Urbana, Illinois, Now a goTenna Mesh Ambassador City!

[MikeL]

UMESH 9 and I went out doing some survey work today. First we poked around UMESH 7 and it seemed to still be suffering a bit. Checking on it proved it was happily burbling away. We did some more poking around and exchanging ideas about why we still seemed to have connectivity problems to the north. Originally, I’d blamed this on the leafing out of trees last spring. Seems to be a part of the problem, but not all of it.

Thinking more about what changed tonight, all of the sudden it hit me like a wall of bricks, because it is I think literally caused by a wall of bricks. UMESH 7 had a good view of UMESH 6 when it went up. Since then, a major school reconstruction project across the street in the direction of UMESH 6 added a large building - brick-faced - much nearer the road that is about 30’ tall! It may also obscure 7’s view of UMESH 3, although I’ve never been sure 7 and 3 are close enough to link.

Thus I’ve come to the conclusion that the solution here is in-fill development, which opens up multiple, optional routes for messages to take. We brainstormed some possible locations to help with that. In particular, we could use node hosts in locations near Florida west of Anderson St. However, we are interested in new node locations all over Urbana, drop us a line and we’ll consider your location seriously so long as it adds to or improves service.

[MikeL]

It’s been awhile since the last update, but the weather has been atrocious. Still, when you’re a node farmer, there always something to be done, like planting…

After the requisite time passes, you harvest useful things that grow…

[winter garden courtesy of Aunt Barbara at UMESH 3 ]

With a little more cultivation – and other parts for new nodes flowing in - battery packs, solar panels, cables, a couple of solar controllers for special installs - these newly-harvested node cases will soon mature into another batch of UMESH relays.

This spring is starting to ramp up and we hope to soon be placing new nodes around Urbana to expand the more than 5 square miles of mesh coverage already available. Drop us a note if you’d like to get involved.

[MikeL]

Nodes are sprouting up…

And Wally is doing his stretches…

[MikeL]

Three more nodes popped up in the last couple of days. While it’s still gloomy in this pic, sunny days are just ahead.

Two I built used a smallish “10 W Maximum” output solar panel, which is roughly half the size of the 5 larger panels in the pic above, which are also rated at 10 W but with no “max output” qualifier. While I’m still skeptical, initial testing showed they do charge pretty well.

The 10 W max panel (outputs 5V):

The 10 W panel(outputs 5V and 12V)

Then there was the oddball, this 18 volt output panel I ended up getting 2 of free because they sent the wrong ones and then resent me the right one.

For comparison purposes, the case used with all three is the same size.

The twin to this one is supposed to be powering UMESH 1, but I didn’t get the MPPT charge controller settings right the first time, so went back to house power, but with an added battery backup. After I get this one tuned, I’ll apply that to the one on the roof.

These 8 nodes will double the number in service currently in UMESH, providing for both expansion of service and fill-in to increase service reliability. One is going up at UMESH 8 and several more have potential hosts, but if you want a node now is a great time to ask for one.

The two 10 W “max” “mini-nodes” utilize the Voltaic V15 battery packs that may have a bit too small capacity for those dark winter days, but they’ll work for now and the battery pack can be replaced before winter with our standard one. I’ll not put them anywhere too difficult to get to until that’s done.

[danielagotenna]

Wow @MikeL, you really have quite a “mesh farm” going.

[MikeL]

Daniela,
Heh-heh, yes we do. The soil is fertile, some of the best in the world. Things just sprout up once you plant them!

As you can see, I’ve been stashing parts and supplies away all winter. As you know, this group of nodes will, once installed, double the number of active nodes in UMESH, putting us at 50% of the goal of completely meshing Urbana. If I can get this group settled into homes with node hosts by the middle of the summer, that leaves enough time to repeat the effort through the end of the year, which would get us to 75% of build out, but I should not get too far ahead of myself.

I don’t want people to fear we won’t have a node in this group for them. Most are as yet still looking for a cozy spot in the sun on someone’s roof. Let us know if you can give them a nice perch that will not disturb your roof, but will help extend free wireless mesh text messaging coverage to thousands more in Urbana at the same time it improves current service.

There is an important technical component to the node farm that should not go unmentioned. I’ve found that building nodes, then letting them run for awhile down where they’re easy to reach helps sort out small issues or significant technical faults before they go where it’s harder to check and correct issues. Everything is hand-built with inexpensive off-the-shelf components by humans and it’s been a sometimes trying learning curve. We’ve got most issues sorted out by now, as well as gained some useful knowledge of fail modes and how to prevent them, so this period is more about ensuring that a node is ready to run for six months or more before we just happen to check on it to find everything’s good, as most do now.
Thanks for the appreciation!

[aspexin]

@MikeL Are you able to measure internal temperature in the box? How high does it get during the summer months? At least for the existing nodes that you have out there.

[MikeL]

Great question, I was just doing some thinking about our work on that…

Short answer: It works.

The design we use for the case and how it’s arranged, painted, etc. I don’t believe we had any failures that I can attribute to heat. I suspected that was a problem a time or two, but eventually found other issues that were the likely culprits. I was thinking just now that I should get some help and make a short video of how the design came about and the components used. It’s not so much the specific brands, but what the items need to do and how to make it work reliably.

The longer answer: The hosts of UMESH 6 installed a Bluetooth environmental monitor recently and are collecting data, at least once the bugs get sorted out to make this an automated task. This is nothing to do with the GTM, just the environment it operates in, but should be informative. I’ll report results when available.

[Briann]

Any reason why you’re going completely solar powered? While a self contained node makes alot of sense as far as setup and placement goes, wouldn’t it be pretty easy to run a hard line to the roof and keep that alive with a standard 110 volt UPS inside the house?

[MikeL]

@Briann

Any reason why you’re going completely solar powered?

Multiple ones.

Practical marketing: It’s a lot easier to convince someone they want to host a node if it requires no electrical hookup, plus other advantages like no roof penetrations, discrete appearance, etc.

Environmental: Sure, it wouldn’t burn a lot of coal to make the electricity needed to power a node. But if you’re at the cutting edge, then you need to act like you care about the future or you undercut your message, especially with younger people. As it turns out, the two most active node hosts here have invested in significant home solar installations. Being up on the roof is normal - and some can see the future from there. Enough philosophy for now, but it just matters to those of us working on UMESH.

Historical: I lived through the 1990 Valentine’s Day ice storm here - then through another 10 days without power. I want a network that can be depended on in times of emergency. 'Nuff said.

[aspexin]

I went through Hurricane Fran in 1996 in Durham. We were without power for 8 days. So definitely having solar is important to keep nodes up. Although I don’t know where Fran would have thrown those nodes if they were not secured. So those in the hurricane states need to think about that.

[MikeL]

Good point on that. The Midwestern equivalent are tornadoes. They sometimes occur with hurricanes, but are plenty destructive all on their own.

For the most part, my nodes just sit there, aided by some attached weight. A couple are tied down. Where they go is generally the roof peak, so unless they’re near a chimney, nothing much to tie to up there. We’ve had 50+ mph winds here several times this spring and it’s bothered none of the nodes. Tornadoes are around 200mph so a big intensity difference.

All the UMESH nodes are housed in Harbor Freight Apache cases. The solar panel might get torn off, but no matter how intense the wind I doubt it could hurt this case. It would need to be located in some cases. But for destruction like that, tying it to something that will get blown away doesn’t gain you much. Sometimes, it pays to just be ready. Eventually when the build out is complete here, I would like to then build a small set of 3 or 4 nodes that would be available for temporary operations or to serve as disaster replacements.

The 2 smaller nodes I built the other day are a good example of what I’m after. They are the center 2 in this pic.

The small panel may not work through the dead of winter. That’s not tested yet, but they seem plenty good for everything else so far. They also fit the case footprint really well, making for a nice compact package once it’s all assembled. One thing not obvious so far, but important is that I’ve started attaching the board that holds the panel to the case so that it allows the case to sit steady and upright so you can transport the node by the handle easily.

[aspexin]

Can you give details on the build out? I’d like to replicate.

[MikeL]

goTenna and I discussed what we thought it would take to cover Urbana with reliable mesh service. The number ended up being 32, a bit larger than I originally thought would be needed. It may still be, but I think I was thinking too much like an old radio guy, trying to maximize the coverage from each node.

But while this early iteration works with some moving around to gain better vantage point on a signal for us old radioheads, it’s better to increase the amount of overlap in signal coverage for most users. That gives you several alternative “looks” at different nodes in most cases. That’s why we’re looking for in-fill hosts as well as hosts to expand the over all footprint of coverage.

I now think the 32 number is pretty good as an endpoint for the buildout here. That will provide substantial coverage before we reach 32 nodes installed (see maps above) and coverage that extends beyond Urbana in many cases ao that those in Champaign and the county beyond the Urbana city limits can eventually connect with UMESH service.

[MrTSolar]

@Briann

Any reason why you’re going completely solar powered?

I agree with Mike’s reasons. Another reason is that it removes some of the hypocrisy of having a decentralized, self-forming comms network at the mercy of the power grid. If the nodes drop out 24 hours after the grid goes down, then where are they any more resilient than cellular? Having had a nasty windstorm blow through our area a couple weeks ago, cell towers for the most part stayed up (although T-Mobile totally dropped out for a few hours). Most of my township was dark for 3 days.

It also helps with the perception of solar power in general if the nodes have good up-time.

[MikeL]

Time for a farm report…

I improved the ergonomics of farming, making it easy to check multiple nodes daily, by setting things up on a portable table.

5 of the 8 nodes have been just fine all along, charging and relaying without interruption. I had a cable issue with another, as it wouldn’t reliably charge a unit even when fully plugged in and the GTM responded appropriately to the benefits of a shiny new charge cable. In another I have some odd issue with getting the battery pack in the right mode, solved by swapping out the pack. And the single 18 volt freebie panel remains a work in progress. It is charging pretty well, but dropped out several time still, so I haven’t got the charge controller set right yet. As of right now 7 of 8 have been up for the last 10 days or so, with 5 going the distance without fail.

I have some experience and do a lot of cross-checking to be sure things are right, but things are hard to get perfect 100%. It’s much easier dealing with the small issues that crop up now, on the ground, before anyone depends on them. Farm those solar nodes for best results.

Now a couple of observations…

This is what the cable entry looks like on my recent nodes.

It’s just a 1/2" hole. Nothing seals it and the charge cable going inside fits very loosely. They sit out in the rain and I’ve detected no water intrusion even when the case is flat on its back. I think the interior foam just inside tends to limit water getting inside. Most installation feature this port being at the lower end of the case pointing downhill and usually on the lee side of things from the prevailing wind direction. This position locates the case’s standard pressure release valve at the upper end, permitting a limited flow thru ventilation. Bottom line: You may not really need to seal up such an opening if it is located properly.

Another issue is one I haven’t completely solved yet, sealing the USB charge cable connection. I use the red translucent 3M made for temporary restoration of busted tail and brake lights. It is somewhat UV resistant and seems to stick well - sometimes.

Sometimes, it’s not so good after a few weeks outside.

I have used clear packaging tape to supplement the stickiness. I think I gained some advantage by using a heat gun to help make the folds more set in position.

So, what to do about all these homeless nodes? I’m having a node adopta-thon at the 2600 meeting at the Lincoln Square Food Court from 5-7pm on Friday, April 5. Stop by, meet a node, and see what all the fuss is about, as well as getting the chance to learn a thing or two about computer and other forms of data security if you are so inclined. Real good people and they have been an immense help with getting UMESH going.

[MikeL]

One new node up and one to be done after the tenants have been notified. So a pretty good day and a shout out to Dave, who helped set this up.

UMESH 10 is up on the south side of Yankee Ridge. Connectivity to the rest of the network is uncertain for it right now. Nonetheless, this is a great location and it should connect to the rest of the network once we add a few more nodes. In addition to serving Yankee Ridge, it should provide good service throughout Meadowbrook Park.

Another relay will go up soon, this one in the critical Florida Ave corridor where we run into issues once the trees leaf out. It’s located just east of Blair Park, an ideal location to overcome that issue and better link the north and south of UMESH. There’s a tower to mount it on. It will definitely improve connectivity to UMESH 9.

[MikeL]

Got approval for the Blair Park area node (will be UMESH 11) to go up.

Also did some surveying and we’re having a power issue again at UMESH 6 and potentially the same at 9. I will replace both with tested nodes from the node farm, then test the faulty nodes to determine the issue and fix to go back out. Much easier than trying to analyze and fix while on the roof.

So after we get that done, then UMESH 10 may actually be connected without further ado. If not, we’ll work at it some more. I think UMESH 11 is the key here, as that will link the north and south node groups much more solidly and hopefully prevent seasonal leafing from being an issue, too.

[MikeL]

UMESH 11 is up! It solidifies reliable connectivity in central Urbana by providing a better link between the north and south node clusters. A solid mesh signal all the way from way down south at Meadowbrook Park all the way north to past Five Points is now easy to find. Not enough time for full testing tonight but will do more soon.

For now, UMESH 10 is isolated from the network, but we hope for better soon. As noted above, it does serve neighboring Meadowbrook Park, Clark-Lindsey Village, etc.

[MikeL]

Got a chance to do some surveying today and found the mesh was substantially strengthened by our efforts yesterday. There is actually a good signal into Yankee Ridge at another location near to UMESH 10. I expect a small relocation on the roof may solve the lack of connectivity to the network there. We’ll experiment some with this after I get done with taxes.

For now, a couple of pics of UMESH 9 climbing up to help put others on the air, first swapping out the node at UMESH 6…

Then putting up UMESH 11.

The small dish located just above where the waterproof box holding UMESH 11’s GTM is attached is a vintage CUWIN item, a 5.8 GHz parabolic dish. It was an attempt to link to the IMC in downtown (location of UMESH 4.) It was said to work episodically, but generally when the leaves were off the trees. UMESH now solidly covers from somewhere south of Windsor Road (Yankee Ridge) to Crystal Lake Park in the north - but in about 6 hops (via 5 nodes.)*

Notable in looking at the Mesh Map is that there are 16 other nodes (besides the relays associated directly with the Ambassador Cities Program) now operating in the C-U area. That’s the ones that someone has put on the node map, which means there are likely many more using the mesh already. One of them is one of my mobile nodes, but the rest seem to have been submitted by people from 8 to 80…, well, who knows? but they are located in a wide range of diverse demographic zones, serving a significant population of nearly 40,000 people who now either have mesh service or who have it nearby - and we hope to soon bring it to them.

• Was thinking more about this tonight and realized that we are just now starting to see leafing out, so I shouldn’t count too many chickens before they are hatched. That said, the signal is more reliable and easier to locate than it has ever been in central Urbana - and beyond. How far we get depends on the next hosts. It’s rather amazing how much doing that can improve coverage across the community. It’s not only extension outward, but also infill, as it adds redundancy and reliability.

Thanks for the appreciation. I could not do it without the help of many others, one of who could next be - you!