New Mobile MeshAP Prototype with Video

I made a new wireless Mobile MeshAP Prototype. It was design to be carried from place to place and run indoors. The unique element of this Mobile MeshAP is the ability to run from many power sources, such as, AC Power, Solar Power, or Car battery. The 12 volt power supply gives me this wonderful flexibility. The fact that it is also fanless and totally solid state ensure a smooth and silent operation.

But this really old news. Other people have made mobile MeshAP that even fly in the ultralight aircraft. What make this one unique is the splash page contain video. I consulted with Jon Anderson from Locustworld who advised me that I can embed the video into the actual MeshAP. The MeshAP has a solid state hard drive of less than 128 Meg so putting a video into the MeshAP can be a challenge.

The video itself can be viewed on line. As you noticed from the video my face is red. In the actual production the colour is normal. The actual video contains more information.

When accessing the Mobile MeshAP from any wireless device the splash page informs the person via video that they have various options. Because the video is embedded in the MeshAP there is no lag or hesitation. It pops up as fast as the web page can load.

With the Locustworld OS and one of the features of a "walled garden" I can give access to my registration page. Ideally I now can sell TV style advertisement using this portable device. This is why open source is so wonderful. I have developed hardware to sell TV style advertisement. What will I think of next?

New Wireless MeshAP Prototype - Outdoor PC

As I previously announce last year, I am working on a Prototype for an Outdoor Wireless Mesh AP. The objective of this new prototype was to have a high powered uplink node in noisy urban areas. After building and do some preliminary tests, I noticed that components had to be upgraded. A single board construction like a wrap board was out of the questioned. I wanted something like a Via board but a lot less expensive but as durable. What I found was an Intel Shelton board. This board is really a Mini-ITX motherboard. It contain advance features like DDR Memory and it is totally fan less. The entire construction is solid state.

The components on the board can be quickly upgraded such as, radio card, power supply, memory, solid state hard disk, antennae and operating system. The bonus was the actual OS. This board is very versatile can run many different OS. The enclosures being NEMA 4 gives the unit the ability to work outdoors. I guess what I made is a PC that can work out doors.

There are many advantages of having flexible equipment. The fact that the hardware can be upgraded as new technologies are available and the fact that one is not married to the operating systems is too an advantage.

Current OS is open source Locustworld and I will be using this software for testing purposes.

Unique features of this Mesh AP is the dual antennae running off a single radio card. I noticed that with two lower power antennae the shadows and null packets decrease. I'm still not sure if it warrants the extra cost but for now it is a powerful Mesh Access Point. Currently the radio cards are 300 mW unfortunately I can only test up to 200 mW (software does not have a 300 mW power setting.)

Other unique features is the smallest PC power supply. This unit is totally awesome. It takes very little space and it has been working flawlessly.

Also I have locked down and the components to prevent them from dislodging themselves.

I have put some pictures and description of unit on my web site.

Myth busting on Outdoor Computers

I have been working with Open Source Wireless Mesh for the last two year developing techniques and hardware for using Locustworld OS. I have read a lot about how to make an outdoor MeshAP or in other worlds a Mini PC that works in the out doors. What I have read is a lot advice with no practiced experience.

The equipment I use is a mini-itx form factor using VIA chip sets or Intel chip set. Nevertheless it is basically a PC with out a keyboard, mouse or monitor.

Now after two years I can say that with the experience I gain one that the manufacturer recommendation is ultra conservative. Many of my units only have 0 C degrees lower limit. I have running my units in in -20 C units continuously with no problems.

Next temperature extremes and condensation. I was told that you have to put an inert gas into you units to prevent condensation happening. Well I didn't. I had in one day temperature variations of 25 degrees. I believe it was -15 C one day and the next it went to 10 C. I check the interior of my enclosures and found no water drops or condensation. And even when the temperature was 10 C one day then -3 the next the units show no frost build up on the walls. Myth busted condensation is not a problem if container are water tight or have fans circulating the air with in the enclosures.

My enclosures vary from Nema 4 to a simple Metal PC case wrapped in plastic. First to the PC. The PC I using are old Compaq slim lines prior to HP take over. These units have been abused. Rain, Snow, Heat, Wind, Vibration and they contain fans, and harddrives. I was told that these unit would be the first two go. And people were right. I put one unit in a plastic case and the gentleman who took care of this unit on his balcony used it for a drink holder. He spilled a few. Nevertheless Compaq do not like swimming in coke cola. Two years later Compaqs are still running strong with a simple plastic wrap. Myth busted PC can run in all temperatures as long as they are running continuously.

The next are Via boards and the Intel boards. Both are excellent and can hold up to extreme temperatures both cold and hot. With enclosures that have fans or no fans the temperature is roughly the same. So having units with moving parts like fans and hard drives have very little effect in a continuous operation.

Most people believe that a total solid state is better with low heat. I have not seen any advantage to solid state other than reduce failure on moving parts. The total solid state is expensive.

Conclusion: More observation is required.