Too Many Experiments

I normally would write about the latest experiment however I have been busy running two many experiments at the same time. I have purchase upgrades to the solar experiment however I have not had the time to put it together and run it. The new experiment is the 3 new MeshAP that I have built. The are currently 2.4 Ghz and 900 Mhz solutions. The 900 is running a special card Antheros type card. So far I have not much success and this is due to the lack of tools that are available to me. Hopefully I will get it to work and write about.

900 Mhz has a better penetration through trees and buildings than the 2.4 Ghz radio cards. The 900 Mhz can be go distances of over 25 miles or 40 Km. The through put runs at about 54 Megs. So as you can see as a backhaul solution this would work pretty well; again however, in theory yes but in practice one can assume only half the throughput and distance.

New telephony call Trixbox is the Asterix replacement. I have plans to up grade these services but just need a few hours to get is running and integrating into the wireless mesh.
Oh, as part of the Trixbox I was planning to build a supercomputer i.e. small parallel system that would help in the computation of the sugarCRM system that goes with the Trixbox.

All of these are great project but I just don’t have enough time in the day to work, play and have fun.

Life cycle on radio cards

My assumption of any electronic product is that they should last forever. With this in mind I have some statistics to share with you. Back 3 years ago when I started to create my wireless mesh I used Prism based PCI radio cards. And in fact I still used them today however; they are more MiniPCI with ether a PCI adapter. Nevertheless the radio cards that I once bought were all less than 100 mw. In today market 200-400 mw is more the standard. I expected that these radios with no moving parts would out last me but what I finding is that radio cards start to fade.

What I mean by fade is the radio cards do not seem to give off as much power as when I first installed them. My thinking is that they should until they burn out. Most of the cards I have should start to burn out.

What happens is the links start to falter and the units tend to go deaf. They became hard of hearing. What I try to do with the radio cards is to put more power through them by changing Wiana power control to something like 200 mw. They don’t actually give off that power but the cards seems to be running a little hotter. This is more my impression rather than fact. In addition the majority of these PCI cards are running in an outdoor environment. This two can cause problems.

The reason that I sharing this information with you is if you use a commercial MeshAP or router seems to be a shelf life to the radio cards. If anyone has more information on this subject please email at don@moskaluk.com

Environmental change create better connectivity

I have been having a number of problems with one installation that I need to share with you all. I have a MeshAP that is sitting on the 7th floor of a balcony. The unit is wired to a Macintosh and has been working flawlessly for the last 3 years. In the last 9 months the problems started to occur. Connectivity was not solid and I lost connections constantly for hours at a time. There were a number of problems that occurred. First someone PC that had a virus tried to constantly connect to the network. Second a 2.4 Ghz telephone system would constantly try to run interference.

The result was that I change the equipment, antennas, radio card etc many times with the same problem. Finally I figure it was the environment. Across the street was a new high rise apartment building was being built. It was finally completed about a month a go and the problems have not escalated any further. I change the antenna location from one location to another. What the result was is a firm and constant signal. The inference went away and that the connection was once again rock solid. I tested the old equipment and it too is now rock solid in the new position.

I won’t believe it. There is no reason for this to occur however, when see the reflection of many large building the antenna position was probably located where a lot of echo and noise was occurring. Imagine throughing a bunch of pebbles into a pond and seeing where all the ripples will converge. I believe that the reflection create these super crest that interfered with antenna. Simply moving the antenna away from this wall or a reflecting service improves the quality of service.

I also have seen a better connection between mesh nodes as a result. I will reposition most of the problem MeshAP to see if this can give a better connection.