 8th July. After that point I won't be able to fit it in and will have to wait until 28th July.  This is really frustrating. I've been planning this since late 2006.

On the upside, I tested out the bike with full baggage tonight. It was just a 10km round-tour in Bonn but it was a good test. I'm hoping to squeeze in one nights camping this week so that I can test all the equipment out together.

I've also heard that there's a possibility to get a Reware Juice Bag. That makes me happy! 

Tonight I'm doing what I've been doing most night for the last few months - researching solar power and solar notebook/laptop solutions. Tonight the Google search subject is expeditions. I figured that these guys must have solar computing pretty much sussed by now but I'm struggling to find a solution that can improve on my, 'draft', solution.

For a start, many of them are using lead-acid batteries, the most dense battery solution around. Is it because they are cheap , rugged and flexible I wonder? Why aren't they looking at lighter Li-Ion solutions? And then there's the panels.  All the expedition panels I've seen have been non-flexible  solutions. Is this because  the efficiency is  better than the flexible (CIGS) solar  solutions? That's actually quite likely and I'm wondering if I would have been better off with a solid panel solution, especially after I just spend two hours making a frame for my flexible solution!

The biggest shock came when I looked at the PC solutions though. None of the websites I've looked at  attempt to  advise on  the  most  efficient laptop solution. Those that do seem to be incredibly out of date. "Laptop computers consume between 15 and 30 watts of power (some as much as 60 watts)" said  one.

It makes me wonder weather I could actually help some of  these expeditions.  I wonder what they  are using the notebooks for? Is it just  emails, logging, weather, mapping solutions, inventory management? If so then I think a UMPC could really help.  Not only  are they much lighter and smaller than your average notebook (1/3rd the weight  which could  save 3-4KG for  an expedition that  is taking two notebooks.)  but the excellent power efficiency would mean knock-on savings  with battery weight and solar panel size.

If you're responsible for the PC solution on your expedition, please let me know what your set-up is and what you're planning to use it for. I think I might be able to help you cut your weight or increase your computing time. For example, the very efficient VIA C7-based Samsung Q1b, retro-fitted with a solid state disk drive could average less than 9Watts of  power and give expedition team members enough time for a lot of extra emails home!

After I've finished the solar-umpc tour I'm planning to take the knowledge gained, continue to add to it and then try and design an ultra-efficient solar-computing solution. Well at least I'll try! I have been speaking to Select-Solar about this and they're also keen to find out more. Maybe together we can take solar computing to the next level and make it available in a cheaper, lighter and more efficient package. I think that Ultra Mobile PC's are going to be an important part of it because battery and solar technology seems to be advancing very slowly compared to the efficiencies of Ultra Mobile PC's

Digital Donkey image courtesy of Alex Papadimoulis.

It was an early start yesterday. 0445 rise and bus at 0530. 5 hours later after three trains and another bus I arrived at InterSolar 2007 in Freiburg, Germany hoping to find out how I could improve the UMPC Solar kit at Europe's biggest solar expo. Unfortunately, there wasn't much there for consumers at all. It was all 200w, $1000 panels and huge thermal heating set-ups. I saw one company that was selling the Voltaic Solar Backpack but I've already assessed this product and at 4W max output, its not powerful enough. I was rather hoping to see someone with the Reware Juicebag which, at 6.3W is much more useful.

Actually the most interesting thing that happened yesterday, apart from some awesomely stormy weather, was that the train journey took me on the Rhein route south to where I will start the Solar UMPC tour. Its a gorgeous route. Lovely scenery and I saw loads and loads of great Rhein-side restaurants and camping sites. I'm looking forward to the tour more than ever!

One bit of relevant news from yesterday which came through RSS and has spawned another bit of research was that Sanyo have broken the record for a production solar cell. We're up to 22% now. In theory, the cell on the left here, a 10x10cm device, should be able to generate over 2W. 100cm2 is about the same area as the face of a UMPC. 4W is the target average power consumption that Intel have set for devices based on Menlow, their UMPC platform for 2008. The interesting thing that I've found out is that this cell (or at least the previous version of it) appears in the Sanyo Eneloop solar charger. The charger houses a Li-Ion battery which can store enough energy to charge 4x2000mah batteries. That's about 10W if my maths is correct. The only problem is the quoted 6 DAYS charging time for the internal Li-Ion battery. I suspect that the cell isn't exactly being used that efficiently because as I said before, that 10x10 cell should be able to kick out 2W, enough to charge the batteries in a few full days of sun. This little bit of tech will set you back over $150. Eek!

To know that there's a commercially available solar-powered mobile phone gives me a lot of inspiration and encouragement that one day in the near future we'll be able to design UMPCs with infinite standby capability. You could have a panel and electronics that senses light levels and keeps the device in standby (or even turned on) when there's enough ambient light. When the light levels drop below a threshold for a certain amount of time (based on remaining battery life perhaps), the device would drop into hibernation.

The mobile phone shown here is produced by HiTech Wealth and is available in China for $510. Full news story at SolSie.

Could the weather forecast be any worse? Not only will I get wet and be unable to test the solar kit fully but with an overnight low of 11 degrees, camping wouldn't be much fun.

I learned a long time ago that WiFi Internet access is not a truly mobile Internet access solution. Its a solution for multi-location access but not a full mobile solution. Fortunately the mobile telephone carriers offer Internet access too and it 'aint bad!Its going to be the perfect solution for the Solar-UMPC tour.

In Europe, we use the GSM and UMTS (3G) methods of mobile phone communications and they are well advanced. 1.8Mbps is common in and around most cities and towns with 3.6Mbps being available in the most populated places. Even out on the fringes, a 384kbps connection is possible so in terms of availability, there are very few issues.

Issues arise, however, when you start looking at equipment and cost. In order to use a 3G data connection, you need to have a 3G data modem. There are three options. Either you buy a dedicated 3G modem that you can connect to your PC, you buy a PC or notebook or UMPC with a 3G modem integrated or, if your contract allows it, you use the modem that's part of all 3G mobile phones. In this case you simply connect the phone via USB or Bluetooth and in general, its an easy few steps to get connected. It really solves the costs of hardware because in 2007 we've seen a number of high speed 3G-capable mobile phones hitting the market and with the normal contractual discounts, the costs is minimal.

It doesn't solve the other cost though. The cost of the data. This has been a real issue for early adopters of this connectivity method. Costs of data can be shockingly high and for the average user its just wasn't worth it. Until, that is, the advent of the data-focused contract. Nearly all the major cellphone carriers in Europe are offering data-oriented connectivity and costs are coming down extremely quickly. In the UK and Finland, 10-15 Euros per month is all it costs to get hundreds of megabytes, sometimes even up to 6GB) of data per month. For the average user that's is using just a few hundred megabytes a month on their normal broadband connection, this is more than enough. In Germany, I have chosen the Vodafone Data Tariff. Vodafone, along with T-Mobile have the widest coverage of high-speed enabled cellular towers and a good selection of data tariffs. You can start with 30MB per month for 10 Euros and go as high as 50 Euro per month for 5GB of data. In the middle are a couple of attractive 200 and 400MB tariffs and the 200MB tariff is the one I've gone for the WebConnect L at 20 Euros per month. I could have got a cheaper tariff but if mobility is an important factor, cellular coverage has to be high on the list of requirements.

Insert SIM card for instant connectivity! Samsung Q1b HSDPA

Combined with the Samsung Q1b HSDPA I've got a 'real' ultra mobile solution in my hands. Vodafone have given me a second SIM card for the UMPC and with the built in HSDPA modem I'm online with broadband speeds within seconds of turning the UMPC on. This is how it should be!

They are expensive and inefficient. Solar panels need to develop a lot before they can provide the average man on the street with enough energy to supplement a worthwhile portion of his energy intake. Mind you, the average man is pretty inefficient. Cars, lighting, heating, cooling. We sap up energy like there's no need to think about tomorrow. I'm sure the average person could cut their energy usage by 90% if they had the skills, time, money and resources to do it. At that point, solar power might be a more interesting proposition.

Mobile computing is one of the areas that is already cutting its energy requirements buy huge amounts. Where notebook PCs can take 20-30W, ultra mobile PCs are taking 10W and in 2008 when we start to see dedicated ultra low power, ultra mobile computing devices, that power requirement is going to get slashed down to the sub 5W level. Intel is talking about sub 4W and VIA have just launched the 3.5W Mobile-ITX board. Of course, the reason for this is not because of energy conservation, its because of heat, size and weight. Mobile devices need to be small and light. Small devices can only dissipate a certain amount of heat and light devices can only contain a certain amount of battery. The last thing we want is a UMPC melting through the bottom of our bag!

One nice side-effect of all this drive for tiny, light devices is that solar energy starts to cash-in. Smaller energy requirements means smaller solar panels which means lower costs and higher mobility. Right now, today, if you took the best solar technology and made a fold-out sun-shade for a UMPC of about 20x20cm, you would be able to power the UMPC non-stop during sunshine. Its true. The best solar tech is reaching 40% efficiency in the lab and at that rate a 20x20cm panel would provide up to 40W of power. Unfortunately, this is lab-tech at the moment and good quality, mass produced solar panels are down at the 15% efficiency level. Also add in the fact that many places don't get much sun and you can see why its not really possible today.

But lets take the Reware Juice Bag as a good example of a useable bit of solar tech. It can provide up to 7W of power from a 20x30cm panel stitched into the backpack. That's about 25W of stored energy per sunny day if you take into account losses on storage. Today, that would give you about 2-3 hours UMPC usage but with a 5W UMPC, you're looking at 5 hours usage. Drop down to 3W (2010 perhaps) and 8 hours of battery life starts to get really useful. Lets assume that by 2010 the efficiency of Solar panels in increased by 50% and you've got full-day power from a panel that's the size of a UMPC.

OK, its not 100% practical for everyone to be sitting outside looking for sun but it would work for a lot of people. And its just one example. How about taking the same calculation and scaling it up to a classroom full of UMPCs. You could power a full class of 30 with a few square meters of solar panel. Costs would be sub $1000 and if you compare it to 30 normal PC's taking 150W per hour for 4 hours per day. (30 x 150 x 5 = 1.8KW) you are not only helping the environment but you're reducing your air-con costs, noise and starting to get to the point where you save significant amounts of money. For a university that has 2000 PC's deployed, well I don't need to show the calculation do I.

After the solar-ump tour is finished I'm going to be following solar tech very closely on this blog and will continue to promote the marriage of ultra mobile PCs and solar tech. I don't see the battery companies shifting towards more efficient battery technology so as solar energy starts to become a really interesting business opportunity we might start to see solar being integrated into specialist UMPCs for long-term outdoor usage. At some point in the future, when the calculations look attractive to the bean counters of this world, investment money will come pouring in to the mass market and the whole thing could take off in a very short time span. I predict, with excitement,  that within 2 years we're going to see solar accessories for UMPCs. Within 3 years we'll see specialist UMPCs with integrated solar panels. I'm hoping that in ten years time, the extended battery will be a thing of the past.

I have been looking all over for a lightweight table that I can use on the Solar UMPC tour (which is at risk of being pushed back by another week due to weather and PR organizational activities unfortunately) but I realised at the weekend that I already have a table.

Its a heavy bike but it has its uses! Since taking these pictures I've fitted a center stand to the bike so that it stays level. All I need to do now is to find a stool. I've tried a few three-legged walking stools but they're not comfortable. I'll try and get a folding camping stool sorted out.

Now all I need is a little beer bottle holder on the side and I'm all set!

Kornel, a member of the UMPCPortal.com team, asked my why I wasn't using a solar backpack on the tour. He was referring to products like the Juice Bag from Reware and the Voltaic Solar Backpack.  [update: And someone has just told me about the Clear Blue Solarpac Series. Another option. Thanks Jon.]

Reware Juice Bag backpack

I had pretty much dismissed the idea based on the fact that I didn't really want anything on my back and I thought I had all that I needed in my 25W solar panel. However, I've had some other thoughts in the last 24 hours mainly because the big 25W solar panel is proving impossible to deploy in a 100% open configuration on my bike. The other thought I had just 30 minutes ago was that if the 25W panel breaks, I'm finished. I need a back-up. The final though I've just had as I sit down here is that I need to test as many possibilities as possible out so that I can, at the end of the tour, be in the best position to advise other people what combination of products is the best for a certain situation.

Voltaic Solar Backpack

The Juice Bag looks like the best option as it provides 6.3 watts compared to the Voltaic's 4 watts. It has a 31 liter capacity (could be good for some light clothing) and is waterproof. It certainly sounds like it could supplement the kit very well so I've decided, just 5 days before I'm due to start, to try and get hold of one to test. I'll see if I can contact Reware to get hold of one under some kind of sponsorship deal where I report back to them how it worked out and whether its worth them promoting them for UMPC usage or even creating a UMPC-focused solar gadget bag!

Thanks Kornel, for making me think about this as a possible solution!

Encouraged by yesterdays progress with the solar kit, I spent most of today finishing off the wiring and packing it all into a plastic box. I then paired down my kit list somewhat (out goes the Nokia N800 and spare mobile phone, two pairs of underwear, one of the t-shirts, the multimeter and a pair of trousers) and loaded everything up on the bike to see what sort of weight I was dealing with. The total wight of the bike and kit together (that's biking, camping and computing gear) comes to 55KG. I have some food and bits and pieces to add to that so lets call it 60KG. Considering the bike weighs 25Kg that means I'm going to be carrying a 35Kg office. I took the bike for a ride and it seems pretty stable although I am a little worried about brake failure. I only have a back-pedal brake and while its very good I'm in serious trouble if it fails. I'm thinking about getting front brakes fitted this week. I also notice a small S in the back wheel. Its out of alignment and I hope that it can be brought true again by a local bike shop. If not, I'm probably in trouble and might have to switch bikes. I don't want to do that as I really like FK398. Its been a work horse for 4 years and it deserves to be in the tour. It also looks great. I love the retro style.

 
About 90% loaded. The Kronan feels OK to ride with 35KG on it.

Another problem I really need to think about is the positioning of the solar panel on the bike. I can easily deploy 50% of it across the rear baggage but that's not really going to be enough. I should have tried to find two separate 12W panels and mounted them front and rear but its too late for that now and I'll have to work round it. In this part of the world, the sun gives you about 4 full-power hours per day. With a 50% deployment of the panel, that's only 48W/hrs. Short of the 60W/hrs I think I need per day. However, what I could do (and I really like the sound of this idea) is take a 2 hours working break between 12 and 2. This should give me close to 50W/hr of energy assuming the sun is shining. Outside these hours I might pick up another 10-20 w/hr so that's 70W/hr of energy from the panel in a day. It looks like its possible and I will test the theory this week.

Under the seat is the battery and electrical kit. It weighs something like 5KG but packs an 80W/hr lead-acid battery and a 56W/hr Li-Ion battery. Enough to take a full days energy from the solar panel. 

There's about 5KG going on the front rack. Its not attached to the forks so doesn't affect steering as much as panniers would. I'm thinking of using it as a computer table in the evening but I can't find a good enough stool. 

Mounted on the front handlebar are the Garmin Etrex GPS logger and a Plexiglas map holder. I will get some rain covers and more straps for the equipment. The Q1b UMPC sits in the right hand pannier its its organizer case and wrapped in clothes. I will have to be careful about making sure the device is in hibernation before I ride. If its in standby and then goes into hibernation it will have to start-up the disk. I don't want that to happen while I'm riding as it could be fatal.

Before I talk about the video I just want to say "where's the damn weather!" Lesson number one in the Solar-UMPC project is that the weather is chaos and if you live in mid-northern Europe, you can't rely on it as a source of instant energy. You'll see how I've had to adapt my solution in my first Solar UMPC video. In the video I present the solar charging and storage setup that I've decided to use for the tour. Its based on the fag-packet drawing I did a few weeks ago after deciding that Li-Ion batteries weren't really the best solution.

I've introduced the lead-acid battery as a buffer and despite the 100-year old technology and 2.6KG weight, provides the perfect stabiliser to the whole architecture.

A quick rundown of the equipment I'm using (mostly bought from my home country, Germany.)

• Sunlinq 12v / 25W foldable solar panel.
• Sonnenschein 12v / 6.6Ah lead-acid battery.
• Solar charge controller. 12v / 4A
• Samsung Q1b UMPC.
• TabletKiosk MP3400 Li-Ion battery pack.
• Voltcraft dual-mode 12v battery charger.
• Voltcraft 12v DC-DC adaptor.

The 20-min Video is below (800kbps divx, 120MB) and also downloadable here. If you can handle a 1.5mbps stream or 200MB download, take the WMV version from here. YouTube version also below. More pictures in the gallery here. Blip.tv link here for sharing. Many thanks to SelectSolar for their help in preparing this solution.

Just like all the best countdowns, this one is paused for a while. The reason? Taiwan. Both VIA and myself have been non-stop busy since Computex started 4 days ago and we've decided to delay for a week. Looking at the weather forecast, its probably a blessing in disguise because cloud and light rain is forecast for much of next week.

Watch out for a video overview of the Solar equipment coming up either later today or tomorrow.

Despite the terrible weather forecast (rain, clouds, heat, thunder) It's time to finalize the pack-list and buy the last items, prepare the solar equipment (which should turn up today) and do some test loading and test cycling. Over the last weekend I fitted the baggage to the bike and loaded it up with random heavy items including the new tent, sleeping bag etc. There should be no problem with space or loading weight. I also tested out the tent and it seems OK although some dampness came through from the floor. Is that normal? I'm surprised.

I've written the pack-list in Google docs and shared it for everyone so you can always see the latest version. Its organized into sections. Feel free to take a copy for your own use.

View the pack-list here.

I will go through it and weigh each section before I go and add that to the pack-list. Despite the sub 1kg PC, it won't be a light load! Here's the list of electronic equipment I'm taking.

12v AA battery charger,12v charger for Nokia phone,12V lead-acid battery (70W/hr),Adaptor cables and tips, Batteries 8xAA for camera, Batteries (2xAA for GPS), Camera case, Camera USB cable, DC-DC adaptor, Digital Camera + 2GB mini-SD card, GPS logger and holder, GPS serial cable and USB-serial adaptor, Li-ion battery (56W/hr), Mini battery tester, Multimeter (voltage, amps, sun, temp), N800 (spare web-access kit), Nokia headphones / adaptor. , Nokia phone (6280) + 2GB mini-SD card, Nokia USB cable, Passive USB mini-hub, Solar panel (25W) +cable kit, Spare mobile phone (voice only), Standard UMPC battery, Torch, Tripod, UMPC (Samsung Q1b) + Extended battery, case + keyboard, USB LED Lamp.

Later today  I should have more details about the solar panel/battery to post. Pictures too.

The new tent has arrived, the bike baggage, and finally, some warmer weather so I'm taking the chance to test things out in the garden. My daughter is nearly asleep in the tent and I'm perched outside with the Samsung Q1b and organizer pack resting on one of my panniers, a clip-on LED lamp a paraffin lamp and a bottle of Germanys best beer!

This is the perfect time to be writing because its dark and you only need the minimum of backlight on the UMPC. As I write this with WiFi on, I'm taking between 6.5 and 8.5W on the UMPC. Its also wonderfully quiet.

Behind me, in the spare bedroom, I have started to lay out everything I need for the tour. I've written the pack list and there are only a few more things to buy before I'm ready to go. I hope it all fits into the panniers I bought for the bike which, incidentally, is going to be the blue Kronan.

A few things that haven't turned up yet are the Lead-Acid battery and the solar panel. They should be here on Monday which will allow me to build the frame that the panel and battery will sit in on the back of the bike. I've ordered a could of panel meters too because I want to see what sort of drain each component puts on the panel.

One thing I was a little disappointed to find out tonight is that in theory, DC-DC conversion will cost at least 25% of my energy. That's rather a lot to be wasting just to transfer energy, especially if I charge the Li-Ion battery from the 12V lead acid battery. I could lose an hours computing time just in that process.

I'll finalise the packlist (V1.0) in the next few days and post it up. I'm also planning to do a video overview of all the equipment I'm taking, a picture-set of the bike and the charging setup and as many other images as I have time to take. The 9-day forecast is looking OK (not perfect, but OK) and so I'm quite confident that the tour will start at some point next weekend.