Good to hear! Look forward to seeing the updated coverage map.
Do you think that UMESH 14 is still need? Would it be best to save the resources for extending the network further?
Itās possible that we could get by without it. But then you lose the redundancy it provides you and vice versa. The more options you have with paths the better for getting through reliably. I guess the answer really is weāre doing better with it than without it.
Then thereās the fact we are almost to the point where the build-out is half-complete. Weāre getting fairly good coverage of the city with fewer than half the estimated number of nodes, so I suspect our resources are in good shape. There are a few gaps, but even more opportunities, so adding the next 16 will provide a good demonstration of how much node density affects performance,
I also canāt help wondering if 14 will solve the riddle weāre having with Andrew Wās relay node. It initially worked great, dropped out when UMESH 1 and 2 went down, but now refuses to perform much better with them both back online. Weāll get it figured out with some more work surveying.
Hereās a shot of the screen while I was a couple of miles south of Urbana, where I managed to hit the home node from Barnhart Prairie, a restoration prairie plot of about 100 acres, via 5 nodesā¦
1 Like
We sorted out Andrew Wās reluctant node. I thought the power had died from the external battery somehow, because it worked for about 24 hours and then quit, but it showed charge and insisted with the 3 flashes that it was in relay mode.
But it wasnāt relaying.
We swapped out the whole node and then we tried the troublesome GTM device as the paired device, where it was fine. Never had that happen before.
Then there was some test messaging and celebrating the link working again.
In the morning UMESH 14 goes up, the third relay this week. Then I may start on a coverage map update.
Any luck getting UMESH14 installed today? I should have my goTennas when I arrive home today, Iāll send a shout to test.
I chose to pause things until we could have someone with the skills to use a safety rope assess. The roof pitch is like yours, but itās a lot farther down at the eaves. Plus this was the first day back on the full dose of a Rx Iām taking and it turns out that it is its side effect that is making my legs ache. The combination of those two issues means weāre waiting for someone with the proper skills to evaluate things. You gotta know your limits.
Iāll update as soon as we have solid info to pass along. Will be looking for your test pings and messages.
Weāre trying to mesh a bit this afternoon, but conditions are ugly. My correspondent is close to max range in term of hops, a little less than 1.5 miles beeline. Sometimes we get through in 4 nodes, sometimes it takes 5 - this afternoon. Yesterday, I had one link where it was just 2 nodes, but that was from the Meshmobile. That was in late evening. I suspect weāre getting some solar effects, although that seemed unlikely at 900 mHz. However, this time of day seems to have something going on. Could also be highest temp time of day inside the cases housing the nodes. This will also pass and weāll have time to prepare for next yearās summer maximums.
Also waiting to see if UMESH 15 got the units he was eagerly awaiting. We hope to hear from him soon.
Things are happening with mesh out here in the cornfieldsā¦we can see the water tower of the next town, St. Joseph, from the roof of UMESH 15ā¦hmmmm. I can tell you weād support efforts at this end for a directional link experiment to see if it was possible if someone in St. Joseph wanted to give it a tryā¦
And once we get a couple of more nodes up, i.e. 16, then weāre halfway to our overall goal of 32 to mesh Urbana and itās probably time to reach out to agencies and interest communities with a potential need for UMESH in case of emergency
Nope I didnāt get the units. Looks like USPS wanted a signature. Hopefully tomorrow!
Bummer, well Friday should do it. We keep our ears up.
I suspect part of the issues weāve had with UMESH 13 is due to the placement of the solar panel on the case. Keep in mind that the goTenna Mesh is in the upper left corner of the case in these pics and the whole thing is typically mounted at a roof ridge, draping the case/solar panel/etc on one side and a counterweight on the other side.
In this pic, we see the original configuration, with the solar panel mounted high (towards the caseās handle) on the board that attaches to the case to mount the panel on as well to keep the case upright as a āleg.ā
In this pic, I have relocated the panel toward the bottom of the mounting board.
In this relocation, the solar panel no longer partially shields the goTenna Mesh inside the case from signals arriving through the top of the case. Iām not sure how big a factor this was, but by making it no longer a factor this solves that part of the issue, if it was an issue.
There will be short delay in dealing with getting UMESH 14 up, so people in Savannah Green have UMESH 15 for nowā¦ When I was on S. Smith Road yesterday, the mostly 2-story structures there were blocking a good signal at street level without 14 being online, so things are a little iffy right now in getting a mesh signal. You likely have to move around to find it on the west side of SG, somewhat less so on the east side of SG. UMESH 14 looks doable, but our climber always has final say on things to ensure safety.
This also delays a look at the potential new near-campus location. This one is also his call, as it will be a rather challenging climb - and may not work if we canāt establish a safe route on what will be a fairly technical climb. The location is excellent, so I donāt want to sound too discouraging, just emphasize the care we take where danger might lurk.
I will update as soon as we know more about when we will be able to implement these service upgrades.
I got my goTennas just now. Tried sending a message to you Mike, right under UMESH15, but it says delivery uncofirmed. I also sent a shout (which I assume should go to everyone within range of any mesh in town). Do you think that UMESH15 isnāt quite talking to the rest of the mesh? Is there anyway I can test UMESH15 to make sure it is still online?
Try again now, to me. I was up dealing with a power failure on the home node just now so probably missed because of that.
A Shout only goes directly to any GTM in range. It does not hop.
An Emergency Shout relays to everyone in the network, but thatās like calling 911 - only for emergencies.
Will be watching for your messages. Iāll be here for the next half-hour, then need to run an errand and will then be back here after 4pm
Iāll try to run back through SE Urbana to check how it should be doing.
Oh ok, I didnāt know that about shouts. I sent you a message, looks like it still didnāt go through.
Now weāre cooking. Turns out both our nodes (1 and 15) failed due to them being the oddball 18 volt ones. Iām sure glad they were free (insert long story about mis-shipped merchandise from overseas here), because theyāve cost plenty of time and effort on getting them right, but they still arenāt ready even after running reliably for months, one on the roof and one on the node farm table that became 15ās first node. Post-mortem pending, but seems definitely power-related since it came right back to life once it restarted.
So I took my last good olā reliable 10 W ābig panelā node down and 15 and I climbed up to swap the boxes out. At first we couldnāt get back here to one from his roof. Then we tried the location sharing and managed to get that working. This eventually proved the new node replaced the bum one. Pretty good range, but will be much better once the leaves dropped. We need a few more nodes in Scottswood and along East Washington to make this work best. Anyone? But it does work 
This post is about a refinement in how weāre installing and keeping in place our relay nodes. They are packaged in thick plastic cases that are Harbor Freight knock-offs of more expensive one like Pelican, etc makes. With the attached solar panels and a battery, they typically weigh under 5 pounds. Their low profile helps further reduce any tendency for the wind to catch at them. Weā've yet to see any movement despite their experiencing 50+ mph winds and at least one small tornado.
So how do we attach them to the roof?
We donāt, they just sit there.
How do we do that?
Generally we drape a node over the roof ridgeline paired with counterweights that attach to it with UV resistant poly rope. Like thisā¦
In the case of my early nodes, I used old window sash weights. I had a bunch, they have a convenient holes to attach them, theyāre heavy as heck, you get the point.
The big design flaw? Theyāre round, so if one gets away while youāre on the roof or after installation, they will do what gravity bids them to - roll right down and fly off the eave - and hopefully land safely.
I put a lot of effort into being careful to check any install using them so theyād end up in a bush or some other safe place if one did get out of control. `
Then I came up with a better idea: casting a custom weight out of concrete that wouldnāt roll away.
Disclaimar: Itās very much up to the installer to make this safe. It could still hurt someone ifr you managed to drop it over the roof edge, which most likely would be someone helping you. Like with all antenna installs Be Safe First. Assess your local environment because you are the sole judge of fitness for your purpose. Where high winds exist, you should bolt things down unless youāre sure things will stay in place for anticipated climate extremes, which are worsening whatever you personally may believe to be the cause. In other words, itās on you to safely deploy and use these for temporary installations while building a network, which may take several years. Be sure and revisit your safety culture at least annually, which is where this improvement came about.
OK, you start with some hardware, a couple of 6" x 3/8" eyebolts and some sort of bar material of channel like this aluminum 3/4" plywood edging.
The complete Mold and a finished counterweight
1 Like
Found a good illustration of the sort of roof the over-the-ridge design works well on, as it shows the angle of repose.
A few more pointers.
These are typilcally installed on asphalt shingled roofs, so that grittiness is part of why it works for us. If you had a slate or metal roof, then it would be rather slick and the new counterweight would be just as likely to slide down as the sah weights are.
The angle is important. You can go a little bit steeper and the new counterweight wonāt slide on itās own. Increase the angle substantially and the new counterweight would slide anyway.
More isnāt necessarily better. You want a counterweight that weighs about the same as the cased relay node that hangs over on the other side. No need for great accuracy here, just need to keep them all about the same so there arenāt vast discrepancies, either.
1 Like
Finally, after many false starts, weather and scheduling issues, we got UMESH 14 up. I hope to have a few pics for you soon, as the view was said to be delicious from up there. Itās the 14th node, despite there being nodes up to UMESH 15, because UMESH 8 still hasnāt gone up for reasons too lengthy to go into at this point. That puts UMESH at 14 out of 32 proposed nodes or 44% of our ultimate coverage.
Given the significant height advantage of 14, it should help coverage significantly in southeast Urbana. Right now though, I need to do some surveying as we seem to have lost an intermediary node in the last day or so. We can ping down there and back, but messages arenāt working, kind of weird but thatās how it is right now. If itās not helping you yet, be patient and weāll get this figured out in the next day or so.
Thanks to Mike, the new node host, for being so patient with us getting our ducks in a row and to Nine, the worldās greatest mesh climber!
We recently received a shipment of battery packs and are expecting a new group of solar panels any day now. There are a number of prospective node hosts interested right now, so we hope to announce more relays going up soon.
1 Like
Yes, fixing UMESH 2 fixed the spotty coverage out east. I was able to get home from Savannah Green North (in 3 nodes) and South (in 4 nodes.) Better yet, a check from the parking lot at Riggs Brewery got me home in 4 nodes.
Had a chance to do some messaging back and forth with UMESH 15 Wednesday evening after things were done. Heās one of several hams who are node hosts. With 14 going online and 2 getting a new battery, messages were just zipping back and forth. He was really impressed by the faster speed provided by the shorter link that putting UMESH 14 made, instead of taking the long way around as was happening before. Only one less link typically occurred, but the speed of getting a confirmation back was substantial.
As I explained, this was also likely facilitated by improved link quality that the new shortcut via 14 provided. Since the Aspen Grove meshware will try twice in sending messages, if the first message gets through more reliably than before, it noticeably speeds things up when better links become available connecting the same two endpoints.
Good thing we got this accomplished Wednesday when I was feeling a little better after feeling under the weather on Tuesday. Because I was really sick Thursday - until my gallbladder came out with emergency surgery Thursday evening! Iām feeling much better now, but this may slow things a little over the next couple of weeks at least as far as any climbing goes. Otherwise Iāll be back to what passes for mostly normal in a few days.